Portal de Periódicos da CAPES

Phytochrome: If It Looks and Smells Like a Histidine Kinase, Is It a Histidine Kinase?

1997; Cell Press; Volume: 91; Issue: 6 Linguagem: Inglês

10.1016/s0092-8674(00)80458-4

1097-4172

Tedd D. Elich, J. Chory,

Plant Molecular Biology Research

Like all organisms, plants continually monitor and respond to changes in their environment. Being photosynthetic, plants are particularly attuned to ambient light conditions. Accordingly, they have evolved sophisticated light-detection systems involving multiple photoreceptors that regulate diverse developmental responses. For example, plants use red and far-red wavelengths in incident light as indicators of time-of-day, seasonal changes, and the proximity of neighboring plants in order to regulate such processes as seed germination, vegetative growth, and flowering. The detection of red and far-red light is mediated by a family of photoreceptors called phytochromes that have the unique ability to exist in two photointerconvertible forms with distinct conformations and photochemical properties: Pr, a red light absorbing form, and Pfr, a far-red light absorbing form (3Fankhauser C Chory J Annu. Rev. Cell Dev. Biol. In press. 1997; Google Scholar) (Figure 1A and Figure 1B). Pfr is generally thought to be the active form since classic phytochrome responses are red light inducible and far-red light reversible. Phytochromes are large soluble proteins of approximately 120 kDa that until recently were thought to be unique to eukaryotes. These photoreceptors are encoded by multigene families in all organisms examined in detail. In Arabidopsis, for example, there are five family members, phyA to phyE. Phytochromes exist as dimers in solution with each monomer folding into two major domains separated by a protease-sensitive hinge region (Figure 1D). The N-terminal chromophore-bearing domain is sufficient for normal photochemistry while the C-terminal domain contains regions necessary for dimerization and biological activity (3Fankhauser C Chory J Annu. Rev. Cell Dev. Biol. In press. 1997; Google Scholarreferences therein). Despite intense scrutiny, however, the primary signal transduction pathways activated by phytochrome photoconversion have not yet been defined. Almost 40 years ago, without direct evidence, it was speculated that phytochromes may act as light-regulated enzymes. Although this simple and obvious hypothesis subsequently grew out of favor, it has now reemerged as the odds-on favorite due to a recurring connection between phytochromes and protein kinases that can no longer be ignored. In the mid-1980s, serine kinase activity was detected in highly purified phytochrome preparations from plants and speculated to be intrinsic to the photoreceptor (reviewed by12McMichael R.W Lagarias J.C Curr. Topics Plant Biochem. Physiol. 1990; 9: 259-270Google Scholar), but this controversial proposal has remained unresolved most notably because phytochromes lack the consensus sequences that define eukaryotic protein kinases. Subsequently, however, it was pointed out that the last approximately 250 amino acids of the phytochrome C terminus show sequence similarity to transmitter histidine kinases of two-component systems, proteins at that time thought to be unique to prokaryotes (16Schneider-Poetsch H.A.W Photochem. Photobiol. 1992; 56: 839-846Crossref PubMed Scopus (78) Google Scholar) (Figure 1D). Furthermore, new analysis indicates that the C terminus actually contains a second transmitter kinase–related domain adjacent to the hinge region (10Lagarias D.M Wu S.-H Lagarias J.C Plant Mol. Biol. 1995; 29: 1127-1142Crossref PubMed Scopus (81) Google Scholar, 19Yeh K.-C Wu S.-H Murphy J.T Lagarias J.C Science. 1997; 277: 1505-1508Crossref PubMed Scopus (429) Google Scholar) (Figure 1D). Interestingly, this second domain contains two PAS repeats (10Lagarias D.M Wu S.-H Lagarias J.C Plant Mol. Biol. 1995; 29: 1127-1142Crossref PubMed Scopus (81) Google Scholar) that overlap with newly identified motifs, the S boxes, that occur in a large family of energy-sensing proteins including other histidine kinases (20Zhulin I.B Taylor B.L Dixon R Trends Biochem.Sci. 1997; 22: 331-333Google Scholar). Finally, the connection between phytochromes and histidine kinases was recently cemented with the exciting discovery of a cyanobacterial phytochrome (6Hughes J Lamparter T Mittmann F Hartmann E Gartner W Wilde A Borner T Nature. 1997; 386: 663Crossref PubMed Scopus (285) Google Scholar, 19Yeh K.-C Wu S.-H Murphy J.T Lagarias J.C Science. 1997; 277: 1505-1508Crossref PubMed Scopus (429) Google Scholar) that shows light-regulated histidine kinase activity (19Yeh K.-C Wu S.-H Murphy J.T Lagarias J.C Science. 1997; 277: 1505-1508Crossref PubMed Scopus (429) Google Scholar). While we are not there yet, it appears that we are rapidly approaching a breakthrough in our understanding of plant phytochrome signaling mechanisms. To try and put these recent findings in perspective, we will review more fully the connections between phytochromes and protein kinase activity. Chemical analysis of purified oat phyA has demonstrated that phytochrome itself is a phosphoprotein containing up to 0.5 mol phosphate per mole phytochrome monomer (11Lapko V.N Jiang X.-Y Smith D.L Song P.-S Biochemistry. 1997; 36: 10595Crossref PubMed Scopus (59) Google Scholar). One site of this in vivo phosphorylation was mapped to the serine rich N terminus, most likely Ser-7 (Figure 1D) (11Lapko V.N Jiang X.-Y Smith D.L Song P.-S Biochemistry. 1997; 36: 10595Crossref PubMed Scopus (59) Google Scholar). Interestingly, the serine-rich N terminus of phyA had been previously implicated as a possible target site for regulatory phosphorylation in two studies where the first 10 serines of either rice phyA (17Stockhaus J Nagatani A Halfter U Kay S Furuya M Chua N.-H Genes Dev. 1992; 6: 2364-2372Crossref PubMed Scopus (99) Google Scholar) or oat phyA (7Jordan E.T Marita J.M Clough R.C Vierstra R.D Plant Physiol. 1997; 115: 693-704Crossref PubMed Scopus (39) Google Scholar) were mutated to alanines. In both cases, the resulting mutant proteins exhibited hyperactivity compared to wild-type phytochrome. Taken together, the above results suggest that phosphorylation of an N-terminal serine acts to attenuate the biological activity of phytochrome. This scenario is reminiscent of the down-regulation mechanism for the photoreceptor rhodopsin where phosphorylation of light-activated rhodopsin leads to the binding of an inhibitory protein, arrestin, thereby interfering with G-protein activation and terminating the light signal. The existence of potential regulatory phytochrome phosphorylation in vivo leads to the obvious question of the identity of the responsible kinase. Surprisingly, as alluded to at the outset, there is evidence that phytochrome itself may be the source of this activity. Lagarias and coworkers were the first to report on the existence of polycation-stimulated protein serine kinase activity associated with highly purified phytochrome preparations (see12McMichael R.W Lagarias J.C Curr. Topics Plant Biochem. Physiol. 1990; 9: 259-270Google Scholarreferences therein). This activity was capable of phosphorylating phytochrome itself as well as several exogenous proteins. Because most kinases are capable of autophosphorylation, and phytochrome was shown by affinity labeling to have an ATP-binding site that was more exposed in the presence of polycations, it was proposed that the kinase activity may be intrinsic to the photoreceptor. The site of phytochrome phosphorylation catalyzed by this activity in vitro was mapped to a serine residue in the blocked N-terminal tryptic dodecapeptide. Since this peptide contains Ser-7, it is tempting to speculate that this activity is identical to that responsible for phytochrome phosphorylation in vivo. If true, the low kinase activity observed in vitro, typically resulting in less than 5% molar incorporation of phosphate into phytochrome, may be due in part to the substantial levels of phosphorylation already existing in purified phytochrome. Since the intial studies described above, additional groups have also reported the detection of a phytochrome-associated serine kinase activity (1Biermann B.J Pao L.I Feldman L.J Plant Physiol. 1994; 105: 243-251PubMed Google Scholar, 4Grimm R Gast D Rudiger W Planta. 1989; 178: 199-206Crossref PubMed Scopus (45) Google Scholar, 9Kim I.-S Bai U Song P.-S Photochem. Photobiol. 1989; 49: 319-323Crossref PubMed Scopus (30) Google Scholar). In two of these cases, however, it was concluded that the activity was due to a low-abundance, tightly bound contaminant rather than to phytochrome itself (4Grimm R Gast D Rudiger W Planta. 1989; 178: 199-206Crossref PubMed Scopus (45) Google Scholar, 9Kim I.-S Bai U Song P.-S Photochem. Photobiol. 1989; 49: 319-323Crossref PubMed Scopus (30) Google Scholar). This conclusion was based in part on the observation that phytochrome lacking kinase activity could be obtained from phytochrome with activity, and on sequence considerations (see below). While valid, these arguments are not conclusive. Furthermore, new techniques have the potential to render the contaminant kinase hypothesis less probable. The ability to attach chromophore to apoprotein in vitro, along with the advent of recombinant phytochrome expression systems, allows one to produce native, photochemically active holoprotein in nonplant systems. If recombinant phytochrome purified from a yeast expression system were found to exhibit kinase activity similar to that exhibited by plant-purified phytochrome, then we believe that the hypothesis that phytochrome is a kinase would be more likely than the alternative hypothesis of a contaminating activity highly conserved between plants and yeast. Studies to clarify these hypotheses are ongoing. A major argument against phytochrome being a kinase has been that it lacks most of the consensus sequences that define the catalytic domains of eukaryotic protein serine/threonine/tyrosine kinases. While leaving no doubt that phytochromes are not members of this superfamily, the ability to catalyze the phosphorylation of serine residues using ATP as a phosphate donor is not exclusive to this group of proteins. Indeed, there are now many examples of proteins that catalyze the ATP-dependent phosphorylation of hydroxy amino acids, but that show no sequence similarity to the eukaryotic protein serine/threonine/tyrosine kinase superfamily (13Min K-T Hilditch C.M Diederich B Errington J Yudkin M.D Cell. 1993; 74: 735-742Abstract Full Text PDF PubMed Scopus (203) Google Scholar, 5Harris R.A Popov K.M Zhao Y Kedishvili N.T Shimomura Y Crabb D.W Advan. Enzyme Regul. 1995; 35: 147-162Crossref PubMed Scopus (66) Google Scholar; and15Ryazanov A.G Ward M.D Mendola C.E Pavur K.S Dorovkov M.V Wiedmann M Erdjument-Bromage H Tempst P Parmer T.G Prostko C.R Germino F.J Hait W.N Proc. Natl. Acad. Sci. USA. 1997; 94: 4884-4889Crossref PubMed Scopus (156) Google Scholarreferences therein). Of particular interest in this regard are the mitochondrial protein kinase family (5Harris R.A Popov K.M Zhao Y Kedishvili N.T Shimomura Y Crabb D.W Advan. Enzyme Regul. 1995; 35: 147-162Crossref PubMed Scopus (66) Google Scholar) and the Bacillus subtilis anti-sigma factor SpoIIAB (13Min K-T Hilditch C.M Diederich B Errington J Yudkin M.D Cell. 1993; 74: 735-742Abstract Full Text PDF PubMed Scopus (203) Google Scholar). Although these particular protein kinases phosphorylate their substrates on serine residues, like phytochromes they exhibit sequence similarity to the catalytic regions of two-component histidine kinases. So, what are histidine kinases and what do they have to do with phytochromes? Signal transduction in bacteria is predominantly accomplished through histidine-aspartate phosphorelay reactions in so-called two-component systems (reviewed by18Wurgler-Murphy S.M Saito H Trends Biochem.Sci. 1997; 22: 172-176Google Scholar). The prototypical two-component system is comprised of a sensor containing an N-terminal signal input domain and a C-terminal transmitter histidine kinase domain, and a response regulator consisting of an N-terminal receiver domain and a C-terminal output domain. Environmental stimuli activate the transmitter kinase resulting in its autophosphorylation on a conserved histidine residue. The phosphate moiety is then transferred to an aspartate residue in the receiver domain of the response regulator thereby altering the activity of its output domain (e.g., transcription activation). This is a simplified description of a prototypical two-component system but there are many variations on this theme. For example, the domains described above are modular and can exist in different configurations. Furthermore, several multistep systems involving sequential histidine-aspartate phosphotransfer reactions have been described in the last several years. Finally, it is now well established that two-component systems exist in eukaryotes including yeast, Dictyostelium, Neurospora, and Arabidopsis. The first connection of phytochrome to a cyanobacterial histidine kinase came with the report that a Fremyella diplosiphon sensor kinase involved in chromatic adaptation had an N-terminal domain that showed limited but significant sequence similarity to the N-terminal chromophore-bearing domain of higher plant phytochromes (8Kehoe D.M Grossman A.R Science. 1996; 273: 1409-1412Crossref PubMed Scopus (303) Google Scholar). While intriguing, the limited nature of the similarity indicates that this protein may not be a true phytochrome. Subsequently, however, complete sequencing of the Synechocystis sp. PCC6803 genome revealed an ORF, locus slr0473 (now called cph1), encoding a protein whose N terminus shows 30%–35% amino acid identity to the chromophore-bearing domain of higher plant phytochromes, and whose C terminus contains the consensus sequences that define histidine kinases (see19Yeh K.-C Wu S.-H Murphy J.T Lagarias J.C Science. 1997; 277: 1505-1508Crossref PubMed Scopus (429) Google Scholar). Furthermore, the genes for cognate pairs of bacterial sensor kinases and response regulators frequently are found in the same operon, and immediately downstream of cph1 is a second ORF (now called rcp1) encoding a protein with strong sequence similarity to the CheY family of response regulators. In an elegant series of experiments, Cph1 was shown to be a true phytochrome by its ability to catalyze the attachment of its own linear tetrapyrrole chromophore and to display red/far-red photoreversible absorbance properties (Figure 1D) (6Hughes J Lamparter T Mittmann F Hartmann E Gartner W Wilde A Borner T Nature. 1997; 386: 663Crossref PubMed Scopus (285) Google Scholar, 19Yeh K.-C Wu S.-H Murphy J.T Lagarias J.C Science. 1997; 277: 1505-1508Crossref PubMed Scopus (429) Google Scholar). Cph1 was further shown to be a histidine kinase by its ability to autophosphorylate on the appropriate conserved histidine residue (19Yeh K.-C Wu S.-H Murphy J.T Lagarias J.C Science. 1997; 277: 1505-1508Crossref PubMed Scopus (429) Google Scholar). As predicted, phosphotransfer from Cph1 to the conserved aspartate residue of Rcp1 was observed, proving that these two proteins do in fact form a cognate pair of a functional two-component system (19Yeh K.-C Wu S.-H Murphy J.T Lagarias J.C Science. 1997; 277: 1505-1508Crossref PubMed Scopus (429) Google Scholar). Most satisfyingly, Cph1 was shown to be a light-regulated enzyme since only the Pr form exhibited substantial kinase activity (19Yeh K.-C Wu S.-H Murphy J.T Lagarias J.C Science. 1997; 277: 1505-1508Crossref PubMed Scopus (429) Google Scholar). Additionally, only phosphorylated Pr, and not phosphorylated Pfr (formed by photoconversion of phosphorylated Pr since Pfr lacks kinase activity), was able to transfer phosphate to Rcp1 (19Yeh K.-C Wu S.-H Murphy J.T Lagarias J.C Science. 1997; 277: 1505-1508Crossref PubMed Scopus (429) Google Scholar). Thus, Cph1 and Rcp1 clearly represent a light-regulated two-component system though their effector function is presently unknown. In addition to being exciting from a phytochrome perspective, this particular two-component system has the potential to further our understanding of histidine kinase regulation in general. Because most histidine kinases are integral membrane proteins whose ligands are unknown, the molecular mechanisms of kinase activation have been difficult to investigate and hence are not well understood. Since Cph1 is a soluble protein with known “ligands” (red and far-red light) and two readily identifiable states (Pr and Pfr), it is an attractive model system for studying sensor kinase activation. That's all fine and good for cyanobacterial phytochrome, but the real excitement has to do with potential functional homology of plant phytochromes. Specifically, do plant phytochromes possess histidine kinase activity and what then do we make of phytochrome-associated serine kinase activity? As discussed by 14Quail P.H BioEssay. 1997; 19: 571-579Crossref PubMed Scopus (70) Google Scholar, while there is presently no experimental evidence supporting the hypothesis that higher plant phytochromes are histidine kinases, this question must still be considered unresolved. In any case, whether or not they have histidine kinase activity, it is clear from sequence analysis and the discovery of Cph1 that plant phytochromes are evolutionarily related to histidine kinases. Given this, and assuming that phytochrome-associated serine kinase activity is intrinsic to the photoreceptor, there are two general models that seem most probable in accounting for serine kinase activity in a histidine kinase–related protein. The first model postulates that plant phytochromes, and perhaps other proteins like SpoIIAB and the mitochondrial protein kinases, represent nonorthodox members of the sensor kinase family that have altered their substrate specificity from histidines to serines (Figure 2A). In this case, the phosphorylated serines would be expected to be stable regulatory modifications rather than intermediates in phosphotransfer reactions due to the comparatively low energy of a phosphoester bond. As discussed previously, phytochrome autophosphorylation would presumably be a down-regulation mechanism; however, this would not preclude the existence of other substrates for phytochrome kinase activity. In this model then, phytochromes would be functionally similar to the eukaryotic protein serine/threonine/tyrosine kinase family even though they are evolutionarily related to histidine kinases. The alternative model is that plant phytochromes are true histidine kinases that autophosphorylate on a nonconsensus histidine, but the phosphate moiety is then transferred to the N-terminal serine residue (Figure 2B). As previously noted, (14Quail P.H BioEssay. 1997; 19: 571-579Crossref PubMed Scopus (70) Google Scholar) there are examples of sensor kinases like CheA that autophosphorylate on nonconsensus histidines. Furthermore, there is precedence for the proposed histidine to serine phosphotransfer reaction. CheY mutants lacking the phosphate-accepting aspartate are phosphorylated by CheA on a hydroxy amino acid presumably due to proximity and the high phosphotransfer potential of a phosphohistidine (2Bourret R.B Hess F Simon M.I Proc. Natl. Acad. Sci. USA. 1990; 87: 41-45Crossref PubMed Scopus (197) Google Scholar). One would predict that such a reaction could be favored if an internal hydroxy amino acid were properly spaced and configured relative to a phosphorylated histidine residue. In this regard, we note that the serine phosphorylated in plant phytochromes resides in a 60–100 amino acid N-terminal extension not found in Cph1 (Figure 1D), suggesting that this domain has a function unique to eukaryotic phytochromes. One could envision this model working either with or without a conventional response regulator. In either case, we once again postulate that phytochrome autophosphorylation would serve as a down-regulation mechanism. If response regulators were not involved in signaling, this reaction might simply be an internal feedback mechanism that may or may not have higher levels of regulation beyond phytochrome photoconversion. If response regulators were involved in phytochrome signaling, the N-terminal serine might be acting as an alternate substrate when the response regulator was not available (e.g., due to its subsequent signaling function) or when it was already phosphorylated (Figure 2B). Presently, the proposed histidine-to-serine phosphotransfer reaction in this model can be tested by systematically mutating all phytochrome histidine residues and assaying for an effect on N-terminal serine phosphorylation. Alternatively, the postulated histidine autophosphorylation may be more readily detected in phytochromes where the known serine phosphorylation site was mutated. In summary, we believe that the recurring connections between phytochromes and protein kinases cannot be coincidental and must be indicative of functional importance. Although it is clear that plant phytochromes arose from an ancestral prokaryotic histidine kinase, the fact that they seem to exhibit serine kinase activity indicates that they are atypical members of this superfamily. Therefore, the extent of functional homology between plant phytochromes and histidine kinases is still in question. Resolution of these ambiguities will mark an important step in our understanding of phytochrome function. This obviously is a required step if phytochrome signal transduction involves phosphorylation of subsequent pathway components.