Essay 17.2

Diversity of Phytochrome Chromophores

Ruairidh J. H. Sawers, Department of Plant Biology, University of Geneva, Geneva, Switzerland

September, 2006

Introduction

The first phytochrome photoreceptors to be characterized were isolated from flowering plants. However, recently, phytochrome and phytochrome-like proteins have been discovered in photosynthetic and nonphotosynthetic bacteria, in fungi, and in slime molds. Such a distribution challenges ideas concerning the evolution of the phytochrome receptors and the nature of their function. Are they all photoreceptors? What was their ancestral function? It has been suggested that the phytochrome family is derived from proteins that evolved as bilin (linear tetrapyrrole) sensors. Of the proteins currently known, all bind a bilin chromophore and it is these light-absorbing prosthetic groups that confer spectral activity. The nature of the chromophore differs among the phytochrome-like protein family, resulting in subtle differences in spectral properties. However, a shared evolutionary history means that the proteins retain the potential to form adducts with certain nonnative chromophores, if they are provided. This observation opens up the possibility for ′spectral tuning′ of crop light-responsive behaviors by chromophore substitution.

Plant Phytochromes Influence Important Crop Traits

Figure 1 shows a mutant corn plant (elm1) that accumulates reduced amounts of the phytochrome chromophore, phytochromobilin (PFB). Although corn contains six different phytochrome photoreceptors, all six require the same PFB chromophore molecule, and without it, the plant becomes blind to red and far-red light. One agronomic benefit of PFB deficiency is an acceleration of flowering time under normally nonpermissive day lengths (Figure 1). However, accelerated flowering is coupled with increased elongation and a resulting tendency to lodge (fall over). The elm1 mutant demonstrates the importance of the phytochromes in the regulation of mature plant development and the potential to alter development by chromophore manipulation. However, the elm1 mutant also shows that dramatic alterations in a pathway (in this case, inactivation) may result in a combination of both beneficial and deleterious effects. Chromophore substitution, instead of chromophore removal, may offer one route to the more subtle manipulation of light-signaling pathways beneficial to agronomic performance.

Figure 1 Mature plant phenotypes of wild type (left) and elm1 mutant (right) plants were grown at a summer field site in New Haven, Connecticut. elm1 mutants are prone to lodging (falling over) but flower earlier than wild-type siblings. (Click image to enlarge.)

Plant and Bacterial Phytochromes Bind Their Chromophores Covalently

How do phytochrome-like proteins bind their chromophores, and how conserved are mechanisms of attachment across the protein family? Phytochrome proteins possess an intrinsic lyase activity and will spontaneously assemble with certain bilins. A covalent thioether linkage is formed between the bilin and a cysteine residue in the phytochrome apoprotein (Figure 2).

Figure 2 Covalent attachment of PFB chromophore to the GAF domain of phytochrome. The reaction proceeds via the formation of a thioether linkage between an ethylidene group on the A-ring of the chromophore and a conserved cysteine residue in the phytochrome apoprotein. As a result of this reaction, a double bond is lost (ringed in red), resulting in a subtle change in spectral properties. (Click image to enlarge.)

Figure 3 shows the domain structure and position of chromophore attachment in higher plant and bacterial phytochrome-like proteins. In plant phytochrome (Phy), chromophore binding occurs via linkage to a cysteine in the GAF domain. Two broad classes of phytochromes can be recognized in bacteria. The first class of proteins (Cph), typical of the cyanobacteria, contains the conserved cysteine residue present in Phy proteins and binds their chromophore in a similar way. A second class of bacterial phytochromes (Bph) does not contain the conserved GAF domain cysteine, and instead, binds chromophore via a cysteine residue in a region preceding the PAS domain. However, although lacking the conserved cysteine residue, the GAF domain of Bph proteins is still present and the topography of the bound bilin may be very similar to that of Phy and Cph. The difference in chromophore binding between Bph and Phy appears to be the direction from which the binding cysteine approaches the A-ring of the bilin. This suggests that the evolutionary development of a Bph-type binding to a Phy type may only have required the introduction of the binding cysteine into an existing GAF domain and that very little further structural rearrangement was required.

Figure 3 Attachment of bilin chromophores to phytochrome and phytochrome-like proteins. Higher plant phytochrome (Phy) binds its chromophore by a covalent linkage formed with a conserved cysteine residue present in the GAF domain. Similar attachment chemistry is used by certain cyanobacterial phytochromes (Cph). A second class of bacterial proteins (Bph) does not contain a similar GAF domain cysteine residue but binds its chromophore via an N-terminal cysteine. Upon folding, the position of the bound chromophore in Bph is believed to be similar to that in Phy/Cph. (Phytochrome and phytochrome-like domain structure after Montgomery and Lagarias 2002.) (Click image to enlarge.)

A Number of Linear Tetrapyrroles Function as Phytochrome Chromophores

Figure 4 shows three naturally occurring bilin chromophores used by phytochrome and phytochrome-related proteins. Biliverdin IX (BV) is both a biosynthetic precursor to PFB and phycocyanobilin (PCB) and a chromophore in its own right, bound by Bph proteins. PFB is the chromophore used by the higher-plant Phy proteins, and PCB by the cyanobacterial Cph proteins. BV, PFB, and PCB are derived from linearization of a heme tetrapyrrole ring, familiar as the oxygenating pigment in mammalian blood. The elm1 mutant described above is deficient in one of the steps required for this conversion of heme to PFB. Although structurally very similar, BV, PFB, and PCB differ in the number of double bonds they contain. These differences lead to differences in spectral properties. Bilins absorb light as a result of a dissociated electron system. More reduced forms, containing fewer double bonds, have a shortened electron system and preferentially absorb light at shorter wavelengths (closer to the blue end of the spectrum). The presumed evolutionary switch from BV to the more reduced PFB and PCB in photosynthetic organisms has been interpreted as an adaptation to allow a greater sensitivity to the most photosynthetically active light.

Given what is known concerning chromophore diversity and the nature of attachment, to what extent are chromophores interchangeable? The attachment of PFB and PCB chromophores by Phy and Cph proceeds via a common mechanism of thioether formation with an ethylidene group present at the C3¹ carbon of the bilin A-ring (Figure 4). As might be expected, these two chromophores appear to be interchangeable and higher-plant Phy will assemble with PCB to generate a functional photoreceptor. The resulting Phy-PCB proteins possess blue-shifted spectral properties. Nonnative chromophores that share this common A-ring structure will also assemble with Phy and Cph proteins. One example is the phycoerythrobilin (PEB), a pigment linked to light harvesting in algae. In this instance, the resulting holoproteins are not spectrally active (PEB lacks the C15 double bond that is the site of light-dependent isomerization in PFB, PCB, and BV) but demonstrate intense orange fluorescence. In contrast to PFB and PCB, BV does not contain the A-ring ethylidene group required for formation of a thioether linkage with the GAF domain cysteine of Phy and Cph (BV has a vinyl group at this position), and therefore, does not covalently bind to Phy or Cph. Structural work has revealed that BV binding to Bph proceeds via linkage at the C3² carbon and that this reaction may be favored by the geometry of the attachment site. Such analyses have begun to reveal both the potential and limitations of chromophore replacement in the phytochrome-like proteins. However, they also suggest that engineering of the apoprotein itself may allow for alterations in the binding capacity and subsequent changes to the spectral activity of phytochrome-like proteins.

Figure 4 Diversity of chromophore groups bound by phytochrome and phytochrome-related proteins. Structural differences and oxidation relationships are highlighted. Note that this figure does not reflect pathways of biosynthesis. (Click image to enlarge.)

Interactions between Chromophores and Proteins Influence Spectral Properties

It was observed early on that the spectral properties of phytochrome holoenzymes did not exactly match those of the free chromophores in solution. Although this may be partially explained by binding chemistry, there is evidence that the bound chromphore is maintained in a linear form, and possibly somewhat distorted by interactions with the phytochrome protein. Determination of the structure of a Bph chromophore-binding domain has shown that the bilin lies in a cleft, buried within the protein. In addition to the covalent linkage, there are many other potential interactions between protein and chromophore. The nature of these interactions, and perhaps their alteration following chromophore isomerization, are believed to play a key role in overall phytochrome function and in specific properties. For example, a naturally occurring allele of Arabidopsis phyA has been isolated containing an amino-acid substitution that also results in changes to the absorption maximum. A second study identified a number of amino-acid changes near the site of chromophore-binding in Cph that result in changes to spectral properties. The most dramatic result in phytochromes is that they become more stable, or even locked, in one or other light-absorbing forms. Amino-acid mutations that sterically prevent chromophore isomerization can also create fluorescent holoproteins, similar to fluorescence induced by PEB adducts. It is now possible to express, and assemble, active phytochrome-like proteins in Escherichia coli. Such systems offer great potential for the identification and analysis of new protein variants and novel combinations of protein and chromophore.

Prospects for Phytochrome Engineering

A greater understanding and appreciation of the broader diversity and evolutionary history of the phytochrome-like protein family has opened up the possibility to rationally engineer the spectral properties of these light-sensing switches. The role of light in plant development has long been recognized and it is now clear that the phytochrome system plays a role in influencing many agronomically important traits. As might be expected, attempts to manipulate certain traits, but not others, can be problematic. Plant responses to light represent the subtle balance between a number of (in some cases, antagonistic) phytochrome-mediated signals in response to changes in the spectral quality of the light environment. Spectral tuning of these photoreceptors offers one way to gently shift this balance without leading to widespread developmental disruption. In addition to agronomic considerations, phytochrome engineering is beginning to provide a novel collection of fluorescent and photoactive molecules for both analytical and therapeutic applications.

Summary and Conclusions

Members of the phytochrome family bind a number of different bilin chromophores and exhibit differences in spectral properties. A shared evolutionary history and mechanism mean that there is the potential to create phytochromes possessing novel properties by interchange of chromophores. Such novel proteins may have agronomic, analytical, and therapeutic applications.

References

Fischer, A. J., and Lagarias, J. C. (2004) Harnessing phytochrome′s glowing potential. Proc. Natl. Acad. Sci. USA 101: 17334–17339.

Kami, C., Mukougawa, K., Muramoto, T., Yokota, A., Shinomura, T., Lagarias, J. C., and Kohchi, T. (2004) Complementation of phytochrome chromophore-deficient Arabidopsis by expression of phycocyanobilin:ferredoxin oxidoreductase. Proc. Natl. Acad. Sci. USA 101: 1099–1104.

Montgomery, B. L., and Lagarias, J. C. (2002) Phytochrome ancestry: Sensors of bilins and light. Trends Plant Sci. 7: 357–366.

Sawers, R. J., Sheehan, M. J., and Brutnell, T. P. (2005) Cereal phytochromes: Targets of selection, targets for manipulation? Trends Plant Sci. 10: 138–143.

Wagner, J. R., Brunzelle, J. S., Forest, K. T., and Vierstra, R. D. (2005) A light-sensing knot revealed by the structure of the chromophore-binding domain of phytochrome. Nature 438: 325–331.