Cyanobacteria are a splendid group of organisms, and much in demand. Formerly "blue green algae", they were appropriated by microbiologists in the 1970s, by which time it was abundantly clear that, as prokaryotes, they better fit the blueprint of a bacterium than that of a plant or alga [1]. Die-hard phycologists argue that a prokaryotic alga is not a contradiction in terms, and point to the important trick, shared with real plants, of using water as the source of hydrogen atoms for photosynthetic electron transport, with molecular oxygen as a benchmark in toxic waste. The consensus is that cell structure and genetics are a better base for taxonomy than photosynthesis, and, in any case, certain species of cyanobacteria are now known to be able to throw off the aerobic, oxygenic mask by abstracting hydrogen from hydrogen sulphide instead of water, carrying out strictly anaerobic photosynthesis of a decidedly bacterial kind.
Having said this, plant scientists cannot safely delegate the cyanobacteria. If in doubt, look at the wealth of information in the synopsis edited by Bryant [2]. First, oxygen-evolving photosynthesis is totally intact - the electrons pass through exactly the same chain as in plant chloroplasts (though cyanobacteria keep open a few special options), while photosynthesis is inherently more amenable to molecular genetic analysis in cyanobacteria than in plants. Second, the chloroplast, the sine qua non of the plant cell, is clearly the descendant of a free-living prokaryote that was as close to a modern cyanobacterium as makes little difference. Why the chloroplast should have retained a core genetic system to synthesise a handful of photosynthetic proteins, likewise the mitochondrion for respiratory ones, is a major problem in cell evolution - one might have thought it tidier, and safer, to hand everything over to the nucleus [3]. Third, cyanobacteria adapt, and it is now becoming clear that the mechanisms by which they detect and respond to environmental change are typically prokaryotic, while, at the same time, a foundation from which eukaryotic signal transduction pathways must have evolved.
Take complementary chromatic adaptation. Described almost a century ago, this phenomenon is now yielding to molecular genetics, and in doing so it reveals some intriguing resemblance to the light responses mediated by phytochrome, the home ground of plant photomorphogenesis. Grow a culture of a Calothrix or Fremyella species in red light, and it appears green. Transfer it to green light, and it turns red. Teleology, endemic to biology, is now a calculated risk: the cells change colour because of changes in the relative proportions of the pigments phycoerythrin (which absorbs blue and green light, and is hence red) and phycocyanin (which absorbs red light, and is hence blue-green). Add the knowledge that these pigments function as components of the light-harvesting apparatus of cyanobacterial photosynthesis, and the functional explanation is obvious: the cells adapt in order to be able to absorb more light for photosynthesis. Thus and their colour (transmitted and reflected light) becomes complementary to that of the light in which they find themselves.
A recent breakthrough by Grossman and co-workers provides the outline of a molecular description of what is going on in complementary chromatic adaptation [4,5]. It was already know that the phycoerythrobilin in and phycocyanobilin chromophores are attached to proteins to form the pigment-proteins phycoerythrin and phycocyanin. Transcription of the genes encoding the apoproteins is induced by light with a spectral composition matching that if the absorption spectrum of the holoproteins, the light-harvesting biliprotein gene products themselves. How can spectral composition be a signal for transcription? The breakthrough follows a familiar path. Mutants of Fremyella diplosiphon are obtained that are unable to adapt. Among these are mutants that stay red. Some of these mutations are complemented - chromatic adaptation is restored - upon transformation of the red-only cells with a gene rcaC (regulator of chromatic adaptation C) [4]. The predicted 73 kDa RcaC polypeptide is a typical bacterial response regulator, that is, it has a conserved pattern of motifs seen elsewhere in proteins involved in a wide variety of bacterial responses to specific environmental change [6]. These motifs include a predicted a-b structure with an aspartate on a surface-exposed loop between the first a-helix and the first b-strand. The paradigm for such response regulators is CheY, which dictates clockwise or anticlockwise rotation of the E. coli flagellum, and hence switches on tumbling, the behavioural response at the centre of bacterial chemotaxis. A high-resolution structure for unmodified CheY has been obtained by X-ray crystallography [7]. CheY is modified by phosphorylation - a response regulators are phosphorylated on the aspartate on the surface-exposed loop. The phosphate group is accepted from a phosphohistidine side chain of the response regulator's partner in crime, the sensor [6]. The sensor becomes phosphorylated (on histidine) in response to the relevant environmental signal. The sensor is thus a response regulator kinase, mediating transfer of phosphate from ATP to the response regulator if, and only if, the environmental conditions are right. Not teleology, you see, but chemistry plus natural selection. Without two-component (sensor and response regulator) signal transduction, bacteria are sans eyes, sans ears, sans teeth, sans everything.
So what is the photosensor for complementary chromatic adaptation? In general terms, this is a foregone conclusion - a histidine sensor kinase. But the latest part of the breakthrough, the discovery of complementation of other mutations with rcaE5, brings us back to plants, and introduces something few would have expected. The predicted chromatic adaptation sensor RcaE indeed contains the four C-terminal motifs expected for a histidine sensor kinase, including the histidine itself. The N-terminal domain contains a region with similarity to a known chromophore attachment domain - that of phytochrome. Homology between phytochrome and bacterial sensor kinases was previously predicted, from imaginative sequence-gazing, by Schneider-Poetsch [8], but a weak link in the argument was replacement of the histidine autophosphorylation site by phytochrome tyrosine. With Kehoe and Grossman's RcaE5, the homology with phytochrome is now undeniable, and the histidine is conserved. Furthermore, two-component regulation came to eukaryotes in 1994 [9], with the sequences of an ethylene response element from Arabidopsis and of another, unspecified receptor, from Saccharomyces. Intriguingly, RcaE also shows similarities to a plant ethylene response element [9].
It is early to expect biochemistry too, but it must come. There are certainly cyanobacterial phosphoproteins [10,11] and two-component systems [12-14]. What is the RcaE chromophore? RcaE lacks the cysteines that ligate the linear tetrapyrrole chromophore of phytochrome. Perhaps RcaE itself has no chromophore, and is a photosensor controlled by a photoreceptor, analogous to the chemoreceptor and chemotaxis sensor and response regulators of E. coli? The phycocyanin and phycoerythrin chromophores themselves are also linear tetrapyrroles. It is tempting to think that the evolutionary prototype of the photoreceptor phytochrome was itself a photosynthetic light-harvesting protein, capable of regulating its own synthesis. While red light repression of phycoerythrin synthesis may well be explained by RcaE's phytochrome-like properties, the green light activation of phycoerythrin synthesis concerns a spectral area outside of the response of phytochrome. It remains to be seen whether the biliprotein spectral window is modified in the photoreceptor, or whether another photosensory system, for which the rhodopsins are especially suited, is responsible for effects of green light.
Complementary chromatic adaptation is not the only means by which cyanobacteria seem to know what their light environment is good for. Indeed, many species do not exhibit the phenomenon at all. A more widespread, long-term response to altered spectral composition is adjustment of the stoichiometry of photosystem I and photosystem II of the photosynthetic electron transport chain [15]. Again, the response is complementary to the change in the light regime: photosystem I absorbs more at higher wavelengths than photosystem II, and becomes relatively more abundant under blue-green illumination, as if to make good by sheer quantity its limited light-harvesting capacity. As a function of light intensity, too, the ratio of photosystem I to photosystem II changes - from near to unity in saturating light to a six-fold increase in favour of photosystem I in low light.
Then there are short-term adaptations - state transitions. These satisfy the same need to balance photosystem I with photosystem II. State transitions involve post-translational, covalent modification of pre-existing proteins, most clearly in chloroplasts, where the functional equivalent of the cyanobacterial phycobilisome is the major light-harvesting chlorophyll-protein LHC II. LHC II becomes phosphorylated by a protein kinase whose activity is regulated by redox state of electron carriers located between photosystem I and photosystem II [10]. Phospho-LHC II, produced in response to a redox signal to the LHC II kinase, moves from photosystem II to photosystem I, restoring balanced delivery of light energy to the two photosystems [10]. Cyanobacteria and red algae also perform state transitions, with a mobile phycobilisome responding to the redox state of electron carriers located between the two photosystems [10]. However, phycobilisome-based state transitions are unlikely to be driven by the same sort of kinase as that found in chloroplasts [16]. The chloroplast LHC II kinase itself is much sought, but elusive. Recent evidence suggests a location in the core of photosystem II [17]
In photosynthetic organisms, redox control means light control. The underlying mechanism of adjustment of photosystem stoichiometry probably involve redox control, rather than direct photocontrol, of gene expression [10,15,18,19]. The universal eukaryotic location of the genes for the core components of the two photosystems in chloroplasts rather than in the nucleus immediately suggests a reason for the maintenance of the extranuclear genomes of chloroplasts (and, by analogy, mitochondria): electron transport holds a tight rein on expression of genes for its own components [3,20].
For cyanobacteria, actually for biology, an extraordinary development is the completion, in May 1996, of the genome sequence of Synechocystis PCC 6803 at the Kazusa DNA Research Institute in Japan [21]. This brings a new era in cyanobacteriology. Thanks both to the research itself, and to the openness and generosity of its participants, the question "is there a cyanobacterial gene with homology to x?" is now trivial - the answer is known, and one can get it straight away from CyanoBase, at http://www.kazusa.or.jp/cyano/. It seems to have taken twenty-four people only nineteen months to obtain the complete sequence of 3,573,470 base pairs. (This is 261 base pairs per person per day, for the statistically-minded.) The presentation of all this data is a model of clarity and accessibility. Apart from clickable Java maps that take you down from the whole genome (at http://www.kazusa.or.jp/cyano/map/java/cmap_small/cmap_sm.html, see figure 1) to any part of it you choose, CyanoBase can be searched with any keyword or sequence. There is less background information on proteins than in Swissprot, for example, but having one complete cyanobacterial genome on one site in this searchable form is more than just useful.
There is a novel [22] in which opposition to the construction of a computer that will find "the answer to the ultimate question about, the universe, and everything" is mounted by a philosophers' trade union on the grounds that it would put them all out of work. The joke, of course, is protectionism, and a vested interest in hindering a goal one pretends to seek. We suggest that CyanoBase would not put out of business a Trends in Theoretical Cyanobacteriology, but in fact give it a new lease of life. See http://www.kazusa.or.jp/cyano/others.html "Comparative analysis to plastid genomes (construction)" and "Comparative analysis to other bacterial genomes (construction)", for example, and read on.
Synechocystis 6803 is not a complementary chromatic adapter. But enter "histidine sensor kinase" in CyanoBase, and you will find several with homology to phytochrome. Maybe colour of light does other things through two-component signal transduction? Triggering development, for example (this is known in other species14)? Maybe the photosystem stoichiometry and state transition redox sensor(s) (we assume) are there at the origin of photomorphogenesis?
Further, perhaps in an idle moment, look up "phytochrome" (see Figure 1). No problem. The gene slr0473 starts at 2,607,812 and ends at 2,610,058. The predicted gene product has 748 amino acids. 35.5 % of its amino acid sequence is identical to that of the phyC gene product of Arabidopsis thaliana. This is not homology - this is phytochrome. If the role of RcaC and RcaE in chromatic adaptation is a breakthrough, CyanoBase is a revolution. They have no seed dormancy, no photoperiodism, no phototropism, but cyanobacteria contain phytochrome - end of argument. Finding unexpected certainties is unusual in science, and can be as disconcerting as the computer "Deep Thought" coming up with the answer to the ultimate question. And what would Stanier1 have thought of CyanoBase?
Oh, and how many histidine sensor kinases are there in Synechocystis? Thirty-one, actually. No, not forty-two [22].
Acknowledgements. We are grateful to Carol Allen and Jasper van Thor for discussion and suggestions.
2 Bryant, D. A., ed. (1994) The Molecular Biology of Cyanobacteria, Kluywer Academic Publishers
9 Alex, L. A. and Simon, M. I. (1994) Protein histidine kinases and signal transduction in prokaryotes and eukaryotes, Trends Genet. 10, 133-138
10 Allen, J. F. (1992) Protein phosphorylation in regulation of photosynthesis Biochim. Biophys Acta 1098, 275-335
11 Mann, N. H. (1994) Protein phosphorylation in cyanobacteria, Microbiol. 140, 3207-3215
22 Adams, D. (1979) The Hitch Hiker's Guide to the Galaxy, Pan Books.