A new type of photoreceptor in algae

Light serves two major purposes in biological systems: the transduction of light to chemical free energy, photosynthesis, and the transduction of light to initiate a signaling pathway, to sense the environment. In animals, retinal pigments underlie signaling by light: vision and photoperiodism. Indeed, the use of retinal, the aldehyde of vitamin A, by visual pigments for their chromophore was the very first example in which the reason that humans require a vitamin was elucidated (1). Plants do more with light than animals and thus have more pigments to transduce it. Besides chlorophyll, used in photosynthesis, at least some plants and algae use other pigments for the processes that underlie photomorphogenesis, photoperiodism, and photomovement. The last category involves both phototropism in sessile organisms and phototaxis in motile organisms, mostly some types of algae. Recently a combination of having the complete genome of a plant, Arabidopsis, and substantial parts for an algae, Chlamydomonas, and being able to manipulate the genes expressed in plants and algae is starting to clarify the role of light in signaling. Several types of photoreceptor molecules that are associated with some well known light responses of plants have been demonstrated. In the model plant Arabidopsis, it seems that there are at least three types of photoreceptor pigment families, each with multiple members: two cryptochromes (each with two chromophores), five phytochromes, and two phototropins (2), none of which use retinal as their chromophore.

However, there are very old reports that some kinds of plants and algae have a class of photoreceptors based on retinal. Wald (3) speculated on such a retinal-based photoreceptor in his 1945 Harvey Lectures, but the first unequivocal evidence came from experiments with cultures of Chlamydomonas (Fig. 1) in which carotenoid (and retinal) synthesis was suppressed, and concomitantly phototaxis disappeared. Phototaxis could be restored with the addition of retinal (4), but progress beyond this point has been very difficult with few candidate genes or pigments actually isolated and none proven to be the holy grail. The paper in this issue of PNAS by Sineshchekov, Jung, and Spudich (5) has changed this.

Figure 1.

Schematic representation of the unicellular green algae Chlamydomonas and its eyespot. The cell is ≈10 μm in diameter, the eyespot is 1 μm in diameter, and the flagella are ≈10 μm long. The blow up to the right is a cross section of the eyespot through the various algal membranes. The photoreceptor pigment most likely is localized in the cell's outer, plasma membrane. As the cell swims with its two flagella, it also rotates around the axis along which it is progressing. The dense pigment indicated as a set of colored circles in the eyespot can shield the photoreceptor pigment from light coming from one direction as compared to the other such that the swimming, rotating Chlamydomonas can continuously sample the spatial distribution of light. The figure is courtesy of Ella Imasheva (University of Washington, Seattle).

First a little background. Until 30 years ago, photosynthesis was thought to always require chlorophyll, but the discovery of bacteriorhodopsin introduced a second system, one based on retinal, which also could convert light to chemical energy (6–8). Bacteriorhodopsin is found in several halophilic Archaea. In the past year or so, a close relative of bacteriorhodopsin was found in eubacteria (proteobacteria) in marine phytoplankton, where it is thought also to function as a light-driven proton pump. It has been estimated that retinal-based photosynthesis might account for as much as 10% of the total photosynthetic energy conversion on the earth (9, 10).

Many of the same halobacteria that use bacteriorhodopsin as a light energy to chemical free energy transducer also have two other retinal proteins that are involved in photosignaling: sensory rhodopsins I and II. The amino acid sequences of these three retinal proteins, as well as that of a fourth pigment involved in light-driven chloride transport, halorhodopsin, are quite similar. They also have many properties in common. For example, transduction is initiated by photoisomerization, leading to a reversible sequence of changes in the pigment, the “photocycle.” All four pigments use the all-trans isomer of retinal as the chromophore configuration for their initial active state. It seems likely that the retinal proteins from the proteobacteria also utilizes the all-trans isomer of retinal. Finally, another cousin retinal protein of the archael rhodopsins has been found recently in Neurospora (11, 12), where it is thought to be involved in blue-light photosensing, although this is not yet established. This pigment can be expressed, and it was shown that it also used all-trans retinal for its chromophore.

The retinal proteins used by vertebrate and invertebrate animals to detect light (13, 14) have little sequence similarity to the archaeal rhodopsins and their fungal and eubacterial relatives. These retinal pigments, usually called visual pigments, have many distinctive properties; for example they uniformly use the 11-cis isomer of retinal for their chromophores. Light causes an 11-cis to all-trans isomerization, leading to a sequence of changes in the pigment requiring enzymatic processes to restore the pigment to its initial state. The visual pigments are the best characterized members of a eukaryotic cell signaling superfamily, the G protein-coupled receptors (GPCRs). Upon activation, the receptors in this superfamily all act by catalyzing the exchange of GDP with GTP in a GTP-binding protein (G protein); the active form of the G protein then acts on some effector to modulate a physiological response (13, 14).

Sineshchekov, Jung, and Spudich (5) found a fragment of a candidate gene in the expressed sequence-tag data bank of the green algae Chlamydomonas by looking for similarity to archaeal rhodopsin-type pigments. Eventually two quite similar genes were identified that had almost all the residues in the retinal binding site of bacteriorhodopsin, and it was suspected that these genes were related to the putative retinal-based pigments that had been suggested for Chlamydomonas (15). There are two especially difficult problems in trying to find the retinal-based signal transduction apparatus in algae. One is the problem of easily measuring the early transduction processes, those most closely related to the photoreceptor. The second problem is showing that a candidate gene product is actually the receptor in the signal transduction system. In the paper by Sineshchekov, Jong, and Spudich, the first problem was overcome by measuring the light-induced electrical currents from the algae, which are part of the signal transduction pathway (5). To show that the pigments are a part of the genuine signaling system, ideally one would like to delete each gene by using homologous recombination, but it is not easy to do such gene knockouts in this (or any) algal species. Instead, this problem was overcome by using RNA interference (RNAi) technology to preferentially suppress the synthesis of one or the other of the two pigments to convincingly show that each pigment is a genuine photoreceptor and exerts control of one segment of the algal phototactic response.

When the first of the two putative pigments, CSRA (Chlamydomonas sensory rhodopsin A) had its gene expression reduced by its RNA interference, the faster photocurrent that saturates at higher light intensities was reduced. When the other pigment, CSRB, had its expression reduced similarly, the slower, more easily saturated photocurrent was reduced. These experiments convincingly show that light absorbed by the CSRA and CSRB pigments do initiate the photocurrents from the algae and thus are the photoreceptors for phototaxis.

Many intriguing questions, of course, are just starting to be pursued. The action spectra that Sineshchekov, Jung, and Spudich measured are a bit unusual; the peaks attributed to the individual pigments are quite narrow, and it may be that rather than each action spectrum containing contributions from two pigments, with one dominating in each case, each action spectrum represents a single pigment. Each of the two pigments would have the same two vibrational peaks but different amplitudes. This and many other questions involving the spectroscopy and photochemistry of the pigments could be answered if one had an expression system producing sufficient amounts of the pigment.

Another question involves how the signaling pathway works. Phototaxis in halobacteria, which uses retinal pigments very similar to the transmembrane portion of the algal pigments, has many components in its signaling pathway that are similar to those of eubacterial chemotaxis (16). But larger cells, such as some algae, seem to rely on an electrical step, a receptor potential, to help link the transduction site to the effector site, where the flagellar motion is controlled (15). So how does the pigment initiate the receptor potential? In vision this is done via the G proteins, but the evidence for G protein-coupled receptor signaling in plants is small (17). Thus the retinal-based photoreception in algae may be based on an entirely new type of signaling pathway.

Finally, there is the question of how widespread the new mechanism of retinal protein signaling that has been revealed in Chlamydomonas is. Is it just in certain kinds of green algae, all green algae, all algae, or all plants and algae?