Light-Responsive Cryptochromes from a Simple Multicellular Animal, the Coral Acropora millepora

See allHide authors and affiliations

Science  19 Oct 2007:
Vol. 318, Issue 5849, pp. 467-470
DOI: 10.1126/science.1145432


Hundreds of species of reef-building corals spawn synchronously over a few nights each year, and moonlight regulates this spawning event. However, the molecular elements underpinning the detection of moonlight remain unknown. Here we report the presence of an ancient family of blue-light–sensing photoreceptors, cryptochromes, in the reef-building coral Acropora millepora. In addition to being cryptochrome genes from one of the earliest-diverging eumetazoan phyla, cry1 and cry2 were expressed preferentially in light. Consistent with potential roles in the synchronization of fundamentally important behaviors such as mass spawning, cry2 expression increased on full moon nights versus new moon nights. Our results demonstrate phylogenetically broad roles of these ancient circadian clock–related molecules in the animal kingdom.

Many organisms possess endogenous clocks that respond to rhythmic changes in light and temperature caused by Earth's rotation (1, 2), allowing them to anticipate daily and annual environmental cycles and to adjust their biochemical, physiological, and behavioral processes accordingly (1). The circadian clock uses cues such as light to entrain endogenous oscillators, which in turn control rhythmic outputs of a wide range of organisms (2). Even simple animals such as medusae (scyphozoan or hydrozoan cnidarians) have specialized light-sensing organs known as ocelli (eyes) or eyespots. These photoreceptors react to changes in light intensity and are responsible for phototaxis and other behavioral responses to light (3). However, anthozoan cnidarians (corals, sea anemones, and sea pens) lack specialized sense organs yet display photosensitive behavior (411). The synchronized mass spawning on the Great Barrier Reef (GBR) in Australia is a spectacular example of the photosensitive responses exhibited by these organisms (79). Over several nights after the full moon in late spring each year, hundreds of coral species spawn en masse, with the final trigger being changes in the lunar irradiance intensity (8, 11).

The specific cellular mechanisms involved in light detection by reef-building corals (Anthozoa, Cnidaria) have remained elusive. Biophysical data (4) show that corals are highly sensitive to blue light, which is also known to entrain the circadian clocks of insects and mammals (12) via cryptochromes (CRYs), which are DNA photolyase–like photoreceptor proteins. The roles of cryptochromes differ subtly between mammals and insects; the proteins function as circadian oscillator components in Mus but as photoreceptors for clock entrainment in Drosophila (13, 14). To date, CRYs have been identified only in higher animals such as vertebrates and insects, although related (and divergent) proteins have been reported in plants and eubacteria (12).

We used degenerate primers based on sequences conserved between Mus, Drosophila, Xenopus, and Danio to clone two cry genes from the coral Acropora millepora (15). The proteins encoded by the genes cry1 and cry2 each contain an N-terminal photolyase-related region (PHR) bearing two chromophore-binding domains and C-terminal domains extending 54 (CRY1) or 27 (CRY2) amino acids in length. cry1 and cry2 were also identified in expressed sequence tag (EST) data sets generated from A. millepora larvae (16), as were two additional genes known here as cry3 (15) and cry-dash. Because the cDNA library was constructed from aposymbiotic larvae, these cry genes are likely to be from coral rather than from associated symbiotic algae or marine microbes.

Phylogenetic analyses (Fig. 1) emphasize the similarity of coral CRYs and their vertebrate counterparts. Coral CRY1 belongs to the mammalian-type (m-type) CRY group. Both CRY1 and CRY2 are only distantly related to the Drosophila-type CRYs. CRY2 more closely resembles the Danio photoreceptor candidate CRY4-type (17) and is basal to the clade comprising both the m-type CRYs and the (6-4) photolyases. Coral CRY-DASH is a typical CRY-DASH protein (18) and is basal in the animal CRY-DASH clade. This analysis suggests that coral CRYs may represent ancestral members of the protein family in the animal kingdom, potentially providing insights into the origins of light perception in animals.

Fig. 1.

Phylogenetic relationships of the coral CRY and photolyase proteins resulting from maximum likelihood analysis of a representative range of CRY and photolyase proteins. Sequences analyzed were retrieved from public electronic databases. The accession numbers of each protein are given together with the genus name. Numbers on nodes represent percentages of 1000 bootstrap replicates, reflecting support for the topology shown.

On the GBR, we investigated (15) whether the expression of cry1 and cry2 in corals is rhythmic only under light/dark (LD) cycles as in Drosophila (13) or driven by an endogenous oscillator as in mammals (19). The coral genes exhibited daily rhythms under LD cycles, with peak expression at zeitgeber (time-giver, ZT) 3 for cry1 (ZT = 0 corresponding to first nautical light, which was 05:00 local time) and ZT 7 for coral cry2 [LD expression ratios were as follows: cry1, 5.4; cry2, 8.1 (Fig. 2, A and B)]. Cry1 mRNA levels were arrhythmic under continued darkness (DD), [analysis of variance (ANOVA) repeated measures (RM) P > 0.05], whereas cry2 expression changed significantly (P < 0.05) only at ZT 17 (P = 0.01) (Fig. 2, A and B). Experiments carried out 4 months later (under LD and DD conditions over 2 consecutive days) confirmed that the expression of both cry1 and cry2 is daylight-dependent, with significantly higher transcript levels under LD at ZT 6 (ANOVA RM P < 0.01). Under DD, cry1 expression exhibited low amplitude with significantly higher mRNA levels at 12:00 and 00:00 (P < 0.05), whereas cry2 expression did not significantly change under DD conditions (Fig. 2, C and D). The aberrant timing of expression under DD versus LD indicates a decay of rhythm in the absence of a ZT, suggesting that rhythmic expression of coral cryptochromes is not driven by an endogenous oscillator (20). Both genes exhibited a rhythmic expression pattern under LD, but with a different phase, suggesting that expression might be controlled by different transcriptional regulatory mechanisms within the same tissue environment.

Fig. 2.

Temporal expression patterns of cry1 and cry2 in A. millepora under LD (white squares) and DD (black circles) cycles analyzed with quantitative polymerase chain reaction. (A and B) A 32-hour cycle with sampling intervals of 4 hours (A). Quantitative analysis of cry1 revealed a significant effect of LD (P = 0.035), as well as a significant effect of sampling time (time, P < 0.001). (B) Expression of cry2 (LD, P = 0.026; time, P < 0.001). (C and D) a 42-hour cycle with sampling intervals of 6 hours. (C) Expression of cry1 (LD, P = 0.016; time, P < 0.001). (D) Cry2 expression (LD, P = 0.013; time, P < 0.001). There was no significant effect of sampling time X cycle (P > 0.05 for both cry1 and cry2 under 42-hour cycle treatment), followed by ANOVA with repeated measures. Each value was normalized to β-actin and converted to a percentage of the maximal level for each gene. Values (mean ± SE) were tested by ANOVA with a linear contrast method within groups in order to distinguish between the LD/DD rhythm amplitude of cry1 and cry2. For Cry1 DD, P > 0.01; cry1 LD, P < 0.01; cry2 DD, P > 0.05; cry2 LD, P < 0.01. Gray areas represent ambient darkness, and time points with asterisks are significantly different (gray asterisks, LD; black asterisks, DD). * represents P < 0.05, ** represents P < 0.01, and *** represents P < 0.001. Sample size = 3 colonies.

To test the involvement of cry genes in sensing and responding to moonlight, fragments from each of four colonies on the reef flat were sampled four times a day during the full and new moon phases of August and September 2005 (data are presented for 18:00 and 00:00 hours). Additionally, we sampled colonies on a full moon night in November 2005 5 days before the mass spawning event (fig. S1).

During full and new moon nights, cry1 showed similar transcript levels during August and September 2005 at 18:00 (ANOVA RM, P > 0.05) (Fig. 3A) (15). In contrast, coral cry2 showed significantly higher expression at midnight under full than new moonlight when compared at 18:00 (P < 0.05) (Fig. 3B). The expression pattern associated with cry2 suggests that it may entrain the intrinsic clock of corals to the lunar phase; however, roles for other photopigments such as opsins cannot currently be ruled out.

Fig. 3.

(A and B). Quantitative analysis of cry1 and cry2 in the 2 consecutive months of August and September 2005, comparing new moon nights to full moon nights at time points of 18:00 and 00:00. Each value is the average time point of the 2 months of sampling (mean ± SE). ANOVA RM, P > 0.05 for cry1, *P < 0.05 for cry2; sample size = 4 colonies. (C to F). Localization of cry1 and cry2 in the coral host tissue. Immunohistochemistry was performed with specific antibodies against CRY1 and CRY 2 in 72-hour larvae and in adult colonies of A. millepora. (C) Control. Larvae probed with preimmune serum were free of signals. (D) CRY2 is expressed (brown staining indicates the antibody-specific staining) in the ectoderm layer of the larval tissue. (E) and (F) Protein localization in the ectoderm layer of adult A. millepora corals when treated with specific antibodies against CRY1 (E) and CRY2 (F) (brown). ec, ectoderm; en, endoderm; mes, mesogloea. Scale bars, 50 μm. Red arrowheads mark expression of the antibody in (D) to (F); the rectangle shows one cell in the ectoderm layer.

Consistent with roles as photoreceptor proteins, immunohistochemistry revealed that both CRY1 and CRY2 proteins (15) are restricted to the ectoderm in both larval and adult corals (Fig. 3, D to F, and fig. S2B; control being negative, Fig. 3C and fig. S2A). The corresponding mRNAs showed the same patterns in in situ hybridization experiments (15, 21) (fig. S2, C and D).

Our discovery of cryptochromes in reef-building corals reveals that the basic mechanisms by which insects and mammal circadian oscillators respond to light were in place at the origins of multicellularity in animals. The presence of CRYs in a phylum close to the base from which all multicellular animals diverge supports the hypothesis that these proteins evolved under the blue light of the Precambrian ocean, possibly as a means to avoid high daytime ultraviolet levels near the surface (22). The expression patterns of coral CRYs in response to daylight (cry1 and cry2) and moonlight (cry2) also suggest that cryptochromes may mediate the spectacular mass spawning event of invertebrates (8, 9), adding an evolutionary dimension to circadian clock biology.

Supporting Online Material

Materials and Methods

SOM Text

Figs. S1 and S2


References and Notes

View Abstract

Stay Connected to Science

Navigate This Article