Dissociation of circadian and circatidal timekeeping in the marine crustacean Eurydice pulchra

Abstract

Background: Tidal (12.4 hr) cycles of behavior and physiology adapt intertidal organisms to temporally complex coastal environments, yet their underlying mechanism is unknown. However, the very existence of an independent "circatidal" clock has been disputed, and it has been argued that tidal rhythms arise as a submultiple of a circadian clock, operating in dual oscillators whose outputs are held in antiphase i.e., ~12.4 hr apart.

Results: We demonstrate that the intertidal crustacean Eurydice pulchra (Leach) exhibits robust tidal cycles of swimming in parallel to circadian (24 hr) rhythms in behavioral, physiological and molecular phenotypes. Importantly, ~12.4 hr cycles of swimming are sustained in constant conditions, they can be entrained by suitable stimuli, and they are temperature compensated, thereby meeting the three criteria that define a biological clock. Unexpectedly, tidal rhythms (like circadian rhythms) are sensitive to pharmacological inhibition of Casein kinase 1, suggesting the possibility of shared clock substrates. However, cloning the canonical circadian genes of E. pulchra to provide molecular markers of circadian timing and also reagents to disrupt it by RNAi revealed that environmental and molecular manipulations that confound circadian timing do not affect tidal timing. Thus, competent circadian timing is neither an inevitable nor necessary element of tidal timekeeping.

Conclusions: We demonstrate that tidal rhythms are driven by a dedicated circatidal pacemaker that is distinct from the circadian system of E. pulchra, thereby resolving a long-standing debate regarding the nature of the circatidal mechanism.

Graphical abstract

Figure 1

Tidal and Circadian Control of Behavior and Physiology in Eurydice (A) Shore-caught Eurydice show robust circatidal swimming in DD. An individual actogram, double plotted on 12.4 hr time base over 7 days, is shown. (B) The same data as in (A) double-plotted on a 24 hr time base to show more clearly the daily modulation of swimming episodes. (C) Periodogram for the animal in (A) and (B). Red line, p < 0.001 level. (D) Dorsal chromatophores of Eurydice and respective pigment dispersion index scale I to V. (E) Chromatophores of animals from the beach show pigment dispersion during the day (mean + SEM, F_1,145 = 2.13, p = 0.003). (F) Chromatophore pigment dispersion (mean + SEM) in Eurydice removed from the shore and released into DD. Gray/black bars show expected light regime on the home beach (see also Figure S1). (G) Chromatophore pigment dispersion (mean + SEM) in Eurydice entrained in reversed LD 12:12 and released into DD. (H) The tidal clock is temperature compensated. The period of swimming rhythms in beach-caught animals free running at 11°C, 17°C (ambient seawater temperature) and 21°C is shown. The red dotted line indicates a 12.4 hr period (mean + SEM, n = 32–58). (I) The daily modulation of tidal activity is temperature compensated (MI data mean + SEM, n = 32–58). See also Figure S1.

Figure 2

Manipulation of Tidal and Circadian Behavior and Physiology in Eurydice by Casein Kinase Inhibitor and Periodic Vibration (A) Chromatophore index (mean ± SEM) for animals in DD exposed to 25 μM PF670462 (red) or vehicle (blue). (B) Dose-response curve for daily modulation of tidal behavior (MI, mean + SEM) by PF670462. (C) Free-running actograms (left) for individuals administered different doses of PF670462 (red arrows) and their corresponding periodograms (right). (D) Free-running period of tidal activity rhythm shows a dose-response relationship for PF670462 (mean + SEM; see Figure S2 for PF4800567 results). (E) Representative activity trace shows entrainment of previously arrhythmic Eurydice tidal behavior by periodic vibration in DD (red arrows) after release to free run. (F) Periodogram reveals a 12.8 hr period in DD after vibration. (G) Representative activity trace shows that entrainment of tidal behavior by vibration does not restore amplitude modulation of swimming. (H) Modulation indices (mean + SEM) of groups of animals in DD taken from the beach (a, n = 14), during vibration entrainment (v, n = 15) and in free run after vibration (fr/v, n = 14). See also Figure S2.

Figure 3

Characterization of Canonical Eurydice Circadian Clock Genes (A) Only Eptim in Eurydice heads shows circadian cycling in DD. Mean abundance (±SEM) in copy number per 100 copies of the reference gene Eprpl32 is shown Horizontal bars, expected light and dark cycles; red arrowheads, time of expected high water; LW and HW, low and high water. LW2 is equivalent to CT0. (See Figures S3A–S3F and Table S1 for details of Eurydice clock genes.) (B) Cartoon of Eurydice brain illustrating the relative position of cells immunopositive to anti-EpPER sera. Red, dorsolateral (dl); pink, lateral (l); OG, optic ganglia; CB, central body; Oe, esophagus. (C) Anti EpPER immunoreactivity in the brain of Eurydice. The upper colored panels show dorsolateral cells at ZT15, counterstained with DAPI. The panel on the left shows strong cytoplasmic immunoreactivity. The right-hand panels show partial nuclear labeling in addition to cytoplasmic localization. The lower left shows paired dorsolateral (DL) cells taken from one hemisphere (ZT8). The lower right shows anti-EpPER-positive cell in the lateral (L) region (ZT8). Scale bars represent 15 μm. The black and white sections show that preabsorption of the EpPER antiserum with the cognate peptide (lower panel) eliminated immunostaining (arrows) of dorsolateral neurons (see also Figure S3G). (D) Anti-EpPER immunoreactivity of dorsolateral and lateral cells at different times of the day (ZT0, lights on; ZT12, lights off) showing predominantly cytoplasmic and partial nuclear labeling. Scale bars represent 10 μm. (E) EpPER antigenicity in dorsolateral and lateral cells under LD 12:12 cycles. (F) EpPER, EpTIM, and EpRCRY2 negatively regulate EpCLK-EpBMAL1-mediated transcription in Drosophila S2 cells. n = 3–6. (See also Figure S3H and Table S2 for EpPER functional analysis in transgenic flies.) (Fi) EpCLK isoforms 5 and 1-7 activate E-box mediated luciferase activity, whereas the other isoforms, 1-9 (missing part of PAS-B) and 1-4 (lost most of polyQ region), do not (see Figure S1E). Deletion of the putative BMAL1 C-terminal transactivation domain (+Δ) does not activate transcription. Mean Luc activity (+SEM) normalized to Renilla is shown. (Fii and Fiii) EpPER (Fii) and EpTIM (Fiii) modestly repress EpCLK-BMAL1 mediated activation. (Fiv) EpCRY2 robustly represses luciferase activity. See also Figure S3 and Tables S1 and S2.

Figure 4

Circadian and Tidal Phenotypes Can Be Separated by Environmental Manipulations (A) Constant light (LL) disrupts the chromatophore rhythm (mean ± SEM) (white bar, subjective day; gray bar, subjective night). (B) LL (right) disrupts the amplitude modulation of tidal swimming (individual plots normalized to maximum activity) evident in DD (left). (C) Mean MI on DD and LL (mean + SEM; n = 21 and 26 for DD and LL, respectively; gray horizontal line, MI = 0.5). (D) Period of tidal swimming rhythm (left; gray line, 12.4 hr) and overall swimming activity (right) under DD and LL (mean + SEM; n = 30 for both LL and DD). (E) Expression of Eptim in heads of Eurydice held under DD or LL (mean ± SEM).

Figure 5

Circadian and Tidal Phenotypes Can Be Separated by Knockdown of Epper by RNAi (A) Knockdown of Epper by dsRNAi injections (mean ± SEM). (B) Mean chromatophore index (±SEM) for control (blue)- and Epper dsRNAi (red)-treated animals maintained in DD. Gray/black bars, subjective LD cycle. (C) Normalized Eptim transcript levels in DD (mean ± SEM) for vehicle control (blue), Epper dsRNAi (red), or control Discoplax celeste molt-inhibiting hormone (Dcmih) dsRNA controls (black). n = 3 for vehicle and Epper manipulation but n = 1 for Dcmih. There was no significant difference between Epper transcript levels in the heads of sham versus vehicle controls (t = 1.9, df = 9, p = 0.11). Gray/black bars, subjective LD cycle. (D) Tidal swimming period in DD is not altered by Epper dsRNAi. The left-hand panels show grouped swimming behavior of vehicle (blue)- and dsRNAi (red)-injected animals. Gray/black bars, subjective LD cycle. The right-hand panels show mean tidal period and mean activity levels (+SEM).