Circadian oscillators are part of the cell’s endogenous mechanism for maintaining synchrony with daily environmental changes during the day and night cycle. They have been studied in many multicellular organisms but are also known from unicellular eukaryotes and prokaryotes (Huang et al., 1990). In some model organisms, such as Arabidopsis thaliana, Neurospora crassa, Mus musculus, Drosophila melanogaster, and Synechococcus, the components and mechanism of circadian clocks are known in great detail (Kondo and Ishiura, 2000; Baker et al., 2012; Mohawk et al., 2012; Nakamichi, 2020; Doležel, 2023), while in others, the system is just being described (Poliner et al., 2019; Farré, 2020; Borges-Pereira et al., 2021).

The simplest circadian mechanism described to date, the post-translationally controlled Kai oscillators, operates in phototrophic prokaryotes and is based on the cooperation of only three proteins: KaiA, B, and C (Johnson et al., 2017; Swan et al., 2018). Studies have shown that the circadian machinery of eukaryotes relies on a similar mechanism of several interlocked transcription-translation feedback loops (TTFL) controlled by environmental factors such as light and temperature (Poliner et al., 2019). Although they operate on the same general principle of TTFL, the key components of the circadian mechanism are substantially different among organisms. The formal rules vary among genetic communities; here we use capital italicized abbreviations for genes (CCA) and capital abbreviations for proteins (CCA). Neurospora crassa served as a model organism to identify the fungal circadian clock. Its mechanism is based on the interplay between FRQ (frequency) and WC-1,2 (white-collar) from PAS (PER/ARNT/SIM) protein family (Crosthwaite et al., 1997; Poliner et al., 2019). The PAS family proteins also form the core of the animal circadian clock mechanism represented by PERIOD, and basic helix-loop-helix PAS (bHLH-PAS) proteins CYCLE, CLOCK, and BMAL, whose activity is inhibited by animal-type cryptochromes (Dunlap, 1999; Poliner et al., 2019). Involvement of bHLH-PAS protein RITMO1 was reported for the marine diatom Pheodactylum tricornutum, and homologous sequences were discovered also in alveolates, cryptophytes and rhodophytes (Annunziata et al., 2019). However, the bHLH-PAS/RITMO1 protein was found in early branching alevolates, heterotrophic cilliates, and in dinotoms, dinoflagelates that harbour the diatom endosymbiont. It is very likely that the bHLH-PAS protein in dinotoms is present in the diatom endosymbiont, which is almost unreduced and contains a diatom nucleus, mitochondrion, and plastids (Yamada et al, 2017). The clock system of phototrophs is unique for the prevalence of repressor elements (Nohales and Kay, 2016). Contrary to expectations, no homolog of the cyanobacterial Kai family gene was found in eukaryotic phototrophs. The circadian mechanism of green algae and plants is characterized by the presence of MYB family proteins (such as LHY, Late Elongated Hypocotyl and CCA1, Circadian Clock Associated) and pseudo response regulators (PRRs; Poliner et al., 2019; Petersen et al., 2022).

The circadian rhythms of eukaryotic phototrophs with plastids of red lineage are much less explored. However, with current tools for omics analyses, new promising targets for the study of primary red algae are emerging (Miyagishima and Tanaka, 2021; Hirooka et al., 2022). Most studies of red lineage circadian clocks have been conducted on stramenopiles, a lineage whose plastids evolved through complex endosymbioses. Self-sustaining 24-hour rhythms (running under constant conditions) have been described in diatoms (Ragni and D’Alcalà, 2007), brown algae (Phaeophyceae) (Schmid and Dring, 1992; Schmid et al., 1992) and the eustigmatophyte Nannochloropsis (Poliner et al., 2019; Farré, 2020). The molecular model has previously been outlined for the marine diatoms Phaeodactylum tricornutum (Annunziata et al., 2019; Farré, 2020; Zhang et al., 2022) and Skeletonema costatum (Zhang et al., 2022). RITMO1 is the first and so far the only clock component for stramenopiles. It resembles an animal bHLH-PAS (basic Helix-Loop-Helix PAS) clock component, and the coding gene is widely distributed in marine algal genomes, with exception of dinoflagellates (Annunziata et al., 2019). Other proteins with bHLH-single PAS domains have also been found in dinoflagellates, ciliates, and the cryptophyte alga Guillardia theta (Farré, 2020). A special chapter in the circadian story of organisms with complex, rhodophyte-derived plastids should belong to dinoflagellates. Their oscillator is based neither on the eukaryotic TTFLs nor on the post-translational prokaryotic Kai system. Studies have revealed proteins with circadian rhythmicity, but with stable mRNA levels (Morse et al., 1989; Milos et al., 1990; Roy et al., 2014). Dinoflagellates have a unique circadian mechanism based mainly on translational level dynamics, with casein kinase 2 possibly playing a crucial role (Roy et al., 2014; Jadhav et al., 2022). Sporozoa, obligate apicomplexan parasites with cryptic complex plastids of red lineage, must cope with incomparable conditions. For example, their circadian rhythms have been found to be synchronized with the host, but also to function independently. The circadian mechanism in sporozoans remains unknown (Rijo-Ferreira and Takahashi, 2020; Rijo-Ferreira et al., 2020).

The metabolism of phototrophs strongly relies on sunlight, which provides the energy for photosynthesis. The circadian clock system of phototrophs is therefore permeated by various photoreceptors that bring information about the quality, intensity, and direction of light (Nohales and Kay, 2016; Lopez et al., 2021). Photoreceptors differ in the wavelength they absorb. Red and far-red light is absorbed by phytochromes, blue light is predominantly perceived by cryptochromes, phototrophins, and Zeitlupe, and finally UV-light is caught by proteins in the UV RESISTANT LOCUS 8 family (Lopez et al., 2021; Petersen et al., 2021). In addition, the function of photoreceptor homologs may differ between plants and algae (Petersen et al., 2021).

Taken together, the core circadian clock components are unique for specific lineages across the tree of life, including the photosynthetic organisms, when cyanobacteria, green lineage phototrophs and red lineage phototrophs built their oscillatory mechanism based on genes from various progenitors. An exception are cryptochromes that seem to be present at least in all lineages of eukaryotic phototrophs (Farré, 2020; Cordoba et al., 2021; Petersen et al., 2021; Jadhav et al., 2022).

Due to its unique position within the alveolates as the closest photosynthetic relative of apicomplexan parasites (such as malaria’s causative agents Plasmodium), Chromera velia has been intensively studied since its discovery (e.g. Moore et al., 2008; Kořený et al., 2011; Oborník et al., 2011; Janouškovec et al., 2013). During the peculiar life cycle, C. velia periodically undergoes the flagellated zoospore stage (Oborník et al., 2016). It has been shown that zoospores peak about six hours after the beginning of the light phase and disappear before its end (Oborník et al., 2011). In this study, we aim to determine whether this periodicity exhibits circadian behavior and identify putative homologs of circadian genes in C. velia.

Our planet revolves around the Sun and rotates around its own axis (Copernicus, 1543). This facts, which causes an alternation of day and night, has influenced all life forms that have ever lived on Earth (Giebultowicz, 2004). The phenomenon of circadian rhythms was first described in 1729 by Jean-Jacques d’Ortous de Mairan using the periodical movement of Mimosa leaves. Circadian rhythms are still being studied and are not fully understood, especially in non-model organisms. Recently, circadian rhythmicity has even been found in unicellular endoparasites (Rijo-Ferreira and Takahashi, 2020).

Chromera velia is a free-living marine alga attracting attention from the scientific community because it is the closest known photosynthetic relative of obligate sporozoan parasites such as Toxoplasma and Plasmodium (Moore et al., 2008). The life cycle of C. velia involves the process of zoospore formation, which is controlled by light and occurs with a periodicity of 24 hours (Oborník et al., 2011). In this work, we 1) investigated the influence of light on zoospore formation of C. velia, 2) searched for the presence of known circadian homologs and 3) described their phylogeny, 4) examined the amino acid sequence of CPF homologs for the presence of conserved motifs and target sequences, 5) developed a method for isolation of pure C. velia zoospores from the culture and 6) measured the rhythmicity of the mRNA level for some of the homologs found.

Although photosynthetic organisms have many physiological processes that are influenced by circadian rhythms, not all of them can be easily followed in vivo (Farré, 2020; Jadhav et al., 2022). Zoosporogenesis of C. velia can be easily observed with an ordinary microscope under appropriate conditions. The release of zoospores occurs rhythmically, beginning at dawn, peaking at midday, and declining before the dusk (Oborník et al., 2011). The zoospore release under constant dark continues only for three cycles. In contrast, zoospore release remains robust under constant light, but is non-cyclical. When a C. velia culture maintained in a standard 12/12 light/dark regime is transferred to constant dim light conditions, the rhythmic release of zoospores can be followed for five more cycles during the subjective day. This periodic release of zoospores under constant conditions suggests the presence of a ‘ticking’ circadian clock in C. velia. The existence of this internal timing mechanism is also confirmed by our phase shift experiments, in which C. velia did not respond immediately to the new light regime. Instead, the release of zoospores continued in a similar phase for another cycle, and the broader distribution of zoospores lasted even longer, eventually leading to a successful phase shift. Indeed, the ability to adjust to a new regime, in contrast to an immediate response, is the hallmark of circadian clocks (Roenneberg and Morse, 1993).

From a practical point of view, the more cycles under constant conditions, the better the experimental possibilities. The transition to constant light led to disruption of the cycle, leading to continuous zoosporogenesis from the second day of the transition ( Figure 1 ), indicating a high light sensitivity of this process. Determining the exact reactivity spectrum of C. velia zoosporogenesis to light intensity would be informative to some extent, but it is beyond the scope of this manuscript.

The role of C. velia zoospores is likely to be to disperse in the environment in search of sufficient living conditions and the zoospores are supposed to be an infective stage. In contrast to the second chromerid V. brassicaformis (Füssy et al., 2017), no fusion of zoospores was observed and the key meiotic genes are also absent (Vazač et al., 2018). C. velia infects both the ectoderm and endoderm of coral larvae (Cumbo et al., 2013; Mohamed et al., 2018). Zoospores are best able to carry out the infection process because they possess the conoid (Oborník et al., 2011; Portman et al., 2014), the conserved core element of apical complex that enables the invasion process in Sporozoa (Kořený et al., 2021). Both activities, opening up a new environment and searching for a host, require a rapid movement that is energetically demanding. The long lifespan of C. velia zoospores may require photosynthetic activity even in the motile stage of the alga, and photosynthesis may also support the proposed photoparasitic lifestyle (Oborník, 2020). With increasing depth, the composition of the light spectrum changes rapidly and red light is almost absent below 15 m. At depths greater than 30 m, blue light becomes increasingly enriched (Duanmu et al., 2017). Here we have shown that for successful transition from coccoids to zoospores, the blue spectrum of the light source is a necessary condition ( Figure 2 ). Blue-light-driven release of zoospores could be an efficient way to maintain this important cell transition process at greater depth. We observed that zoospores are photophobic. Therefore, C. velia zoospores most likely actively search for an environment with lower radiation intensity. By colonizing a shaded environment, C. velia expands its ecological niche where it can thrive thanks to its adapted photosynthetic apparatus (Kotabová et al., 2014). It was found that zoospores are able to actively invade the coral larvae (planula, Cumbo et al., 2013). Planulae form new coral colonies by dropping to the seafloor (Dana, 1846). Therefore, pointing the light source in the opposite direction of the zoospore may be helpful in finding a new host. Since we know the average speed and lifespan of the C. velia zoospore (see chapter 3.2), we can assume that the zoospore can swim over eight meters in search of a suitable environment after its release from the zoosporangium.

We searched the currently available genome of C. velia (Woo et al., 2015) for homologs of known blue light receptors and were successful only for cryptochromes. The only light receptors present in most evolutionary lineages (Lopez et al., 2021) that evolved from photolyases, blue light-dependent DNA repair enzymes (Ahmad and Cashmore, 1993; Hegemann, 2008; Petersen et al., 2021). In C. velia we found six homologs of cryptochrome photolyase family proteins ( Table 1 ). Our phylogenetic analyses show that the C. velia homologs of the CPF family are scattered throughout the tree and have a representative in plant-like, animal, and CRY-DASH1,2 families ( Figure 4 ). Cvel_8422.t1 is located at the base of the CRY-DASH group along with the V. brassicaformis homolog. Some CRY-DASH proteins retain DNA repair activity under stringent conditions (Chaves et al., 2011) and may reside in organelles (Kleine et al., 2003; Rredhi et al., 2021). Nuclear localization signal has been confirmed at the C terminus of mammalian cryptochromes (Mei and Dvornyk, 2015). Localization of cryptochromes in plants can vary between the nucleus and cytosol and is light-dependent (Xu et al., 2009). We screened CPF homologs of C. velia for the presence of conserved motifs and localization signals. In all homologs, we found both alpha/beta photolyase and FAD domains that typically form a photolyase homology region located at the N-terminus of the cryptochrome, suggesting an active role in the light-mediated response and a possible role in the mechanism of C. velia circadian clock ( Figure 3 .; Petersen et al., 2021).

The C-terminus of cryptochromes is much more variable but has a major impact on cryptochrome function and is often involved in interactions with partner proteins (Chaves et al., 2011). The longest C-terminus is located in Cvel_8422.t1, where the two NLSs were also detected, suggesting that this CPF homolog can be targeted to the nucleus ( Figure 3 ). The potential NLS domain was also predicted at the N-terminus of the Cvel_23588.t1, the CPF homologue with second longest aa sequence. The potential DNA repair activity of these homologs is difficult to verify because there is no expression system in C. velia. Organellar targeting signals were detected for the animal-like CPF homolog (Cvel_7245.t1) and one of the two CRY_DASH homologs (Cvel_13989.t1), both of which have a mitochondrial transit peptide ( Figure 3 ). None of the examined C. velia CPF homologs had a positive plastid targeting signal. Mitochondrial targeting of the above- mentioned CPF homologs may enhance the physiological response of C. velia to environmental changes, thereby increasing its flexibility.

In Arabidopsis, two closely related myb proteins expressed in dawn, CCA1 and LHY, are required for clock function. Their homologs are likely found in all green plants, including the moss Physcomitrium patens (Okada et al., 2009; Cervela-Cardona et al., 2021). Our BLAST search revealed only two candidate myb-like proteins in C. velia. Although phylogenetic analysis was relatively limited due to the short and divergent sequence, both C. velia homologs branched at the base of a large clade containing CCA1 and LHY from the green plastid lineage. However, several other myb-containing proteins from A. thaliana are part of this clade. So far, there is no functioning transformation system in C. velia. Thus, without functional evidence, the role of the myb homologs from C. velia in the clock remains unclear.

A characteristic feature of circadian rhythms is their rhythmic oscillation within a 24-hour period. This rhythmicity is entrained by external cues (such as a light source) but persists under constant conditions within a nearly 24-h period even in the absence of external stimuli (Kuhlman et al., 2018; Farré, 2020). An easily detectable physiological feature with circadian properties in C. velia is zoosporogenesis ( Figure 1 ). Currently, three main types of circadian oscillators are known: eukaryotic systems relying on interlocked transcription-translation feedback loops (TTFL; Lande-Diner et al., 2013), prokaryotic post-translationally controlled Kai oscillators (PTO; Golden and Canales, 2003), and presumably ancient metabolic/redox oscillators conserved in all lineages (Edgar et al., 2012).Within the eukaryotic TTFL system, we can also identify the type of circadian clock conserved in bilaterian animals (Kotwica-Rolinska et al., 2022), with some components present even in basal animal lineages such as cnidarians (Reitzel et al., 2010). The plant system relies on different proteins involved in TTFLs that are conserved, with some idiosyncrasies, from moss to Arabidopsis (Holm et al., 2010). The existence of another oscillator cannot be ruled out, since dinoflagellates do not fit into either the TTFL or PTO oscillators and the metabolic/redox oscillator function has not been proven in them (Jadhav et al., 2022). The circadian cycle of most dinoflagellate proteins has been found to be regulated at the translational level (Milos et al., 1990; Roy et al., 2014; Jadhav et al., 2022).

Although the bHLH-PAS protein has been found in ciliates, an early branching alveolates (Annunziata et al., 2019), it is probably absent in advanced alveolates such as dinoflagellates, apicomonads (including C. velia), and sporozoans. Dinotoms, dinoflagellates with the diatom endosymbiont are probably the only exception, but the bHLH-PAS/RITMO1 protein is probably present in the unreduced diatom endosymbiont rather than in the dinoflagellate host cell. Therefore, it appears that the particular type of circadian clock that uses bHLH-PAS/RITMO1 proteins has been lost with the acquisition of the complex plastid in alveolates (dinoflagellates, apicomonads, and sporozoans).

C. velia, together with other Apicomplexa, forms a sister branch of dinoflagellates (Moore et al., 2008) and inhabits a biogenic sediment of stony corals (Mathur et al., 2018), periodically releasing zoospores into the environment (Oborník et al., 2011). We developed a new method for separating the zoospores of C. velia that will certainly be helpful in experimental work with this life stage, which is probably involved in the infection of coral larvae. Our observations showed that the periodicity of zoosporogenesis follows almost a 24-hour free-running rhythm. In our search for homologs of circadian clock genes, we found two genes possibly related to A. thaliana CCA/LHY. We also found six homologs of blue light photoreceptors from the cryptochrome family. However, without functional transformation tools in C. velia, we cannot demonstrate their role in the circadian rhythm of C. velia. Thus, although the circadian clock exists in C. velia, its molecular mechanism remains elusive and could either rely on clock gene homologs that are so modified that they could not be identified, or the C. velia clock uses completely different clock components.

The original contributions presented in the study are included in the article/ Supplementary Materials , further inquiries can be directed to the corresponding authors.

JR prepared Figures 1 - 3 and Supplementary Figure 1 . AH prepared Figures 4 , 5 and Supplementary Figure 2 . DD prepared Figure 6 , Table 1 , and Supplementary Table 1 . JR and DD designed and analyzed experiments for circadian rhythmicity and the influence of light wavelength. JR and VGG performed the time laps counting. AT run the cultivations for expression analyses. OB and DD designed and analyzed the gene expression part of the research. AH and DD identified homologs and reconstructed their phylogeny. DD and MO designed and led the study. JR, DD, AH, and MO wrote the paper, with input from all authors. All authors contributed to the article and approved the submitted version.