Transcriptome survey of phototransduction and clock genes in marine bivalves

INTRODUCTION

Marine ecosystems are governed by a multitude of environmental cycles, such as the light-dark cycle and lunar-related cycles, which are linked to the periodic recurrence of the sun or the moon, and are known as “marine rhythms” (Tessmar-Raible et al., 2011). Most terrestrial organisms possess a circadian clock, which controls metabolic reactions and physiological processes according to daily light-dark cycles (Panda et al., 2002; Schrago, 2006; Golombek and Rosenstein, 2010; Shi and Zheng, 2013). In addition to the circadian rhythm, marine organisms also exhibit lunar-related rhythms that are regulated by lunar periodicity (Ikegami et al., 2014). In contrast to the circadian rhythm, lunar cycles in marine organisms are most prominently involved in reproductive events, such as synchronous mass spawning (Tessmar-Raible et al., 2011). Lunar-controlled reproductive rhythms are widespread and are of fundamental importance for many marine organisms, including corals, sea urchins, mollusks, and fishes (Vize, 2009; Fukushiro et al., 2011; Ikegami et al., 2014). For these marine species, the lunar cycles provide a steady timeframe that can be used for developmental synchronization and mass spawning, even across widespread populations (Tessmar-Raible et al., 2011). For example, in order to ensure successful fertilization, spawning corals utilize an extraordinarily accurate timing system to reproduce according to solar and lunar cycles (Vize et al., 2005). As reported, many species spawn on just one or two evenings per year in a time window only 20-30 min long, which is thought to be controlled by both lunar and solar cycles (Levitan et al., 2004; Vize, 2009).

A number of molecular components involved in circadian rhythms have now been discovered, including the circadian locomotor output cycles kaput (clock), period (per), timeless (tim), and cryptochrome (cry; Wager-Smith and Kay, 2000; Golombek and Rosenstein, 2010; Pairett and Serb, 2013). However, the molecular understanding of circadian rhythms is almost exclusively related to terrestrial species. A number of studies have revealed that the expression of some clock genes, such as per, cry, and melatonin, is affected by both daylight and moonlight, suggesting interconnections between solar and lunar rhythms in animals (Levy et al., 2007; Sugama et al., 2008; Li et al., 2015a). For marine species, molecular components of “marine rhythms” are essential to understand how marine species detect solar and lunar cycles and translate this into the sequence of physiological processes that result in the maturation of gametes and determine the timing of spawning.

Identification and understanding of the expression patterns of clock-controlled genes (CCGs) will help elucidate “marine rhythms” found in marine ecosystems, improve aquaculture, and enhance conservation of natural populations. As a high-throughput method for transcriptome sequencing, RNA-Seq has been widely applied to uncover genes central to the circadian clock at the transcriptome level (Vize, 2009; Pairett and Serb, 2013; Shoguchi et al., 2013). To explore the potential genes underlying biological rhythms in marine species and investigate how these genes are correlated with terrestrial models, we made use of the recent generation of a large expressed sequence tag (EST) dataset developed from the mantle transcriptomes of marine bivalves (Pairett and Serb, 2013; Sun et al., 2015). The analysis allowed us to identify most of the key circadian genes from insects and mammals in the transcriptomes of marine bivalves, which indicated that many of these genes are very highly conserved in the animal kingdom. The findings provide a valuable gene resource for studies of “marine rhythms” and further our understanding of the diversification and evolution of rhythms in marine species.

MATERIAL AND METHODS

RESULTS

Summary of genes related to phototransduction and the circadian clock in bivalves

The available expression data made it possible to compare the phototransduction and circadian clock genes among four bivalve species, P. magellanicus, A. irradians, P. yessoensis, and C. gigas. The 29 phototransduction and 8 circadian clock genes are summarized for the four bivalve species in Table 2. Most of the known genes were consistently found in P. yessoensis and C. gigas, having 28/29 phototransduction genes and 8/8 circadian genes expressed in their mantle tissue. Of the phototransduction genes, PDE gamma expression was not identified in P. yessoensis, while trp-like expression was not detected in C. gigas (Table 2). For the other two species, 21/29 phototransduction genes and 6/8 circadian genes were found in P. magellanicus, while 9/29 phototransduction genes and 0/8 circadian genes occurred in A. irradians. Based on the transcriptome data, we predicted the model of the circadian clock in scallops. In the first feedback loop, the Clock:Cycle complex activates transcription of repressor genes Per and Tim, and their proteins inhibit the transcriptional activity of Clock:Cycle (Figure 1). In the second feedback loop, because the expression of Rev-erb and Vri are not detected, the Clock:Cycle may activate the transcription of the RAR-related orphan receptor (Ror) and the PAR domain protein (Pdp), and their proteins feedback to repress or activate Clock mRNA transcription.

Table 2

Expression of key phototransduction and circadian clock genes in the mantle or eyes of mollusks.

Pathway	Gene	Mollusk species
		P. magellanicus	A. irradians	P. yessoensis	C. gigas
Phototransduction	Arrestin	√		√	√
	Calmodulin	√	√	√	√
	CDSA (CDP diglyceride synthetase)			√	√
	CNGA (cGMP-gated cation channel alpha)	√		√	√
	CNGB (cGMP-gated cation channel beta)	√		√	√
	G-protein, alpha subunit	√	√	√	√
	G-protein, beta subunit	√	√	√	√
	G-protein, gamma subunit	√		√	√
	guanyl cyclase			√	√
	inaC (inactivation no afterpotential C)	√		√	√
	inaD (inactivation no afterpotential D)			√	√
	ninaC (neither inactivationnor afterpotential C)		√	√	√
	Opsin	√	√	√	√
	PDE alpha	√		√	√
	PDE beta	√		√	√
	PDE gamma				√
	Phosducin			√	√
	phospholipase C	√		√	√
	PI synthase	√		√	√
	PI3K	√		√	√
	PIP kinase	√	√	√	√
	protein kinase A	√	√	√	√
	RdgA (retinal degeneration A)	√	√	√	√
	RdgB	√		√	√
	Recoverin		√	√	√
	Rhodopsin			√	√
	Rhodopsin kinase	√		√	√
	TRP	√		√	√
	TRPL	√		√
Circadian clock	Arntl			√	√
	casein kinase I			√	√
	Clock	√		√	√
	Cry	√		√	√
	Cycle	√		√	√
	Doubletime	√		√	√
	Period	√		√	√
	Timeless	√		√	√

Figure 1

Hypothesized molecular framework of the circadian clock for the Yesso scallop Patinopecten yessoensis compared to that in two model organisms, a fly and a mammal, as summarized by Shi and Zheng (2013). Dashed black lines indicate the hypothetical relationships based on expression data from the current study. P, phosphorylation; Per, Period; Tim, timeless; Cry, Cryptochrome; Pdp, PAR domain protein; Bmal1, Brain and muscle arnt-like protein 1; Vri, Vrille; Ror, receptor-related orphan receptors.

The expression data for Placopecten magellanicus and Argopecten irradians are summarized from Pairett and Serb (2013). Data for gene expression in the mantle of Crassostrea gigas were downloaded from an online database (http://www.oysterdb.com/). Data for Patinopecten yessoensis are from the current study.

DISCUSSION

As an important marine species, scallops are present in all of the World’s oceans. The mantle of the scallop not only secretes biomineralization proteins along its edge to induce shell formation, but it is also responsible for light-mediated functions and circadian rhythms due to 30-200 single chambered eyes along the mantle edge (Parsons et al., 1992; Pairett and Serb, 2013). In the present study, comparison of the phototransduction and circadian clock genes among four bivalve species indicated that the lack of target genes in scallops of P. magellanicus and A. irradians may be due to the low read depth in their transcriptomes, which both have less than one million reads (114,052 reads for A. irradians and 654,002 reads for P. magellanicus) generated from only eye tissue. In contrast, there were a total of 55.9 million reads in our assembled transcriptome of P. yessoensis produced from the mantle tissue-containing eyes, in which most phototransduction genes and all key genes related to circadian clock function were detected during this study. Indeed, experimental studies have demonstrated that 10-15 million reads in flies and 20-25 million reads in mice may be sufficient to characterize the majority of the circadian transcripts (Li et al., 2015b). Based on that, ~1/10 of a million reads is insufficient to detect the circadian transcripts in A. irradians in comparison with ~56 million reads in the mantle transcriptome of P. yessoensis.

These results suggest that the number of identified phototransduction and circadian transcripts mainly depends on read depth, rather than tissue type, e.g., eye or mantle. For scallops, although their eyes along their mantle edge are potentially an important entry for light entrainment of the clock, some of the clock genes may originate from extraocular sources in the mantle tissue (Pairett and Serb, 2013). Moreover, scallop eyes are generally so small that they may be insufficient to meet the sample requirement for the next-generation sequencing platform. It is therefore suggested that scallop mantle, including eyes, is the optimal tissue used in the next-generation sequencing platform to identify phototransduction and circadian transcripts, and the requirement of sequencing depth should be at least 10-20 million reads (Li et al., 2015b).

Identification of most phototransduction and circadian clock genes in this study indicates that these genes are highly conserved, not only in insect and mammals, but also in marine bivalves and corals (Shoguchi et al., 2013). Six key genes central to the circadian clock in insect and mammals, including clock, cry, cycle, doubletime (DBT), period (per), and timeless (tim), are consistently detected in three bivalve species, which suggests that the circadian network may also function in marine bivalves. For terrestrial model species, such as mammals and insects, the circadian clock mechanism consists of both negative and positive feedback loops that drive rhythmic expression patterns of core clock components (Lowrey and Takahashi, 2000). In mammals, the negative feedback loop involves a rhythmic transcription of per and cry (Golombek and Rosenstein, 2010). The transcription factors Clock and Bmal1 interact and activate per and cry expression by binding to an E box in their promoter regions. PER and CRY proteins form the heterodimeric complex and translocate to the nucleus, which acts on the BMAL1:CLOCK heterodimer to repress its own transcription. The degradation of PER and CRY proteins is phosphorylated by CK1ε (casein kinase 1 epsilon). For the positive feedback loop, BMAL1:CLOCK activates mRNA expression of nuclear hormone receptor genes Rev-erb and Ror, and their proteins feedback to negatively and positively regulate Clock and Bmal1 transcriptions (Golombek and Rosenstein, 2010). Similarly, in the fruit fly, the negative and positive feedback loops also operate in the circadian clock system (Hardin, 2005). The Clock:Cycle heterodimer complex activates transcription of repressor genes Per and Tim (Figure 1). Accumulated PER and TIM proteins form heterodimers to repress the transcriptional activity of the Clock:Cycle. The PER and TIM proteins are phosphorylated by the CK1ε homolog doubletime, followed by degradation. In another feedback loop, Clock:Cycle activates the transcription of Vri and Pdp, and their proteins feedback to repress and activate Clock mRNA transcription, respectively (Golombek and Rosenstein, 2010).The present results support the inference that the basic framework of clock regulation is highly conserved not only in terrestrial species, from insects to rodents and humans, but also in marine mollusks (Panda et al., 2002; Lowrey and Takahashi, 2000; Shi and Zheng, 2013), which suggests a convergent evolution of circadian timing mechanisms and raises the possibility that the underlying timekeeping mechanism may also be conserved.

Interconnections between circadian and lunar rhythms are found in some marine organisms, such as corals and fishes (Levy et al., 2007; Vize, 2009; Fukushiro et al., 2011; Ikegami et al., 2014). It is suggested that the rhythmic expression of melatonin may serve as a possible transducer between lunar spawning rhythmicity and circadian rhythms in fishes (Ikegami et al., 2014). Consistently, we detected a strong negative relationship in the expression level between melatonin and the clock genes (per and cry), wherein melatonin generally has a 10-fold higher expression level (RPKM) than per and cry. However, the interconnections between circadian rhythms and lunar-related reproductive activity in marine bivalves need to be further examined in physiological and experimental studies. Additionally, the transcriptome survey of clock genes in this study will not only help understand the principles, interactions, and evolution of rhythms in marine species, but will also be useful to measure the effects of climate change or light pollution in the marine ecosystem.

We performed a comprehensive survey to detect phototransduction and circadian genes in the mantle transcriptome of the scallop P. yessoensis, and compared the results with three other bivalves. The results revealed the transcripts of most core members from phototransduction and circadian networks in terrestrial model species, such as arrestin, calmodulin, guanyl cyclase, inaC, opsin, phosducin, recoverin, clock, period, timeless, and cycle. The present findings indicate conservation of the timekeeping mechanism from terrestrial species to marine bivalves. Moreover, the identification of clock genes will provide a valuable gene resource for “marine rhythms” and help understand the diversification and evolution of rhythms in marine species.

Supplementary material