Abstract
The regulation of circadian gene expression remains largely unknown in farmed fish larvae. In this study, a high-density oligonucleotide microarray was used to examine the daily expression of 13,939 unique genes in whole gilthead sea bream (Sparus aurata) larvae with fast growth potentiality. Up to 2,229 genes were differentially expressed, and the first two components of Principal Component Analysis explained more than 81% of the total variance. Clustering analysis of differentially expressed genes identified 4 major clusters that were triggered sequentially, with a maximum expression at 0 h, 3 h, 9–15 h and 18-21 h zeitgeber time. Various core clock genes (per1, per2, per3, bmal1, cry1, cry2, clock) were identified in clusters 1–3, and their expression was significantly correlated with several genes in each cluster. Functional analysis revealed a daily consecutive activation of canonical pathways related to phototransduction, intermediary metabolism, development, chromatin remodeling, and cell cycle regulation. This daily transcriptome of whole larvae resembles a cell cycle (G1/S, G2/M, and M/G1 transitions) in synchronization with multicellular processes, such as neuromuscular development. This study supports that the actively feeding fish larval transcriptome is temporally organized in a 24-h cycle, likely for maximizing growth and development.
Similar content being viewed by others
Introduction
The evolution of many organisms has been driven by circadian rhythms to adapt to periodic events in their external environments. These self-sustained and entrainable 24-h rhythms begin in the early stages of development and rely on tight regulation of gene expression. Fish are the most diverse vertebrate group and have evolved in quite different habitats (e.g., freshwater vs. sea water, tropical vs. temperate, diurnal vs. nocturnal), coordinating metabolic processes in a timely fashion1. This metabolic coordination is particularly relevant for farmed fish species, whose growth, health, and well-being may rely on synchronization between endogenous rhythms and external clues imposed by production practices. Therefore, information on daily transcriptome organization for a given species would enable further comparison of farming conditions, evaluate the adaptation capacity to changing environments, identify reliable markers of health and performance, and refine feeding practices by matching diet composition or the time of diet delivery to the temporal requirements of the organism. Thus far, massive gene expression analyses addressing clock-driven transcription in fish larvae have mostly been conducted in zebrafish2,3. Both microarray and RNA-Seq gene expression profiling of whole-body larvae of other teleostean fishes have been used to unravel key issues of development4,5, domestication6, and the effects of different environmental factors such as pollutants, diet, stress, and infection7,8,9,10,11,12 but not to elucidate overall daily transcriptomic organization.
Daily gene expression is driven, although not exclusively, by circadian clocks. The current vertebrate circadian model involves a positive core loop with heterodimer BMAL1/CLOCK that transactivates promoters with circadian clock-responsive elements such as E-box (e.g., in period genes, per) and triggers different transcriptional cascades13. The negative core loop of the model includes PER and cryptochromes (CRY), which after gradual accumulation can suppress BMAL1/CLOCK transcriptional activity14. PER and CRY are progressively phosphorylated and targeted for degradation, allowing reactivation of clock/bmal1 15. Additional ancillary loops drive the alternate activation and repression of bmal1 16, per and nr1d1 (rev-erbα) genes17. Significant advancements have been made in understanding the functioning of these clock components in commercially important fish species. The available studies have focused mostly on juvenile or adult stages and on specific tissues such as the pineal18 and liver19 in salmon, the pineal in the European sea bass18, and the brain and liver in the gilthead sea bream (Sparus aurata) 20. However, a study in S. aurata demonstrated that, as in other vertebrates, there are differences in the entrainment of central and peripheral clocks, being the liver clock regulated by feed rather than by light cycles20. Different to juveniles, larvae from farmed fish are cultured under constant feed availability and, in theory, with both central and peripheral clocks in phase as reported in higher vertebrates21. Actually, under these conditions, whole S. aurata larvae exhibit clear circadian rhythms in clock genes expression, being the expression of bmal1/clock and voluntary ingestion closely correlated22. In addition, most circadian clock studies on early stages of development of cultured fish have focused on the expression of a discrete number of clock genes22,23,24, and there is still little information on the overall metabolic output of the molecular clock. Thus, by analyzing clock components and transcriptomic variations in entire fish larvae, a better representation of the temporal organization of metabolism can be achieved. This approach has been used in early (5 days post-fertilization) zebrafish larvae2, but transcriptomic variations in more advanced and actively feeding fish larvae have not yet been examined, even though feeding exerts a prominent effect on fish metabolism (e.g., up to a 136% increase in metabolic rate)25.
The present study aims to depict the daily transcriptomic changes of the feeding larval stage of S. aurata under culture conditions and their relationship with the circadian clock. This is a perciform fish of high value among temperate farmed species of European aquaculture. Since fish larvae of S. aurata exhibit a remarkably high growth rate, which can reach 12–15% day−1 26, we hypothesized that the metabolism of actively feeding larvae of this species has a highly synchronized and time-based organization. To test this hypothesis, the S. aurata nucleotide database (www.nutrigroup-iats.org/seabreamdb)27 was updated with sequences from pyrosequencing of 454 libraries of larval origin28, and a specific high-density oligonucleotide microarray was constructed to examine, over a single day, the expression profile of more than 13,900 unique genes in whole larvae kept under a light/dark (LD) cycle with continuous feed availability. This gene expression analysis demonstrated the coordinated daily progression of various cellular and metabolic processes and highlighted the putative role of the circadian clock in the organization of daily metabolism and growth in whole fish larvae.
Results
Several genes are sequentially expressed during the daily cycle
To investigate whether the whole fish larvae transcriptome exhibits a daily pattern, 30-day-old S. aurata larvae were sampled every 3 h during a 24-h cycle under continuous feed availability and 12 h light: 12 h dark photoperiod. Then, a customized high-density oligo-microarray was used to profile the expression of 13,939 unique genes of S. aurata. One-way ANOVA showed that 2,229 genes were differentially expressed throughout the day (Supplementary Table S1). Principal Component Analysis (PCA) of differentially expressed genes showed a cyclic distribution of the groups along two components that accounted for 81% of the total variance (Fig. 1). Of note, minimal transcriptome differences (7 genes) were found when the comparison was made between fish sampled at 0 h and 24 h zeitgeber time.
Differentially expressed genes are grouped into four clusters
The k-means clustering of differentially expressed genes identified 4 major clusters with a sequential expression profile (Fig. 2a). The entire sets of genes included in each cluster are listed in Supplementary Table S1, with fold-change expression values referring to fish at 24 h zeitgeber time. Cluster 1 comprised the lowest number of genes (132), although the magnitude of response was higher than in the other clusters, with a peak of expression at 0 h zeitgeber time and the minimum 12 h later (Fig. 2b). Clusters termed 2, 3 and 4 contained 675, 758 and 650 genes, respectively; they also showed a circadian pattern of expression with intensity peaks at 3 h (cluster 2), 9–15 h (cluster 3) and 18–21 h (cluster 4) zeitgeber time (Fig. 2c).
Daily activation of canonical pathways of intermediary metabolism, development and cell cycle
Ingenuity Pathway Analysis (IPA) software was used to gain insight into the biological functions and pathways that were most significant in the different detected clusters. Up to 92.6% (2,052 genes) of differentially expressed genes were eligible for pathway analysis in the IPA software. Regarding molecular and cellular functions, the most significant for cluster 1 were “molecular transport”, “lipid metabolism”, “amino acid metabolism”, “nucleic acid metabolism”, “carbohydrate metabolism”, and “vitamin and mineral metabolism” (Fig. 3a). Notably, the most significant canonical pathways in this cluster were Thyroid hormone receptor and retinoid X receptor (TR/RXR) activation, represented by collagen alpha-3(VI) chain, cholesterol 7 alpha-monooxygenase (cyp7a1), phosphoenol pyruvate carboxykinase 1, and the mitochondrial uncoupling proteins (ucp1, ucp2, and ucp3). In addition, Phototransduction was significantly over-represented by five genes: arrestin 3, cyclic nucleotide-gated channel alpha 1 and 3, phosducin, and guanylatecyclase activator 1B. In cluster 2, the molecular functions “RNA post-transcriptional modification” and “DNA replication, recombination, and repair” were clearly the most significant (Fig. 3b), and overlapping analysis of associated canonical pathways identified a group of 30 related genes (Table 1), which were mainly involved in repair response to DNA damage and cell cycle regulation. “Cell cycle” and “DNA replication, recombination, and repair” were also among the top molecular and cellular functions of cluster 3, together with “cellular assembly and organization” (Fig. 4a). In this cluster, significant overlapping of canonical pathways identified 72 genes (Table 2), mostly related to unfolded protein responses, oxidative stress, and regulation of cell cycle. In contrast, there were no clearly prominent molecular and cellular functions in cluster 4 (Fig. 4b), and overlapping analysis only resulted in six related pathways with 26 overlapping genes (Table 3), including those of the cell cycle and signaling processes related to development (Neuregulin signaling, Agrin interactions at neuromuscular junction, and TWEAK-TNF-like weak inducer of apoptosis signaling).
Expression of several genes in different clusters is correlated with core clock genes
Seven genes in clusters 1–3 were unequivocally identified as core clock genes. These were the circadian protein homolog 3 (per3) in cluster 1; per1, per2 and cryptochromes 1 and 2 (cry1 and cry2) in cluster 2; and aryl hydrocarbon receptor nuclear translocator-like (bmal1) and circadian locomotor output cycles kaput (clock) in cluster 3. We next sought to establish whether the expression of genes within each cluster may be related to the expression pattern observed for these core clock genes. Interestingly, correlation analysis (Spearman coefficient > 0.95) showed that several genes in clusters 1, 2 and 3 shared the expression dynamics of the identified core clock genes within the corresponding cluster. Figure 5 shows the number of related genes for each clock gene and their average expression profiles. The entire list of significantly correlated genes to these clock genes and their normalized intensity values is provided as a supplemental material (Supplementary Table S2). Nine differentially expressed genes, including three clock genes, covering a wide range of hybridization intensities and fold-change variations were chosen for real-time qPCR analysis, and results were consistent (r = 0.87) with those of the microarray analysis (Supplementary Fig. S1).
Discussion
While the use of whole-body fish larvae in microarray studies was suggested to be a limiting factor because some tissue-specific effects may be buried by overall gene expression29, we found significant daily variations in the expression of both genes expressed in poorly represented tissues (e.g., pineal gland, retina) and those ubiquitously expressed. Additionally, the number of genes detected to have a significant variation through the day, after false positive correction, was relatively high (2,229) compared to those found by other microarrays to vary through ontogeny (~200)4 or in response to nutritional programming interventions (924–1,787)9. This may result from the remarkable variation in fish daily physiology, and from the update of the nutrigroup-IATS S. aurata nucleotide database that yields a powerful microarray tool with approximately 14,000 unique sequences, specifically enriched in actively expressed genes of the intestine and whole-larvae tissues. This represents a significant improvement over previous S. aurata arrays used to wide-underline gene expression in fish challenged with nutritional and environmental stressors30,31,32,33. Noteworthy, all genes present in this new microarray were annotated, and a single probe was designed for each gene. This might avoid potential deviation due to over-representation of multiple probes for a unique transcript and likely contributed to the reliable and clear analytical results, especially when statistical and functional analyses were envisaged. A first glimpse of the differential transcriptomic expression at each sampling point by PCA showed the sequential displacement of each sampling point along the two components that explained 81% of the whole variance, resulting in the striking figure of a 24-h circular cycle (Fig. 1). The shortest distances were observed between successive samples, while the highest distances were found between samples taken with a 12 h difference. This pattern clearly reveals that several genes are expressed sequentially along the day cycle in the entire larva.
Daily transcriptome organization in the growing larvae
Clustering analysis revealed the existence of 4 clusters of genes differentially expressed throughout the day (Fig. 2). Temporal and functional analysis of the first cluster, referred to as cluster 1, suggested that phototransduction genes are increasingly expressed during the dark period, likely as preparation for the next light phase. Concomitantly, we observed the up-regulation of genes involved in the TR/RXR activation pathway, in agreement with the role of the thyroid hormone (TH) cascade in light signal transduction34. TH and its receptors also affect a wide range of metabolic processes. In this regard, the expression of key genes related to lipid and carbohydrate metabolism were particularly enhanced during the night (Fig. 3a). This is the case of CYP7A1, which catalyzes the first reaction in cholesterol catabolism for bile acid synthesis35 and PCK1, which is a rate-controlling step of gluconeogenesis and hepatic glucose output36. TH action in vertebrates is also associated with changes in metabolic efficiency, modulating the up-regulation of mitochondrial uncoupling proteins (UCPs)37,38. This close association was evidenced for ucp1 and ucp2–3, which were all included in cluster 1 (Fig. 3a). Previous results in S. aurata revealed that ucp2 and 3 are up-regulated under feed restriction in aerobic muscle tissues or with aging or nutrient deficiencies in the glycolytic skeletal muscle39,40. This metabolic feature would reflect an increased flux of fatty acids towards skeletal muscle, which might also occur in larvae during overnight fasting and the first hours after the light onset. Hence, cluster 1 could be considered a metabolic regulator, considering its temporal pattern, high amplitude of expression, and the inclusion of TR/RXR activation pathway genes.
Cluster 2 also included genes with key roles in both direct light response (e.g., opn4) and transmission/amplification of the visual signal (e.g., pde) (Supplementary Table S1), which may act in concert with phototransduction genes from cluster 1. However, cluster 2 genes remained actively expressed 3 h after lights were turned on (Fig. 2c), suggesting that they may play a role in light entrainment. This is supported by the presence of the clock genes involved in light entrainment, per2 41 and cry1 42, and other light-responsive genes in retinal ganglions (opn4)43 and pineal gland (e.g., pinopsin)44 (Supplementary Table S1). Functional analysis of genes in cluster 2 outlined several interrelated canonical pathways involved in the transition through the G1/S stages of the cell cycle (Fig. 3b). The G1/S transition stage is a boundary between one active growth stage of cells (G1 phase) and DNA replication (S phase)45. Notably, key genes for G1/S transition such as cyclin D and cdk4 46, and DNA replication such as pcna 47, were observed in this cluster (Table 1). Response to DNA damage during S phase appears to derive from the up-regulation of interrelated molecular pathways such as those of BRCA1, GADD45, and p53 (Fig. 3b). Taken together, it is plausible that the moderate feed intake of S. aurata larvae during the morning22 (Supplementary Fig. S2) plays a role in determining the growth scope of the new day: cells check starting DNA quality and whether there are sufficient raw materials to replicate the DNA.
Genes in cluster 3 began to significantly increase their expression when the light was turned on, with maximal values 9–12 h later (Fig. 2c), coincident with the maximal feeding activity (Supplementary Fig. S2). The main canonical pathways in cluster 3 were related to cell cycle, including “G2/M DNA Damage Checkpoint Regulation” (Fig. 4a). During G2, cells continue to grow after DNA duplication in S, and during M phase mitosis occurs. ccna2 (cyclin A2), cdk1 (cyclin dependent kinase 1), and ccnab1 (cyclin B1) (Table 2) were represented in this cluster. Cyclin A modulates the activity of CDK1 during the transition from G2 to M48 and is replaced by cyclin B to regulate the progression of M49, which may occur about 6-hours after maximal mitotic gene expression50. Interestingly, some genes in this cluster are involved in myogenesis, such as mef2c, clock and myh 51,52,53 (Supplementary Table S1). Moreover, our data indicate that during most of the feeding period and probably during early night, S. aurata larvae achieve a low level of ROS by the Nrf2-Mediated Oxidative Stress Response pathway (Fig. 4a) and resultant up-regulation of antioxidant defense proteins (e.g., glutathione S-transferase, heme oxygenase, Table 2). This shift in antioxidant defense strategy (i.e., limiting ROS production during overnight fasting to ROS-scavenging during feeding) likely results from the high energy cost of growth during G2 phase and muscle differentiation. Protein synthesis, an energetically expensive process, is likely to be enhanced during the second half of the feeding phase, as genes involved in the folding of nascent proteins (e.g., hsp40, hsp70, and hsp90) (Supplementary Table S1) were at their maximum of expression.
Analysis of cluster 4 genes revealed that molecular processes related to cell morphology, proliferation, growth, and development, were up-regulated over nearly the entire dark period (Figs 2c and 4b). Significant canonical pathways in this cluster included Neuregulin Signaling and Agrin Interactions at the Neuromuscular Junction (Fig. 4b). The concomitant up-regulation of these pathways54,55 suggests that neuromuscular junction formation in S. aurata larvae occurs during the second half of the dark period and likely extends to the next morning. Cluster 4 also appears to prepare non-differentiated cells for upcoming processes the next morning, such as proliferation, as myogenic progenitor cells (MPCs) are strongly stimulated by IGF-II56 present in cluster 4. This is also supported by G1/S Checkpoint Regulation and Estrogen-mediated S-phase Entry in cluster 4. In addition, we found dio3 in this cluster (Supplementary Table S1), whose protein product might protect already differentiated tissues from T3-induced muscle differentiation57 or gate overall TR/RXR (cluster 1) effects to the next morning.
Cell cycle-regulated transcription in single cells is grouped into three main waves, coincident with the transition from G1 to S, G2 to M and M to G146, but studies on whole organisms, including fishes, are scarce58,59,60. A striking outcome of this study is that 24-h variations in whole S. aurata larvae transcriptome resemble cell cycle progression (Supplementary Fig. S2). This observation revealed an unexpected synchrony in the whole larvae, as it is established that somatic cell cycles are usually not synchronous after early embryonic stages58. This situation at the whole larvae may be equivalent to local cell cycle synchrony of vertebrate adult tissues under high cell proliferation conditions (e.g., regeneration)61. As it has been proposed that cells with an interdivision time near to 24 h proliferate faster62, the 24-h synchrony in the cell cycle at the whole organism level may result in the high growth rate of fish larvae26 and led us to suggest a cell cycle-based model for daily growth of S. aurata larvae as detailed in Supplementary Fig. S2. As it is well established that clock genes play key roles in the circadian organization of metabolism16, cell cycle63, and myogenesis64, we next sought to establish whether core clock genes may drive the observed expression patterns.
Putative role of clock genes in the daily organization of the larvae transcriptome
Virtually nothing is known about clock-driven transcription in the larvae of fishes other than zebrafish2,3. Our study provides evidence that the circadian clock plays a key role in daily transcriptome organization in S. aurata larvae. PER3 promotes stability and nuclear translocation of PER1/PER265 and thus, it was not surprising that per3 expression peak occurred (cluster 1) before per1/per2 peak (cluster 2) (Figs 2, 5). Expression of per3 (cluster 1) and per2 (cluster 2) may be related to the negative control of PPARγ transcriptional activity and adipogenesis66,67 (pparg in cluster 2, Supplementary Table S1) during the last hours of the dark phase and the first hour of the next morning, suggesting that clock genes from one cluster may be functionally related to genes from the same or the next cluster.
The expression of several genes in cluster 2 was significantly correlated with per1/per2 and cry1/cry2 during the dark period (Fig. 5). The activation of per and cry genes in larvae occurred after 0–3 h of maximal clock/bmal1 expression, in agreement with the activation of per and cry genes by CLOCK/BMAL113. PER and CRY may act over several cluster 2 genes carrying E-box sequences13, and these primary clock-controlled genes will in turn regulate second order clock-controlled genes in cluster 2 and 3. The per and cry expression later decreased through the light phase, likely as a result of their feedback regulation assisted by casein kinase I (CK1)68. We observed ck1 in cluster 4, attaining maximal expression three hours before the onset of light (Supplementary Table S1). Thus, a putative high availability of all CRY, PER, and CKI in the morning may facilitate their translocation to the nucleus and the interaction with CLOCK/BMAL1 heterodimers, thereby inhibiting transcription of cry and per genes.
The bmal1 and clock (cluster 3) up-regulation in S. aurata occurred during the light phase; bmal1 rhythmicity is known to be driven by changes in its promoter occupancy by RORα (activator) and REV-ERBα (repressor)69. Both rev-erba and bmal1 exhibited a similar expression pattern in S. aurata cluster 3, while rora was observed in cluster 2 (Fig. 5, Supplementary Table S1). These results indicate a finely tuned regulation of clock/bmal1 expression during the day; after transcription of rora during the night, RORα proteins may mediate the up-regulation of bmal1 during the day under negative regulation of REV-ERBα. This last transcription factor is also a key coordinator between circadian rhythms and metabolism, regulating metabolic pathways such as lipid and bile acid metabolism, adipogenesis, and gluconeogenesis70. Intriguingly, no core clock genes were identified in cluster 4, and thus this group of genes may be controlled indirectly by clock components of cluster 3 through their network output, especially transcription factors (6 and 13 in clusters 3 and 4, respectively). We are aware, however, that not all cycling genes described in our work must be directly or indirectly controlled by the clock but instead may be driven by other signals such as feeding in phase with the clock. We also acknowledge that functional interpretation of our correlation analysis may be hampered by the variable phase delay between the peak in protein accumulation and the peak in its respective transcript. Further experimental and computational approaches are needed to investigate this issue.
In general, main clock genes from clusters 1 and 2 (per1, 2, and 3, cry1 and 2) peaked at dawn and those from cluster 3 (bmal1, clock, and nfil3) at dusk. These temporal patterns are in line with those observed before in S. aurata brain20 and whole larvae22, and in zebrafish muscle71 and whole larvae3. However, zebrafish cry1 and cry2 were observed in antiphase3,71, while these genes were in phase in S. aurata. Additionally, in whole zebrafish larvae, Li et al. found that per2 rapidly increased its expression after light exposure (peak after 3 h)2, whereas we found that per2 was transcribed through all night in S. aurata to peak in the morning (also after 3 h of light), indicating differences among fish species in the responsiveness to light of per2. In addition, it is well known that chromatin remodeling is crucial for the clock function72. We observed histone acetyltransferases in cluster 2 (KAT2A and KAT7) and cluster 4 (KAT2B and KAT5) (Supplementary Table S1), which are transcriptional activators. Subsequently, genes coding for transcriptional repressors such as histone deacetylases (HDAC3) and methyltransferases (EHMT2, SETDB1-B) were found in cluster 3 (Supplementary Table S1). Hence, at least two waves of expression of chromatin remodeling enzymes may mediate clock-driven circadian transcription in S. aurata larvae.
Paving the way to indicators of circadian synchrony discovery
A plethora of studies during the last decade and recent research have resulted in an extended list of useful biomarkers for nutritional status, growth, metabolism and health in various farmed fish species (e.g., www.nutrigroup-iats.org/arraina-biomarkers). However, compared to juveniles, relatively few biomarkers are available for larval stages. We suggest that genes showing significant correlated synchrony with core clock genes may be suitable candidate biomarkers of circadian physiology, and those retained after filtration by pathways overlapping have the potential to be highly informative, as they may reflect the concerted action of several pathways leading to a particular phenotype. Taking this into account, we identified different candidate biomarkers of cell cycle progression, neuromuscular development and growth, and protection against oxidative stress (Supplementary Table S3) that offer many evaluation possibilities as briefly explained below.
It is well established that in animals, growth, health, and well-being depend on the synchronization of endogenous biological rhythms and environmental clues. Conventionally, circadian rhythm disturbance assessment requires repeated measurements of biomarkers through the 24 h cycle. Our study provides extensive information on daily variations of candidate genes (Supplementary Table S3) to follow this approach in S. aurata larvae. However, when several sampling points are not practical under routine aquaculture operations, we suggest to alternatively use a single sampling point to assess the expression of two subsets of selected genes from antiphase clusters (e.g., cluster 1 and 3), to find ratio indicators of synchrony. The use of these algorithms has the additional advantage of decreased variability of synchrony indicators among individuals or groups.
Also, we suggest that ratios resulting from the combined analysis of candidate biomarker genes at two different times may be informative about a given process scope. For example, it may be worthwhile to explore whether some combination of indicators of cell cycle progression (e.g., G1/S specific cyclin D1 at ZT3: G2/M specific cyclin B1, 2, or 3 at ZT12, Supplementary Table S3) are suitable predictors of larval growth scope. In addition, the occurrence of a higher number of MPCs resulting from proliferation during early development in teleosts may be advantageous for future growth due to increased fiber number (i.e., hyperplasia)53,73 as occurred in salmon74 and cod75. For this reason, we suggest the combined evaluation of pcna and other indicators of proliferation at ZT3 and myh-striated muscle and others indicators of differentiation at ZT 12 (Supplementary Table S3), as predictors of musculature development in S. aurata larvae. Likewise, the combined assessment of genes involved in antioxidant responses would lead to the selection of larvae batches with higher metabolic capacity. For instance, voluntary energy intake in fish was suggested to be limited by oxidative metabolism capacity, likely due to the detrimental effects of ROS76. Up-regulation of genes involved in the maintenance of the oxidative status was observed in Senegalese sole77 and cod78 larvae under high growth rate conditions. We suggest the combined evaluation of upc1/2/3 at ZT0 and gst, hmox, and several hsp at ZT12 (Supplementary Table S3) as putative indicators of the scope of oxidative metabolism of larvae.
In summary, this study demonstrates that the molecular circadian clock and metabolic rhythms are highly synchronized in early life stages, likely allowing S. aurata larvae to grow at a high rate. Our results also offer the possibility to identify early predictors of fish performance taking into account the changes in daily physiology. Given that S. aurata is a commercially relevant fish species for which genomic resources are increasingly available, this study opens new opportunities to unravel the complexity of daily gene regulation, with implications for fundamental and applied research.
Methods
Experimental setup
S. aurata larvae were reared at ICMAN-CSIC animal experimentation facilities (REGA number ES110280000311) in three circular 250-L tanks under constant temperature (19 °C) and salinity (34‰) and a 12 h light: 12 h darkness cycle. The light was switched on at zeitgeber time (ZT) 0 (09:00 h local time) and off at ZT 12 (21:00 h local time). The larvae were fed ad libitum with rotifers (Brachionus rotundiformis Bs-strain and B. plicatilis S-1-strain) supplied at a density of 10 rotifers/mL and enriched with the microalgae Nannocholopsis gaditana from day 4 post-hatching (dph) and subsequently with Artemia sp. nauplii from 18 dph until the end of the experiment79. Larvae were sampled at 30 dph (middle of the larval stage) during a 24 h cycle. Six individuals (two per tank) were taken every 3 hours (00:00, 03:00, 06:00, 09:00, 12:00, 15:00, 18:00, 21:00 and 24:00 h ZT) and preserved in RNAlater (Ambion). All experimental procedures complied with the Guidelines of the European Union Council (2010/63/EU) for the use and experimentation of laboratory animals and were reviewed and approved by the Spanish National Research Council (CSIC) bioethical committee.
RNA extraction for microarray analysis
Total RNA was extracted from whole larvae using an Ultra-Turrax T8 (IKA®-Werke) and the NucleoSpin® RNA II kit (Macherey-Nagel), including the on-column RNase-free DNase digestion included with the kit. RNA quantity was measured spectrophotometrically at 260 nm with a BioPhotometer Plus (Eppendorf) to a yield of 3 to 15 µg. RNA quality was checked in a Bioanalyzer 2100 and with the RNA 6000 Nano kit (Agilent Technologies). RIN (RNA integrity number) measurements ranged between 8.4 and 10, indicative of clean and intact RNA.
Transcriptome database construction and annotation
Blast comparisons were conducted between the assembled sequences of the Nutrigroup-IATS nucleotide database27 and those from pyrosequencing of 454 libraries of larval origin28 to combine both into a unique database. In the case of overlapping sequences with shared homology, the longest was retained. Larval sequences with no equivalent in the previous Nutrigroup-IATS database were annotated by searching sequence homologies against 24 different nucleotide and protein databases previously reported27, and subjecting them to the same algorithm of frame shift detection to correct potential 454 sequencing errors at homopolymer regions80. The updated S. aurata nucleotide database was hosted at www.nutrigroup-iats.org/seabreamdb, and it contains 3,388 larval annotated sequences (e-value < 1e-5) of a total of 20,565 non-redundant sequences encoding for 14,546 unique transcripts.
Microarray construction, hybridization and data analysis
The updated S. aurata database was the basis for a custom high-density oligo-microarray (8 × 15 K) (sea_bream_nutrigroup_array v.3), that was designed and printed using the eArray web tool (Agilent). The array comprised 60-oligomer probes for 13,939 different S. aurata genes and was used herein for circadian transcriptomic profiling of 30-day-old larvae. The design of the array was stored in the NCBI Gene Expression Omnibus (GEO) database under accession identifier GPL19579. Total RNA (150 ng) from individual fish (n = 6 for each group) were labeled with cyanine 3-CTP (Low Input Quick Amp Labelling Kit, Agilent), and 600 ng of each labeled cRNA were hybridized to microarray slides that were analyzed with an Agilent G2565C Microarray Scanner according to the manufacturer´s protocol. Data were extracted using the Agilent Feature Extraction Software 11.5.1.1 and deposited in the GEO database under accession identifier GSE64481. Microarray data analysis was performed with Genespring GX 13.0 software (Agilent). After quality control assessment, raw data (median intensity of each spot) were extracted and corrected for background with the Agilent Feature Extraction plug-in, and the intensity values were normalized using the 75th percentile shift. Functional pathway analysis was performed with IPA software (www.ingenuity.com). For each gene, the Uniprot accession of the annotation equivalent for one of the three higher vertebrates model species in IPA (human, rat or mouse) was assigned.
Real-time qPCR validation of microarray results
Nine differentially expressed genes covering a wide range of low and high hybridization intensity levels and fold-change variations were chosen for real-time qPCR analysis: cry1 (GenBank Accession Number JQ965014), clock (JQ965015), bmal1 (JQ965013), ucp3 (EU555336), pcna (KF857335), catalase (JQ308823), 26 S proteasome non-ATPase regulatory subunit 4 (psmd4; KM522789), cyp7a1 (KX122017) and surfeit locus protein (surf; KC217650). Validation was performed on the same individual samples used for microarray analyses (6 individuals for each of the nine sampling points) by real-time qPCR, using PerfeCTa™ SYBR® Green FastMix™ (Quanta BioSciences) in a Mastercycler ep gradient S Realplex2 (Eppendorf). Primer design, reverse transcription, qPCR optimization, and qPCR reactions were performed as previously detailed33. Specificity of reaction was verified by melting curve analyses and electrophoresis. Data were normalized to β-actin using the ΔΔCt method81 and fold-changes were referred to fish at 24 h zeitgeber time.
Statistical analysis
Microarray results from the nine experimental groups were analyzed by one-way ANOVA (corrected P-value < 0.05, Benjamini-Hochberg), PCA, k-means clustering and correlation analysis with similar entities by means of the Genespring GX 13.0 software (Agilent). PCA eigenvalues were determined by means of Genesis software (release 1.7.7). Optimal number of clusters was determined on the within-group sum of squares calculated by means of the k-means script in R. Ingenuity Pathway analysis used Fisher’s exact test was used to calculate a P-value reflecting the probability that the association between the set of molecules and a given pathway was due only to chance. Threshold of the P-value for association was set to 0.01. In overlapping pathway analysis, settings were selected to guarantee a minimum of 4 common genes between different canonical pathways.
Data availability
The updated S. aurata nucleotide database is available at www.nutrigroup-iats.org/seabreamdb. The design of the array was stored in the NCBI Gene Expression Omnibus (GEO) database under accession identifier GPL19579 and data obtained were deposited in the GEO database under accession identifier GSE64481. All other data generated or analyzed during this study are included in this published article and its Supplementary Information files.
References
Paranjpe, D. A. & Sharma, V. K. Evolution of temporal order in living organisms. J Circadian Rhythms 3, 7, https://doi.org/10.1186/1740-3391-3-7 (2005).
Li, Y., Li., G., Wang, H., Du, J. & Yan, J. Analysis of a gene regulatory cascade mediating circadian rhythm in zebrafish. PLoS Comput Biol 9, e1002940, https://doi.org/10.1371/journal.pcbi.1002940 (2013).
Boyle, G. et al. Comparative analysis of vertebrate diurnal/circadian transcriptomes. PLoS ONE 12, e0169923, https://doi.org/10.1371/journal.pone.0169923 (2017).
Sarropoulou, E., Kotoulas, G., Power, D. M. & Geisler, R. Gene expression profiling of gilthead sea bream during early development and detection of stress-related genes by the application of cDNA microarray technology. Physiol Genomics 23, 182–191, https://doi.org/10.1152/physiolgenomics.00139.2005 (2005).
Kaitetzidou, E., Xiang, J., Antonopoulou, E., Tsigenopoulos, C. S. & Sarropoulou, E. Dynamics of gene expression patterns during early development of the European seabass (Dicentrarchus labrax). Physiol Genomics 47, 158–169, https://doi.org/10.1152/physiolgenomics.00001.2015 (2015).
Christie, M. R., Marine, M. L., Fox, S. E., French, R. A. & Blouin, M. S. A single generation of domestication heritably alters the expression of hundreds of genes. Nat Commun 7, 10676, https://doi.org/10.1038/ncomms10676 (2016).
Olsvik, P. A., Lie, K. K., Nordtug, T. & Hansen, B. H. Is chemically dispersed oil more toxic to Atlantic cod (Gadus morhua) larvae than mechanically dispersed oil? A transcriptional evaluation. BMC Genomics 13, 702, https://doi.org/10.1186/1471-2164-13-702 (2012).
Marancik, D. et al. Whole-body transcriptome of selectively bred, resistant-, control-, and susceptible-line rainbow trout following experimental challenge with Flavobacterium psychrophilum. Front Genet 5, 453, https://doi.org/10.3389/fgene.2014.00453 (2015).
Balasubramanian, M. N. et al. Molecular pathways associated with the nutritional programming of plant-based diet acceptance in rainbow trout following an early feeding exposure. BMC Genomics 17, 449, https://doi.org/10.1186/s12864-016-2804-1 (2016).
Sarropoulou, E. et al. Transcriptomic changes in relation to early-life events in the gilthead sea bream (Sparus aurata). BMC Genomics 17, 506, https://doi.org/10.1186/s12864-016-2874-0 (2016).
Rebl, A. et al. Microarray-predicted marker genes and molecular pathways indicating crowding stress in rainbow trout (Oncorhynchus mykiss). Aquaculture 473, 355–365 (2017).
Murray, H. M. et al. Effect of early introduction of microencapsulated diet to larval Atlantic halibut, Hippoglossus hippoglossus L., assessed by microarray analysis. Mar Biotechnol 12, 214–229, https://doi.org/10.1007/s10126-009-9211-4 (2010).
Muñoz, E. & Baler, R. The circadian E-box: when perfect is not good enough. Chronobiol Int 20, 371–388 (2003).
Ye, R. et al. Dual modes of CLOCK:BMAL1 inhibition mediated by Cryptochrome and Period proteins in the mammalian circadian clock. Genes Dev 28, 1989–1998, https://doi.org/10.1101/gad.249417.114 (2014).
Busino, L. et al. SCFFbxl3 controls the oscillation of the circadian clock by directing the degradation of cryptochrome proteins. Science 316, 900–904 (2007).
Green, C. B., Takahashi, J. S. & Bass, J. The meter of metabolism. Cell 134, 728–742, https://doi.org/10.1016/j.cell.2008.08.022 (2008).
Mohawk, J. A., Green, C. B. & Takahashi, J. S. Central and peripheral circadian clocks in mammals. Annu Rev Neurosci 35, 445–462, https://doi.org/10.1146/annurev-neuro-060909-153128 (2012).
McStay, E., Migaud, H., Vera, L. M., Sánchez-Vázquez, F. J. & Davie, A. Comparative study of pineal clock gene and AANAT2 expression in relation to melatonin synthesis in Atlantic salmon (Salmo salar) and European seabass (Dicentrarchus labrax). Comp Biochem Physiol A Mol Integr Physiol 169, 77–89, https://doi.org/10.1016/j.cbpa.2013.12.011 (2014).
Betancor, M. B. et al. Daily rhythms in expression of genes of hepatic lipid metabolism in Atlantic salmon (Salmo salar L.). PLoSOne 9, e106739, https://doi.org/10.1371/journal.pone.0106739 (2014).
Vera, L. M. et al. Light and feeding entrainment of the molecular circadian clock in a marine teleost (Sparus aurata). Chronobiol Int 30, 649–661, https://doi.org/10.3109/07420528.2013.775143 (2013).
Hamada, T. et al. In vivo imaging of clock gene expression in multiple tissues of freely moving mice. Nat Commun 7, 11705, https://doi.org/10.1038/ncomms11705 (2016).
Mata-Sotres, J. A., Martínez-Rodríguez, G., Pérez-Sánchez, J., Sánchez-Vázquez, F. J. & Yúfera, M. Daily rhythms of clock gene expression and feeding behavior during the larval development in gilthead seabream. Sparus aurata. Chronobiol Int 32, 1061–1074, https://doi.org/10.3109/07420528.2015.1058271 (2015).
Davie, A., Sanchez, J. A., Vera, L. M., Sanchez-Vazquez, J. & Migaud, H. Ontogeny of the circadian system during embryogenesis in rainbow trout (Oncorhynchus mykyss) and the effect of prolonged exposure to continuous illumination on daily rhythms ofper1, clock, and aanat2 expression. Chronobiol Int 28, 177–186, https://doi.org/10.3109/07420528.2010.550407 (2011).
Martín-Robles, A. J., Whitmore, D., Pendón, C. & Muñoz-Cueto, J. A. Differential effects of transient constant light-dark conditions on daily rhythms of Period and Clock transcripts during Senegalese sole metamorphosis. Chronobiol Int 30, 699–710 (2013).
Secor, S. M. Specific dynamic action: a review of the postprandial metabolic response. J Comp Physiol B 179, 1–56, https://doi.org/10.1007/s00360-008-0283-7 (2009).
Yúfera, M., Pascual, E., Polo, A. & Sarasquete, M. C. Effect of starvation on the feeding ability of gilthead seabream (Sparus aurata L.) larvae at first feeding. J Exp Mar Biol Ecol 169, 259–272 (1993).
Calduch-Giner, J. A. et al. Deep sequencing for de novo construction of a marine fish (Sparus aurata) transcriptome database with a large coverage of protein-coding transcripts. BMC Genomics 14, 178, https://doi.org/10.1186/1471-2164-14-178 (2013).
Yúfera, M. et al. Transcriptomic characterization of the larval stage in gilthead seabream (Sparus aurata) by 454 Pyrosequencing. Mar Biotechnol 14, 423–435, https://doi.org/10.1007/s10126-011-9422-3b (2012).
Mazurais, D., Darias, M., Zambonino-Infante, J. L. & Cahu, C. L. Transcriptomics for understanding marine fish larval development. Can J Zool 89, 599–611, https://doi.org/10.1139/Z11-036 (2011).
Calduch-Giner, J. A. et al. Use of microarray technology to assess the time course of liver stress response after confinement exposure in gilthead sea bream (Sparus aurata L.). BMC Genomics 11, 193, https://doi.org/10.1186/1471-2164-11-193 (2010).
Calduch-Giner, J. A. et al. Dietary vegetable oils do not alter the intestine transcriptome of gilthead sea bream (Sparus aurata), but modulate the transcriptomic response to infection with. Enteromyxum leei. BMC Genomics 13, 470, https://doi.org/10.1186/1471-2164-13-470 (2012).
Calduch-Giner, J. A. et al. Transcriptional assessment by microarray analysis and large-scale meta-analysis of the metabolic capacity of cardiac and skeletal muscle tissues to cope with reduced nutrient availability in gilthead sea bream (Sparus aurata L.). Mar Biotechnol 16, 423–435, https://doi.org/10.1007/s10126-014-9562-3 (2014).
Martos-Sitcha, J. A. et al. Unraveling the tissue-specific gene signatures of gilthead sea bream (Sparus aurata L.) after hyper- and hypo-osmotic challenges. PLoS ONE 11, e0148113, https://doi.org/10.1371/journal.pone.0148113 (2016).
Bedolla, D. E. & Torre, V. A component of retinal light adaptation mediated by the thyroid hormone cascade. PLoS ONE 6, e26334, https://doi.org/10.1371/journal.pone.0026334 (2011).
Zhang, Y.-K. J., Guo, G. L. & Klaassen, C. D. Diurnal variations of mouse plasma and hepatic bile acid concentrations as well as expression of biosynthetic enzymes and transporters. PLoS ONE 6, e16683, https://doi.org/10.1371/journal.pone.0016683 (2011).
Gut, P. et al. Whole-organism screening for gluconeogenesis identifies activators of fasting metabolism. Nat Chem Biol 9, 97–104, https://doi.org/10.1038/nchembio.1136 (2013).
Samec, S., Seydoux, J., Russell, A. P., Montani, J. P. & Dulloo, A. G. Skeletal muscle heterogeneity in fasting-induced upregulation of genes encoding UCP2, UCP3, PPARgamma and key enzymes of lipid oxidation. Pflugers Arch 445, 80–86 (2002).
Vaitkus, J. A., Farrar, J. S. & Celi, F. S. Thyroid hormone mediated modulation of energy expenditure. Int J Mol Sci 16, 16158–16175, https://doi.org/10.3390/ijms160716158 (2015).
Bermejo-Nogales, A., Calduch-Giner, J. A. & Pérez-Sánchez, J. Gene expression survey of mitochondrial uncoupling proteins (UCP1/UCP3) in gilthead sea bream (Sparus aurata L.). J Comp Physiol B 180, 685–694, https://doi.org/10.1007/s00360-009-0441-6 (2010).
Bermejo-Nogales, A., Calduch-Giner, J. A. & Pérez-Sánchez, J. Tissue-specific gene expression and functional regulation of uncoupling protein 2 (UCP2) by hypoxia and nutrient availability in gilthead sea bream (Sparus aurata): implications on the physiological significance of UCP1-3 variants. Fish Physiol Biochem 40, 751–762, https://doi.org/10.1007/s10695-013-9882-7 (2014).
Ziv, L. & Gothilf, Y. Period2 expression pattern and its role in the development of the pineal circadian clock in zebrafish. Chronobiol Int 23, 101–112 (2006).
Tamai, T. K., Young, L. C. & Whitmore, D. Light signaling to the zebrafish circadian clock by Cryptochrome 1a. Proc Natl Acad Sci USA 104, 14712–14717, https://doi.org/10.1073/pnas.0704588104 (2007).
Semo, M. et al. Melanopsin retinal ganglion cells and the maintenance of circadian and pupillary responses to light in aged rodless/coneless (rd/rd cl) mice. Eur J Neurosci 17, 1793–1801 (2003).
Takanaka, Y., Okano, T., Iigo, M. & Fukada, Y. Light-dependent expression of pinopsin gene in chicken pineal gland. J Neurochem 70, 908–913 (1998).
Bartek, J. & Lukas, J. Pathways governing G1/S transition and their response to DNA damage. FEBS Lett 490, 117–122, https://doi.org/10.1016/S0014-5793(01)02114-7 (2001).
Bertoli, C., Skotheim, J. M. & de Bruin, R. A. M. Control of cell cycle transcription during G1 and S phases. Nat Rev Mol Cell Biol 14, 518–528, https://doi.org/10.1038/nrm3629 (2013).
Essers, J. et al. Nuclear dynamics of PCNA in DNA replication and repair. Mol Cell Biol 25, 9350–9359, https://doi.org/10.1128/MCB.25.21.9350-9359 (2005).
Pagano, M., Pepperkok, R., Verde, F., Ansorge, W. & Draetta, G. Cyclin A is required at two points in the human cell cycle. EMBO J 11, 961–971 (1992).
Pines, J. & Hunter, T. Cyclins A and B1 in the human cell cycle. Ciba Found Symp 170, 187–196 (1992). discussion 196–204.
Peyric, E., Moore, H. A. & Whitmore, D. Circadian clock regulation of the cell cycle in the zebrafish intestine. PLoS ONE 8, e73209, https://doi.org/10.1371/journal.pone.0073209 (2013).
Ornatsky, O. L., Andreucci, J. J. & McDermott, J. C. A dominant- negative form of transcription factor MEF2 inhibits myogenesis. J Biol Chem 272, 33271–33278 (1997).
Andrews, J. L. et al. CLOCK and BMAL1 regulate MyoD and are necessary for maintenance of skeletal muscle phenotype and function. Proc Natl Acad Sci USA 107, 19090–19095, https://doi.org/10.1073/pnas.1014523107 (2010).
Vélez, E. J. et al. Understanding fish muscle growth regulation to optimize aquaculture production. Aquaculture 467, 28–40 (2017).
Buonanno, A. The neuregulin signaling pathway and schizophrenia: from genes to synapses and neural circuits. Brain Res Bull 83, 122–131, https://doi.org/10.1016/j.brainresbull.2010.07.012 (2010).
Ruegg, M. A. & Bixby, J. L. Agrin orchestrates synaptic differentiation at the vertebrate neuromuscular junction. Trends Neurosci 21, 22–27 (1998).
Rius-Francino, M. et al. Differential effects on proliferation of GH and IGFs in sea bream (Sparus aurata) cultured myocytes. Gen Comp Endocrinol 172, 44–49 (2011).
Salvatore, D., Simonides, W. S., Dentice, M., Zavacki, A. M. & Larsen, P. R. Thyroid hormones and skeletal muscle-new insights and potential implications. Nat Rev Endocrinol 10, 206–214, https://doi.org/10.1038/nrendo.2013.238 (2014).
Sampetrean, O. et al. Reversible whole-organism cell cycle arrest in a living vertebrate. Cell Cycle 8, 620–627, https://doi.org/10.4161/cc.8.4.7785 (2009).
Dekens, M. P. S. et al. Light regulates the cell cycle in zebrafish. Curr Biol 13, 2051–2057 (2003).
Mendieta-Serrano, M. A., Schnabel, D., Lomelí, H. & Salas-Vidal, E. Cell proliferation patterns in early zebrafish development. Anat Rec 296, 759–773, https://doi.org/10.1002/ar.22692 (2013).
Matsuo, T. et al. Control mechanism of the circadian clock for timing of cell division in vivo. Science 302, 255–259, https://doi.org/10.1126/science.1086271 (2003).
Bernard, S. & Herzel, H. Why do cells cycle with a 24 hour period? Genome Inform 17, 72–79 (2006).
Bieler, J. et al. Robust synchronization of coupled circadian and cell cycle oscillators in single mammalian cells. Mol Syst Biol 10, 739 (2014).
Lazado, C. C. et al. Daily rhythmicity of clock gene transcripts in Atlantic cod fast skeletal muscle. PLoS ONE 9, e99172, https://doi.org/10.1371/journal.pone.0099172 (2014).
Yagita, K. et al. Dimerization and nuclear entry of mPER proteins in mammalian cells. Genes Dev 14, 1353–1363 (2000).
Costa, M. J. et al. Circadian rhythm gene Period 3 is an inhibitor of the adipocyte cell fate. J Biol Chem 286, 9063–9070 (2011).
Grimaldi, B. et al. PER2 controls lipid metabolism by direct regulation of PPARγ. Cell Metab 12, 509–20, https://doi.org/10.1016/j.cmet.2010.10.005 (2010).
Lee, H., Chen, R., Lee, Y., Yoo, S. & Lee, C. Essential roles of CKIγ and CKIɛ in the mammalian circadian clock. Proc Natl Acad Sci USA 106, 21359–21364, https://doi.org/10.1073/pnas.0906651106 (2009).
Guillaumond, F., Dardente, H., Giguère, V. & Cermakian, N. Differential control of Bmal1 circadian transcription by REV-ERB and ROR nuclear receptors. J Biol Rhythms 20, 391–403, https://doi.org/10.1177/0748730405277232 (2005).
Yin, L., Wu, N. & Lazar, M. A. Nuclear receptor Rev-erbα: a heme receptor that coordinates circadian rhythm and metabolism. Nucl Recept Signal 8, e001 (2010).
Amaral, I. P. G. & Johnston, I. A. Circadian expression of clock and putative clock-controlled genes in skeletal muscle of the zebrafish. Am J Physiol Regul Integr Comp Physiol 302, R193–206 (2012).
Feng, D. et al. A circadian rhythm orchestrated by histone deacetylase 3 controls hepatic lipid metabolism. Science 331, 1315–1319, https://doi.org/10.1126/science.1198125 (2011).
Rossi, G. & Messina, G. Comparative myogenesis in teleosts and mammals. Cell Mol Life Sci 71, 3081–3099, https://doi.org/10.1007/s00018-014-1604-5 (2014).
Levesque, H. M., Shears, M. A., Fletcher, G. L. & Moon, T. W. Myogenesis and muscle metabolism in juvenile Atlantic salmon (Salmo salar) made transgenic for growth hormone. J Exp Biol 211, 128–137, https://doi.org/10.1242/jeb.006890 (2008).
Katan, T. et al. A little goes a long way: improved growth in Atlantic cod (Gadus morhua) fed small amounts of wild zooplankton. Aquaculture 451, 271–282, https://doi.org/10.1016/j.aquaculture.2015.09.014 (2016).
Saravanan, S. et al. Constraints on energy intake in fish: the link between diet composition, energy metabolism, and energy intake in rainbow trout. PLoS ONE 7, e34743, https://doi.org/10.1371/journal.pone.0034743 (2012).
Fernández-Díaz, C., Kopecka, J., Cañavate, J. P., Sarasquete, C. & Solé, M. Variations on development and stress defences in Solea senegalensis larvae fed on live and microencapsulated diets. Aquaculture 251, 573–584, https://doi.org/10.1016/j.aquaculture.2005.06.014 (2006).
Rise, M. L. et al. Transcriptome profiling reveals that feeding wild zooplankton to larval Atlantic cod (Gadus morhua) influences suites of genes involved in oxidation-reduction, mitosis, and selenium homeostasis. BMC Genomics 16, 1016, https://doi.org/10.1186/s12864-015-2120-1 (2015).
Polo, A., Yúfera, M. & Pascual, E. Feeding and growth of gilthead seabream (Sparus aurata L.) larvae in relation to the size of the rotifer strain used as food. Aquaculture 103, 45–54 (1992).
Loman, N. J. et al. Performance comparison of benchtop high-throughput sequencing platforms. Nat Biotechnol 30, 434–439, https://doi.org/10.1038/nbt.2198 (2012).
Livak, K. J. & Schmittgen, T. D. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C (T)) method. Methods 25, 402–408 (2001).
Acknowledgements
This research was supported by the Spanish Ministry of Economic Affairs and Competitiveness (MINECO) by the project RIDIGEST (AGL2011-23722), by FEDER/ERDF contribution, by the project PROMETEO (FASE II-2014/085) from Generalitat Valenciana, and by the EU Seventh Framework Programme through ARRAINA (KBBE-2011-288925) project. This work does not necessarily reflect the views of the funders. The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. JAMS was supported by a doctoral grant (ID 215473) from the Mexican National Council for Science and Technology (CONACYT). The authors thank Ms. Rosa Vázquez (Servicio Central de Investigación de Cultivos Marinos, SCI-CM, CASEM-University of Cádiz, Spain) for supplying the S. aurata eggs.
Author information
Authors and Affiliations
Contributions
M.Y., J.P.S. designed the experiments. J.A.M.S., G.M.R. and M.Y. performed the experiments. J.C.G., G.M.R., J.P.S., J.A.M.S. and E.P. analyzed the data. E.P., M.Y., J.C.G. and J.P.S. wrote the manuscript. All authors reviewed the manuscript.
Corresponding author
Ethics declarations
Competing Interests
The authors declare that they have no competing interests.
Additional information
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Yúfera, M., Perera, E., Mata-Sotres, J.A. et al. The circadian transcriptome of marine fish (Sparus aurata) larvae reveals highly synchronized biological processes at the whole organism level. Sci Rep 7, 12943 (2017). https://doi.org/10.1038/s41598-017-13514-w
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-017-13514-w
- Springer Nature Limited
This article is cited by
-
Chronotypes-personality behavioural syndromes in wild marine fish
Scientific Reports (2023)
-
Transcriptomic profiling of Gh/Igf system reveals a prompted tissue-specific differentiation and novel hypoxia responsive genes in gilthead sea bream
Scientific Reports (2021)
-
A Transcriptome Insight During Early Fish Larval Development Followed by Starvation in Seriola rivoliana
Marine Biotechnology (2021)