The Expansion of Sirtuin Gene Family in Gilthead Sea Bream (Sparus aurata). Phylogenetic, Syntenic, and Functional Insights Across the Vertebrate/Fish Lineage

1. Introduction

Sirtuins (SIRTs) are a family of evolutionarily conserved enzymes that couple protein deacylation with the energy status of the cell via the cellular NAD⁺/NADH ratio [1]. These NAD⁺-dependent enzymes share different cellular locations (nucleus, cytoplasm, and mitochondria), and a wide tissue-specific gene expression pattern that reflects their involvement in a variety of fundamental biological processes [2,3,4]. Indeed, targets of SIRTs include regulators of DNA repair, inflammation, and energy metabolism [5]. In mammals, this protein family is composed of seven paralog counterparts (SIRT1-7), and their down-regulated expression has been associated with the pathophysiology of aging reviewed by [6]. Certainly, there is now evidence that SIRTs regulate multiple processes associated with the pathogenesis of Alzheimer’s disease and, thereby, therapeutic activation of SIRTs is emerging as an active area of research in the field of age-related disorders in humans [7]. However, a comprehensive evolutionary survey of the SIRT family remains unclear, though a recent study recognized the expansion of “Class I SIRT (SIRT1-3)” and “Class IV SIRT (SIRT6 and 7)” around the origin of animal branching, raising up to nine the number of SIRT paralogs [8]. Moreover, recent studies highlighted the occurrence of an additional SIRT isotype in the genome of non-amniote vertebrates (cartilaginous fish, bony fish, coelacanth, lungfish, and amphibians) that was annotated as Sirt3.2 by [9]. However, a universal taxonomy assignment remains difficult, being of relevance to untangle the evolutionary history and functional shift of the SIRT/Sirt gene family across the evolution of the vertebrate lineage.

In gilthead sea bream (Sparus aurata), a highly cultured fish in the Mediterranean basin, transcriptional studies revealed a ubiquitous sirt gene expression that was tissue-specific for each sirt isotype [10,11,12]. According to this, sirt1, but also sirt2 and sirt5 are expressed at relatively high levels. In contrast, sirt3, sirt4, sirt6 and sirt7 have been categorized as sirt isotypes with relatively low transcriptional activity. The exon-intron organization is also specific to each sirt paralog, ranging the number of exons from 3 in sirt4 to 16 in sirt2, though the gene organization (length and number of exons) seems to be relatively conserved for each SIRT ortholog through the vertebrate evolution. Conversely, the number and location of CpG islands (CGIs) are quite variable among fish genomes [13], and sirt genes in particular, despite its recognized role as an evolutionarily conserved mechanism that protects DNA from methylation, shaping the epigenome and the cell and tissue-specific transcriptional patterns through development [14]. Thus, CGIs in close association with SP1 binding sites were predicted in the sirt1 promoter of several fish, including gilthead sea bream (gene ID: 1847824780), fugu (gene ID: 101061405), zebrafish (gene ID: 797132), tilapia (gene ID: 100700447) and Atlantic salmon (gene ID: 106576833). However, the CGIs of the sirt3 promoter does not appear to be conserved in all teleostean fish, which will be indicative of different permissive transcriptional rates, though it is not always true that genes with CGI promoters shaped a ubiquitous gene expression [15].

The aim of this study was to go further in the function and regulation of Sirt family in fish with special focus in the farmed gilthead sea bream, an eurytherm, euryhaline and protandrous hermaphrodite species with a well-recognized capacity to adapt to intensive aquaculture practices and alternative fish feeds [16,17,18,19,20,21,22]. The ultimate mechanisms responsive to this high functional plasticity remain elusive, though it can be favoured by a high rate of short segmental duplications (SSD) in a fish species with a third round (3R) of whole genome duplications (WGD) [23]. Indeed, the genome of gilthead sea bream (1.24 Gb) spans more than the genome of turbot (Scophthalmus maxima) (0.54 Gb) [24] or European sea bass (Dicentrarchus labrax) (0.68 Gb) [25], but less than that of common carp (Cyprinus carpio) (1.7 Gb) [26] and Atlantic salmon (Salmo salar) (2.90 Gb) [27] with a recent tetraploidization event (4R) [28]. All this makes gilthead sea bream an interesting farmed fish for studying the adaptive expansion/contraction of some gene families to increase fitness and adaption to a changing environment in a context of climate change. This can be especially relevant in the case of the Sirt family, and we have used herein a combined approach based on structural genomics, phylogenetic and synteny analyses to disclose/re-annotate uncertain paralogs from the two already assembled gilthead sea bream genomes [23,29]. The study was completed by analysing developmental and tissue-specific gene expression patterns of sirt genes and other syntenic genes focusing on insulin-like growth factor binding proteins (igfbp1, igfbp3).

2. Results

2.1. Updated Sirt Catalogue

Data of PhylomeDB evidencing up to three new gilthead sea bream sirt gene isotypes that were embraced by SIRT3 and SIRT5 branches (Figures S1 and S2) (http://phylomedb.org; PHY00CLOW5_SPAAU and PHY00CKZTK_SPAAU accession numbers). These new sequences were uploaded to GenBank with accession numbers OR394775 (sirt3.1a), OR394776 (sirt3.1b) and OR394777 (sirt5b), being renamed their gilthead sea bream paralogs as sirt3.2 (AHX56275, before sirt3) and sirt5a (AHX56277, before sirt5) based on sequence, phylogeny and synteny analyses. Graphical representation of all gilthead sea bream Sirt proteins with data on amino acid sequence identity and similarity is shown in Figure 1. A conserved catalytic domain of about 250 amino acids in length was evidenced in all the reported sequences, and mitochondrial peptide targets of 17-66 amino acids in length were identified in Sirt3, Sirt4 and Sirt5 paralogs (Figure 1A). According to the current SIRT classification of [30], the highest sequence conservation was found between Sirt1-3 (Class I), as well as Sirt4 (Class II) and Sirt5 (Class III), and Sirt6 and Sirt7 (Class IV). This rendered amino acid sequence similarities of 53-99% and 74% for the comparisons made within the expanded gene families of Sirt3 and Sirt5, respectively (Figure 1B).

2.2. Gene Structure and Organization

For comparative purposes, the exon-intron organization of human and gilthead sea bream SIRT3/sirt3 and SIRT5/sirt5 genes is shown in Figure 2. As a general feature, the retrieved gilthead sea bream sirt3 genes (sirt3.1a, sirt3.1b, sirt3.2) were shorter than the human SIRT3 counterpart (4.3-5.1 kb vs. 19.6 kb), changing the number of protein-coding exons from 7 in the human SIRT3 to 8-9 in the duplicated gilthead sea bream sirt3.1 gene (Figure 2A). In contrast, the human and gilthead sea bream SIRT5/sirt5 genes showed the same number of protein-coding exons (8), but the gene length varied from 48.3 kb in sirt5a to 27.8 kb in SIRT5, and 5.1 kb in sirt5b (Figure 2B), conserving, more accurately than SIRT3/sirt3, the length of their coding region.

2.3. Phylogenetic and Syntenic Analyses

The reconstructed phylogeny of the SIRT3 family evidenced two main branches according each one with the present hierarchy of vertebrates (Figure 3). The branch named SIRT3/Sirt3.1 embraced the Sarcopterygians (amniotes, amphibian, coelacanth), and the Actinopterygians (ray-finned fish) that disclosed separately the node of primitive Actinopterygii fish (Polypteriformes/Acipenseriformes/Lepisosteiformes) and that of modern fish of the infra-class Teleostei represented in the branch Sirt3.1. This last taxonomic group also disclosed two main monophyletic groups corresponding to Salmoniformes/Cypriniformes and Perciformes/Gadiformes, showing most species of this fish lineage duplicated sirt3.1 genes as a result of whole or segmental tandem genome duplications. In contrast, the node named Sirt3.2 was a long branch that encompassed a single copy of sirt3 paralogs from cartilaginous fish (Chondrichthyes), lobe-finned fish (Coelacanthiformes) and aquatic tetrapods (frogs) to primitive and modern fish (Actinopterygii). The synteny analyses also rendered a different genome cartography for each one of these two SIRT3/Sirt3 branches. Thus, sirt3.1 has evolved in gilthead sea bream as a duplicated synteny block of 5 genes (psmd13/sirt3.1/drd4/cdhr5/ctsd) located in the superscaffold/chromosome 4 (Figure 4A). This segmental tandem genome duplication also occurred in other Percomorphaceae (Parambassis ranga, Sphaeramia orbicularis), and it persisted through the evolution of Sarcopterygians as a well conserved psmd13/sirt3.1/drd4 single-synteny block. Likewise, the branch of Sirt3.2 evolved across amphibians, cartilaginous fish, and the Actinopterygian fish class as a highly conserved single-synteny block (tmem263/sirt3.2/ric8b/rfx4/polr3b) that is located in the superscaffold/chromosome 8 of gilthead sea bream (Figure 4B).

The reconstructed phylogenetic tree of SIRT5 also rendered a long branch that embraced separately Sarcopterygians, primitive fish and teleosts, according to the present hierarchy of vertebrates with the exception of Chondrichthyes that were more related to Sarcopterygians rather than to bony fish (Figure 5). On the other hand, most primitive fish (Acipenseriformes/Lepisosteiformes/Polypteriformes) evolved as a main outgroup that constitutes together with Sarcopterygians the clade SIRT5, whereas the Teleostei disclosed two additional nodes (Sirt5a, Sirt5b) due to the conservation of two gene copies of sirt5 (superscaffold/chromosome 11 and 21 in gilthead sea bream) in almost all the members of this fish lineage. The analysis of synteny also supported this dichotomy trend with two highly conserved synteny blocks (foxc1/foxf2/foxq1/irf4 /dusp22; adcy1/igfbp1/igfbp3/tns3) in each one of these two teleostean Sirt5a/Sirt5b nodes (Figure 6). One or both of these gene synteny blocks were lost in the primitive Actinopterygii (Sterlet sturgeon) and Sarcopterygian fish (coelacanth), being weakened through the evolution of tetrapods (except in Gallus gallus and Xenopus tropicalis), which becomes especially evident for the conserved order of aligned genomic blocks of sirt5b (Figure 6B).

2.4. Tissue-Specific Gene Expression

The gene expression pattern of the complete catalogue of gilthead sea bream sirts was assessed in a wide range of tissues of adult fish with different metabolic and physiological features (Figure 7A). Specific PCR-primers were designed for each sirt paralog, with the exception of sirt3.1 primers that amplify both the sirt3.1a and sirt3.1b transcripts due to the difficulty in designing specific primers for each one of these duplicated genes (codifying sequence, 99% nucleotide identity, see Figure 1B). Overall sirt1, sirt2 and the now named sirt5a and sirt3.1 were the most ubiquitously expressed sirts across all the analysed tissues, whereas low expression levels were reported for sirt4 and sirt6 in almost all the analysed tissues. Relatively low expression levels were also disclosed for sirt7 in almost all tissues, except for liver and hypothalamus. The same for the novel sirt5b, which only shared relatively high levels of expression in liver and gills in comparison to the other sirt counterparts. Likewise, the now named sirt3.2 showed a relatively high expression level in head kidney (HK) and adipose tissue (AT), whereas the sirt3.1 paralogs (sirt3.1a + sirt3.1b) displayed high expression levels in liver and muscle tissues (white skeletal muscle, WSM; red muscle, RM and heart). At a closer look, the different expression pattern of duplicated genes of sirt3 and sirt5 paralogs is disclosed as a quotient expression ratio of sirt3.1/sirt3.2 and sirt5a/sirt5b (Figure 7B,C).

2.5. Differential Expression Patterns of Sirts and Syntenic Igfbps across Development

Gene expression across development with whole larvae, WSM and liver as targeted tissues was focused on sirt1, sirt2, sirt3, sirt5 and synteny igfbp genes of sirt5a (igbp1a, igbpb3a) and sirt5b (igfbp1b, igfbp3b). As shown in Table 1, the expression level of all the analysed genes was changed through the development. As a general feature, the expression level of sirt3.1 increased consistently with advancing age. In contrast, the expression level or sirt1, sirt3.2 and sirt5b decreased markedly. Less clear is the pattern in fish development of sirt2 and sirt5a, which achieved the maximum expression level in three-year old fish after an invariable or decreasing trend during early life stages. This also resulted in a pronounced increase of the expression ratio of sirt3 (sirt3.1/sirt3.2) and sirt5 (sirt5a/sirt5b) paralogs in advanced development. In parallel, the expression level of igfbps was markedly decreased (igfbp1a) or almost silenced (igfbp1b, igfbp3b) through the development, whereas a pronounced up-regulation was evidenced in the case of igfbp3a. In comparison to WSM, the developmental changes in the hepatic gene expression signature were less pronounced (Table 2). In any case, the hepatic expression of sirt1 and sirt3.2 was decreased significantly through development, which in turn increased the expression ratio of sirt3.1/sirt3.2. In parallel, the hepatic gene expression of igfbp1 paralogs (igfbp1a, igfbp1b) decreased with advancing age, but remained always detectable with a relatively high level of expression in the case of igfbp1b. In contrast, the gene expression of igbp3 paralogs (igfbp3a, igfbp3b) was almost silenced through the development in the liver tissue.

3. Discussion

SIRTs build a gene family of deacylases with a complex evolutionary history across eukaryotic organisms that makes difficult to establish a common SIRT repertoire, but at least seven paralogs can be traced back to the last common ancestor of vertebrates [9]. The present study confirmed this statement, increasing the number of Sirt family members from seven to ten in a Perciform fish, the gilthead sea bream (Figure 1A), which served to disclose up to three gene copies of sirt3 (sirt3.1a, sirt3.1b, sirt3.2) and two copies of sirt5 (sirt5a, sirt5b). All of them are nuclear-encoded mitochondrial proteins as evidences the conservation of a mitochondrial targeting sequence of 17-66 amino acids in length that would direct the newly synthetized proteins towards the mitochondria (Figure 1A). From an evolutionary perspective, phylogenomic approaches indicate that the mitochondrial proteome expanded through vertebrate evolution not only by gene/genome duplications of mitochondrial proteins, but also by re-localization of the paralogs to the organelle in a tissue-specific manner [31]. The most common process is the intra-mitochondrial duplication, but regardless of this, the establishment of new gene balances is always challenging, and the polyploids that survive a WGD event typically undergo a rediploidization process that leads to genome fractionation. Conversely, the preservation of duplicated genes can signify an increase in its activity (gene dosage), a novel acquired function (neo-functionalization), or the division among the copies of the functions of the ancestral gene (sub-functionalization) [32]. This is the basis of the evolution of the teleostean lineage that experienced a third WGD event (3R), followed by a 4R WGD in the branch belonging to the modern lineage of cyprinids and salmonids [28,33,34]. Conversely, alternative splicing arises in terrestrial vertebrates as the preferred mechanism to increase gene diversity [35,36]. Thus, recent human genomic searches disclosed up to 38 and 43 proteins isoforms of SIRT3 (https://www.ncbi.nlm.nih.gov/gene/23410; last accessed: 17/01/2024) and SIRT5 (https://www.ncbi.nlm.nih.gov/gene/23408; last accessed: 17/01/2024), respectively. These alternative variants often show a tissue-specific gene expression pattern, extending the way in which mitochondrial, cellular and organism homeostasis can be regulated in an effective manner [37,38]. Certainly, meta-analyses approaches have identified the mitochondria as a key player of stress adaptive responses in a wide range of living organisms, including fish [39]. With these considerations in mind, we aimed to better understand the evolution and functional significance of the expanding Sirt3 and Sirt5 family in a farmed fish model, with a well-recognized predisposition of gene expansion [23,40].

At a closer look, phylogenetic attempts disclosed two SIRT3 ortholog (2R WGD) at the early vertebrate evolution [8] with still vestiges in the genome of the modern Sarcopterygians (Figure 3) that supports the Sirt3 division in Sirt3.1/Sirt3.2 in non-amniote aquatic-living vertebrates [9]. The Sirt3.1 branch also discerned sirt3.1a and sirt3.1b duplicates in gilthead sea bream and in two other Percomorphaceae (S. orbicularis, P. ranga), which might mean a gene retention after the 3R WGD (Figure 3). However, these paralogs accumulated a very low intra-sequence polymorphisms (Figure 1B), which can lead to theorise that the sirt3.1 duplication had occurred recently in the evolution [41]. This is further supported by the conservation in these fish species of a duplicated tandem repeat of 5 genes (psmd13-sirt3.1a/3.1b-drd4-cdhr5-ctsd) in each one of these two sirt3.1 paralogs (Figure 4A), which differ largely from the synteny block (tmem263/sirt3.2/ric8b/rfx4/polr3b) of the sirt3.2 paralog retained through the evolution as a unique gene copy (Figure 4B).

The importance of SIRT3 in metabolic homeostasis is widely documented in the skeletal muscle of humans and rodents, in which the TRF2-SIRT3 axis connects telomere shortening with muscle adaptive metabolism, development and aging [42,43]. Thus, not surprisingly, TRF2 (telomeric repeat-binding factor 2) ablation in mouse skeletal muscle leads to Sirt3 down-regulation, highlighting the functional importance of the TRF2-mediated chromatin loop in regulating Sirt3 gene expression, and subsequently oxidative metabolism and cellular senescence. Mechanistically most of these processes remain elusive, though it is known that the anti-aging effects of the adjudin drug are exerted by elevating the expression level of SIRT3, which in turn reduces cellular levels of reactive oxygen species (ROS) by deacetylating forkhead box O3a (FOXO3a), a transcription factor that transactivates antioxidant genes, such as catalase (CAT) and manganese superoxide dismutase (SOD2) [44,45]. In agreement with this, in gilthead sea bream, the expression level of sirt3.1 reached its highest expression level in RM and WSM (Figure 7B), with an enhanced expression through development in the case of WSM (Table 1 and Table 2) that becomes largely constitutive in the liver (Table 2). Otherwise, it must be noted that the sirt3.1 gene lies very close to pmsd13 (Proteasome 26S subunit, non-ATPase 13) through almost all the vertebrate evolution (Figure 4A), with the occurrence in humans of a bidirectional promoter that reinforces the link of metabolic condition, and mitochondrial and cellular stress responses with the proteolysis of unfolded proteins [46,47]. If this is the case of other organisms, such observation gives support to a subrogate marker of deviations of biological age in a wide range of living organisms, including farmed fish.

Unlike sirt3.1, both in this (Figure 7B) and a previous study [10], the regulation of sirt3.2 was driven towards an active gene expression in mucosal and immunorelevant tissues of gilthead sea bream (HK, spleen, gills, intestine), which was concurrent herein with a developmentally regulated gene silencing in both liver and WSM (Table 1 and Table 2), also reported for the ortholog of Xenopus tropicalis (https://www.bgee.org/gene/ENSXETG00000014800; last accessed: 17/01/2024; [48] (Figure 4B). Since SIRTs are considered to have anti-inflammatory properties due to their regulatory effects on several transcription factors and their downstream pro-inflammatory effectors [49], such gene expression pattern can be viewed as a negative feedback regulation of inflammation. Certainly, SIRTs are highly activated during a number of conditions known to enhance NAD⁺ bioavailability including nutrient restriction (fasting), exercise, and late acute inflammation. On the contrary, aging, nutrient overload and activation of the hypoxia-ROS-early inflammation triad lead to decreased cellular NAD⁺ levels, which translates into reduced gene expression and enzymatic activity of SIRTs. This was supported by recent observations in gilthead sea bream, where the ectoparasite Sparicotyle chrysophrii, parasitizing the gill epithelium, triggers the up-regulation of apoptotic markers in combination with a down-regulation of sirt3.2 and hypoxia-related genes [50]. Likewise, also in gilthead sea bream, there is now experimental evidence that high stocking density and mild-hypoxia pre-conditioning down-regulated sirt3.2 expression in heart [51] and WSM [18], respectively. Altogether, in the context of climate change with an increase in water temperature and reduced oxygen availability strengthens the interest in targeting sirt3 on livestock-fish farming.

As reported for SIRT3/Sirt3, the SIRT5/Sirt5 branch also expanded through the evolution of vertebrates, rendering two Sirt5 isotypes that remained conserved in the lineage of modern teleost as two separated Sirt5a and Sirt5b clusters (Figure 5). The Sirt5a branch was phylogenetically closer to its human ortholog, forming a conjoint clade composed of teleost and non-teleost organisms. By contrast, the Sirt5b branch was exclusive of bony fish and persisted as a separate node with no vestiges of it in Sarcopterygians. This, together with a relatively low phylogenetic divergence of Sirt5a/Sirt5b (74% amino acid similarity) in comparison to Sirt3.1/Sirt3.2 (53% amino acid similarity) supported the acquisition of a novel Sirt5b isotype as a teleost-specific Sirt after the 3R WGD event of teleosts, rather than a duplication event at the early vertebrate evolution (2R WGD). In addition, the sirt5a was expressed more actively than sirt5b across a large range of tissues, with the exception of gills where the expression of sirt5b was at least two-fold higher than that of sirt5a (Figure 7). If this is indicative of a tissue-specific activity related to aquatic life remains elusive, though it has been suggested that the Sirt5 counterpart of marine mussels would drive different cellular stress responses to alleviate signs of heat stress [52]. In particular, in gilthead sea bream, sirt5a becomes expressed actively in both anaerobic (WSM) and aerobic (RM, heart) muscle tissues. Furthermore, the expression level of sirt5a was maintained high in WSM through development, whereas it was almost suppressed in the case of sirt5b few days after hatching. The close association of sirt5a and muscle growth and metabolism was also reported in previous gilthead sea bream studies, in which growth suppression with fasting [10] and advancing age [15] was associated to a pronounced increase of sirt5a gene expression. In both cases, this occurred in coincidence with a reduced energy wastage, evidenced by the down-regulation of muscle mitochondrial uncoupling proteins and changes in the gene expression patterns of markers of lipid metabolism and oxidative phosphorylation, according to the general idea that SIRTs restrain any energy-consuming cellular activity, including growth and inflammation, until the abatement of metabolic stress disturbances [49,53].

Early studies also support a crosstalk between SIRTs and the GH/IGF axis, a key endocrine system regulating growth in vertebrates. Attention has been focused on SIRT1, and studies in mice have evidenced that in vivo knockdown of hepatic SIRT1 restores the fasting-induced decrease in circulating levels of IGF-I [54]. Additionally, SIRT1 acts at the brain level as a link between somatotropic signalling and calorie restriction [55], and brain SIRT1 knockouts displayed dwarfism and reduced plasma GH and IGF-I levels [56]. According to this, SIRT1 activation contributes to suppress the GH/IGF tonus, which would serve to drive a decreased supply of energy for growth purposes in a cellular milieu with a reduced availability of metabolic fuels [53]. The relationship of GH/IGF axis with other SIRT paralogs remains mostly understood, but intriguingly our synteny analysis highlighted a high conservation through all the vertebrate evolution of a genomic synteny block with SIRT5 and IGBP1/3 as neighbouring genes (Figure 6). This offers the possibility of the maintenance of genomic regulatory blocks, where the regulatory domain of a regulatory gene can extend into and beyond adjacent transcriptional units to shape the co-regulated expression of neighbouring genes [57,58]. Focusing on IGFBPs, it is now recognized that the ancestral IGFPP gene was duplicated at an early animal stage to produce a pair of IGBPs that gave rise in subsequent duplication events the two IGFBPs clades of modern vertebrates (IGFBP1/2/4; IGBP3/5/6) with a differentially regulated expression in liver and skeletal muscle that is early accomplished through development in gilthead sea bream [18,53]. In adult fish, such regulatory feature rendered a clear dominance of the igbp3a expression in the WSM of adult fish, which might depict a co-regulated expression of sirt5a and igbp3a as a synteny block in this muscle tissue (Table 1). Conversely, the silencing of sirt5b should be shaped by a genomic regulatory block involving the co-regulated inhibition of igfbp1b. However, less clear are these types of associations at the liver level, in which other regulatory mechanisms might serve to preserve a more constitutive gene expression of all sirt3 and sirt5 paralog pairs (Table 1 and Table 2).

In summary, this research highlighted the expansion of Sirt3 and Sirt5 family in gilthead sea bream at different times through the evolution, increasing the repertoire of Sirt genes from seven to nine in our farmed fish model. As their mammalian SIRT counterparts, the named sirt3.1a/b and sirt5a were mostly expressed at a relatively high level in muscle tissues, with perhaps a different contribution to proteolysis of unfolded proteins and muscle growth regulation via Gh/Igf system as inferred from gene synteny block analysis. By contrast, the named Sirt3.2 evolved as an exclusive isotype of aquatic organisms with a main envisage immunoregulatory role based on gilthead sea bream tissue-specific gene expression patterns. Likewise, Sirt5b emerged as an exclusive teleostean Sirt isotype with an enhanced expression in gills that might support some specific adaptive features like osmoregulation and heat stress. Overall, these findings offer valuable insights into the adaptive strategies of fish across development and in response to environmental challenges, reinforcing the importance of gene duplication shaping the landscape of the plastic gilthead sea bream

4. Materials and Methods

4.1. Ethics Statement

All procedures involving experimental animals were approved by the Ethics and Animal Welfare Committee of IATS, CSIC Ethics Comittee (permission 1295/2022) and Generalitat Valenciana (permission 2022-VSC-PEA-0230). Such interventions were carried out in a registered installation facility (code ES120330001055) in accordance with the principles published in the European Animal Directive (2010/63/EU) and Spanish laws (Royal Decree RD53/2013) for the protection of animals used in scientific experimentation.

4.2. Fish Husbandry and Tissue Sampling

Fish were raised from early life stages to harvest with commercial pellets (0.2-0.8 mm Skretting; 1.5-6.5 mm Biomar) at the indoor experimental facilities of the Institute of Aquaculture Torre de la Sal (IATS, CSIC) under natural photoperiod and temperature conditions at our latitude (40° 5 N; 0° 10 E). Feed was offered daily near to visual satiety or in alternate days (3-7 times per week) depending on season and fish size. Water temperature ranged from 10°C in winter to 27°C in summer. The water oxygen concentration was always higher than 75% saturation, and unionized ammonia remained below toxic levels (< 0.02 mg/L) irrespective of season. For the tissue-specific gene expression profiling, one-year-old fish (120 g average body weight) were overnight fasted in summer and representative tissues (liver, white skeletal muscle, red muscle, heart, gills, perivisceral adipose tissue, anterior intestine, posterior intestine, head kidney, spleen and hypothalamus) with different metabolic capabilities were rapidly excised, frozen in liquid nitrogen, and stored at -80°C until RNA extraction. Additionally, samples of whole body (larval stages), liver and white skeletal muscle were taken through fish development (May-July) at 60-, 81- and 127-days post-hatching (dph) from individuals with an averaged body weight of 137 mg, 1 g and 4 g, respectively. At the latest sampling point, liver and muscle tissues were also taken from 100 g (one-year old fish) and 1 kg (three-year old fish) fish.

4.3. Phylogenetic and Synteny Analyses

The reconstructed gilthead sea bream PhylomeDB with protein-coding sequences from the IATS-CSIC assembled genome [23] was interrogated for sirt paralogs. This rendered up to three new sequences that were recognized as sirt3 and sirt5 gene isotypes that were manually curated by homology searches in the genome draft of [29]. This was followed by extensive searches for orthologs of Sirt3/Sirt5 genes across the vertebrate lineage in the OrthoDB v11 database (Kuznetsov et al., 2023). Multiple sequence alignments and similarity/identity calculations were carried out using GeneDoc v2.7 (http://genedoc.software.informer.com/). Mitochondrial target sites and sirt domains were predicted using the TargetP- 2.0 (https://services.healthtech.dtu.dk/services/TargetP-2.0/) and PROSITE-Expasy (https://prosite.expasy.org/) online tools, respectively. Separate phylogenetic trees for SIRT3/Sirt3 and SIRT5/Sirt5 clades were constructed with the Maximum- Likelihood algorithm in MEGA v11 [59], using JTT [60] substitution matrix based on the lowest BIC (Bayesian Information Criterion) score. Non-uniformity of evolutionary rates among sites were modeled by using a discrete Gamma distribution (+G). All positions containing gaps and missing data were eliminated (complete deletion). For syntenic studies, neighboring genomic regions of SIRT3/sirt3 and SIRT5/sirt5 paralogs were established by querying the IATS-CSIC genomic database and the Genomicus tool v100.01 [61]. The syntenic blocks were the result of gene order arrangements from representative species of Sarcopterygians (order Coelacanthiformes and Superclass Tetrapoda) and Actinopterygians (Chondrostei and Neopterygii Subclasses). The lineage-representative group of species used in each approach, as well as their corresponding NCBI protein ID, can be found in Table S1. The graphical representation of gene organization was carried out with the online tool Exon-Intron Graphic Maker (https://wormweb.org/exonintron). The manuscript followed the ZFIN Zebrafish Nomenclature Guidelines for gene and protein names and symbols, with a/b assigned suffixes based on the suffix of the already annotated surrounding genes (https://wiki.zfin.org/display/general/ZFIN+Zebrafish+Nomenclature+Conventions).

4.4. Gene Expression Profiling

RNA was extracted using the MagMAX-96 total RNA isolation kit (Life Technologies, Carlsbad, CA, USA). The RNA yield was 50-100 μg, with absorbance ratios (A260/A280) of 1.9-2.1. RNA integrity number (RIN) values of 8-10 (Agilent 2100 Bioanalyzer; Agilent, Santa Clara, CA, USA) were indicative of clean and intact RNA. Reverse transcription of 500 ng of total RNA was performed with random decamers using a High-Capacity cDNA Archive Kit (Applied Biosystems, Foster City, CA, USA). Negative control reactions were run without reverse transcriptase. Different 96-well PCR arrays were designed for the simultaneous gene expression profiling of gilthead sea bream sirts and syntenic igfbp genes (igfbp1a, igfbp1b, igfbp3a, igfbp3b). Two housekeeping genes (β-actin and 18S rRNA) and controls of PCR performance were included in each array. Briefly, 660 pg of total cDNA was used in 25 μL PCR reactions. PCR wells contained 2x SYBR Green Master Mix (Bio-Rad, Hercules, CA, USA) and specific primers at a final concentration of 0.9 μM (Table S2). All pipetting operations for the PCR arrays were performed by an EpMotion 5070 Liquid Handling Robot (Eppendorf, Hamburg, Germany) to improve data reproducibility. Real-time quantitative PCR was carried out in an Eppendorf Mastercycler Ep Realplex (Eppendorf, Hamburg, Germany). The PCR amplification program consisted of an initial denaturation step at 95°C for 3 min, followed by 40 cycles of denaturation for 15 s at 95°C and annealing/extension for 60 s at 60°C. The efficiency of the PCR reactions was consistently higher than 90% and similar among all the genes. The specificity of the reactions was verified by melting curve analysis (ramping rates of 0.5°C/10 s over a temperature range of 55-95°C). Negative controls without a template were performed for each primer set. Gene expression was calculated using the delta-delta Ct method [62]. For multigene analysis, all values for whole larvae and muscle tissue were referenced to the expression level of sirt1 in whole larvae (60dph). Liver gene expression levels were referenced to the expression of sirt1 at 81dph, for which a value of 1 was arbitrarily assigned.

4.5. Updated Sirt Catalogue

Hierarchical clustering was made using the Genesis software [63] to assess the tissue-specific gene expression pattern of the complete catalogue of sirt genes in gilthead sea bream. Statistical analyses were performed using SigmaPlot version 14.0 (Systat Software, San Jose, CA) with all P-values set to 0.05. Normality and equal variance of data were tested by Shapiro-Wilk and Levene tests, respectively. Developmental and tissue-specific differences in gene expression were analyzed by One-way ANOVA followed by Holm-Sidak test.