http://orcid.org/0000-0002-5036-2558
Figures
Abstract
We have created a new, fast-growing strain of common carp with higher unsaturated fatty acid content in muscle. To better understand the impacts of gene regulation in intestinal tissue on growth and unsaturated fatty acid content, we conducted a comparative RNA-Seq transcriptome analysis between intestine samples of Selected and Control groups (and corroborated selected results by PCR). After eight weeks of cage culture, weight gain of the Selected group was 20.84% higher. In muscles of the control group, monounsaturated fatty acids (FAs) were more abundant, whereas polyunsaturated FAs were more abundant in muscles of the Selected group. In total, we found 106 differentially expressed genes (DEGs) between the two groups. Only the endocytosis pathway was significantly enriched in DEGs, with two upregulated genes: il2rb and ehd1. The latter is involved in the growth hormone/insulin-like growth factor (Gh/Igf) axis, which plays a key role in the regulation of growth in animals. tll2, which is known to be associated with intestinal regeneration, was extremely highly upregulated in both transcriptomic (infinite) and qPCR (610.70) analyses. Two of the upregulated genes are associated with the fatty acid metabolism, several genes are likely to be indicators of heightened transcription levels, several are associated with metabolic and developmental roles, several with neuronal functions (including two with vision), several with the immune system, and two downregulated genes with the development of vasculature. The higher growth rate of the Selected group is likely to be at least partially attributed to increased endocytosis efficiency and genetically-driven behavioural differences (higher aggression levels). There are some indications that this new strain might have slightly impaired immune responses, and a higher propensity for inherited diseases leading to sight impairment, as well for neurodegenerative diseases in general, but these indications still need to be confirmed.
Citation: Zhang C, Su S, Li X, Li B, Yang B, Zhu J, et al. (2018) Comparative transcriptomics identifies genes differentially expressed in the intestine of a new fast-growing strain of common carp with higher unsaturated fatty acid content in muscle. PLoS ONE 13(11): e0206615. https://doi.org/10.1371/journal.pone.0206615
Editor: Marc Muller, Universite de Liege, BELGIUM
Received: April 13, 2018; Accepted: October 16, 2018; Published: November 5, 2018
Copyright: © 2018 Zhang et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Data are deposited in the NCBI’s BioProject (PRJNA414702) and SRA (SRR6202422, SRR6202423, SRR6202418, SRR6202419, SRR6202424, SRR6202425, SRR6202421, SRR6202426, SRR6202420, SRR6202429, SRR6202427, SRR6202428) databases.
Funding: This work was supported by grants from Central Public-Interest Scientific Institution Basal Research Fund, Chinese Academy of Fishery Sciences (CAFS) (2016RC-LX03 to SS); the Key Projects in the National Science & Technology Pillar Program during the twelfth Five-Year Plan Period (2012BAD26B02); and the National Nonprofit Institute Research Grant of CATAS-TCGRI (2013JBFM14). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Growth rate, regulated by environmental factors and genetics, is a primary trait of interest in selection programs of most cultured fish species due to its intrinsic link with productivity and profitability of aquaculture enterprises [1]. The common carp (Cyprinus carpio Linnaeus 1758) is probably the oldest and most common cultured fish species, with third largest total aquaculture production output globally. In 2014, the total global output was over four million tonnes (almost 10% of the global annual freshwater aquaculture production), producing the economic value of almost six billion dollars, with China producing around 70% of the total output [2].
Growth of skeletal muscles of fish is the primary topic of interest for the aquaculture industry [3]. Several growth-related features set the fishes apart from mammals: continuous accretion of muscle tissue results in the increase of body length and mass (albeit at a slowing rate) until mortality or senescence occur, both hyperplasia and hypertrophy contribute to muscle growth, and fishes predominantly accumulate functional protein (as opposed to storing the excessive energy as adipose tissue) [3–5]. Although some molecular aspects of growth in fish are relatively well-studied, such as the growth hormone—insulin-like growth factor I (GH-IGFI) axis of the neuroendocrine system [3,6–8], our understanding of the molecular regulation of growth in fish remains fragmentary [3,9].
Fatty acids are essential cellular components, whose composition is important for the nutritive value and taste of meat [10]. In humans, saturated FAs are risk factors for cardiovascular diseases, whereas unsaturated and especially polyunsaturated FAs, such as omega-3 and -6 (ω-3 and ω-6), have antiatherosclerotic effects and a number of other beneficial health effects [11,12]. Fish meat is generally the main source of polyunsaturated fatty acids (FAs) in human diet, and thus of remarkable significance in human nutrition [13,14].
The gastrointestinal tract acts as a selectively permeable barrier for dietary nutrients, electrolytes and water, while maintaining an effective defense against pathogens [15]. Regardless of anatomical and physiological differences (common carp is an agastric, omnivorous fish), teleost fishes are characterised by the presence of at least two intestinal segments: the first is responsible for the absorption of lipids and the second for pinocytotic uptake of macromolecules, including proteins [3,4]. Uptake of nutrients from food via the intestinal barrier is a prerequisite for growth, and impacts of different diets on intestinal transcriptome have been studied in fish [16,17]. However, individual differences in the efficiency of this process in fish, and particularly the genetic background thereof, remain almost completely unstudied.
The molecular control of growth in fish muscles is relatively well-understood [5,9], but molecular control of intestinal nutrient uptake, although likely to play a part in the growth rate [9], remains poorly understood. High-throughput RNA sequencing technology (RNA-Seq) has been used to study the molecular mechanisms underlying growth in bacteria [18], plants [19], mammals [20], and fish [21–26]. Furthermore, detailed RNA-Seq transcriptome analyses have been applied to study gene expression profiles in 19 different tissues of the common carp, but the analysis did not include intestinal samples [27]. Intestinal transcriptomes have been used to search for drug candidates [28], study regeneration mechanism [29], and genome polyadenylation [30], but (to our knowledge) none of previous transcriptome-based studies of fish growth have used intestine among the sampled tissues [21,22].
A new, fast-growing strain of common carp with higher unsaturated fatty acid content in the muscle was recently created by combining the best linear unbiased prediction and molecular markers-based breeding methods [31]. To better understand the overall gene expression levels in the intestine of common carp, as well as the impacts of gene regulation in intestinal tissue on the growth and unsaturated fatty acid content in the muscles of this important cultured fish species, we conducted a comparative RNA-Seq transcriptome analysis of intestinal tissues of the new strain and a control group. This study also aims to contribute to the understanding of the genetic basis of growth and unsaturated fatty acid content in fish and vertebrates in general.
Results and discussion
Growth performance and effects of selection on fatty acid composition of muscles
Growth performance test of 48 families showed that after eight weeks of cage culture experiment, weight gain of the Selected group was 20.84% higher than that of the Control group (Table 1). Fatty acid composition analysis showed that saturated FAs (SFA) were more abundant (albeit non-significantly) in the muscles of Control group (Table 1). Among eight SFAs, six were more abundant in the Control group, but only one significantly. Regarding the two that were higher in the Selected group (C:18 and C:20), only C:18 was significantly higher. Unsaturated FAs (UFA), however, were more abundant in the Selected group: among 15 UFAs, only four were more abundant in the Control group, and only C18:1 significantly. Among the remaining eleven UFAs, three (C20:4, C22:3 and C22:4) were significantly higher in the Selected group. Intriguingly, this difference was driven exclusively by polyunsaturated FAs (PUFA), as monounsaturated FAs (MUFA) were significantly more abundant in the Control group: three out of four were more abundant. Among the PUFAs, only one (C18:3 n-3) was more abundant in the Control group (Table 1).
A large number of factors play a part in determining the fatty acid composition of muscles of common carp [32,33]. However, as environmental factors were controlled for in this experiment, the differences observed should be largely attributable to genetics [33]. The observed reduction in the level of oleic acid (MUFA: C18:1) is not a desirable result, as this FA is beneficial both for the total cholesterol and low-density lipoprotein cholesterol levels in plasma in humans and for the fish meat flavour [34]. However, the lower proportion of this MUFA in the Selected group is compensated by the increase in essential PUFAs, the consumption of which is associated with numerous health benefits [11,13]. Therefore, we can conclude that selection for faster growth did not negatively affect the meat quality (at least not in terms of FA composition) of the Selected line.
Transcriptome analyses
The sequencing of twelve intestine samples, which includes eight specimens from the Selected group and four specimens from the Control group, yielded 93.78 Gb of clean data, reaching 6.17 Gb for a single sample, with Q30 over 87% (S1 Table). Clean reads of each sample were compared to the designated common carp reference genome, with mapping successfulness rates ranging from 62.97% to 64.91%. On the basis of the results of alternative splicing prediction, gene structure optimization, and queries of seven different databases, 5 346 new genes were discovered. Differentially expressed genes (DEGs) were identified by statistically comparing the results between the two groups. Functional annotation and enrichment analysis of 106 discovered DEGs was carried out subsequently. Data are deposited in the NCBI’s BioProject (PRJNA414702) and SRA (SRR6202422, SRR6202423, SRR6202418, SRR6202419, SRR6202424, SRR6202425, SRR6202421, SRR6202426, SRR6202420, SRR6202429, SRR6202427, SRR6202428) databases.
Overall distribution of gene ontology terms following the GO database [35] annotation results are comparable to those previously reported for liver samples of bighead carp [23], with the exception of a strong increase in the percentage of DEGs associated with synapse (cellular component) in the intestinal samples of common carp. Comparison of functional category distribution between DEGs and the background transcriptome revealed notable differences in the percentages of genes associated with the following biological processes: biological adhesion, reproductive process and growth, all of which represented between 4 and 8% of all annotated genes, but were completely absent from the DEGs. In the cellular components category, notable differences were observed in the percentage of genes associated with macromolecular complexes, membrane-enclosed lumen, extracellular region part, extracellular matrix and synapse part, all of which appear to be notably less abundant, or even completely absent, from the DEGs. In the molecular function category, notable differences were observed in the percentages of genes associated with nucleic acid binding transcription factor activity and enzyme regulator activity, both of which represented between 3 and 7% of all annotated genes, but were completely absent from the DEGs. Among the categories with higher proportion of DEGs than all annotated genes were (in approximately descending order): guanyl-nucleotyde exchange factor activity (1% among all genes and ≈ 5% among DEGs), synapse, cell junction, receptor activity, metabolic process, multi-organism process, catalytic activity, and structural molecule activity (Fig 1).
‘DEG Unigene’ refers to differentially expressed genes and ‘All Known Gene’ to all annotated genes in both groups.
Functions of DEGs were also predicted and classified by searching COG [36] and KOG [37] databases (Fig 2). Results of these two functional classifications were only partially congruent: by far the highest proportion of DEGs could be merely assigned to the ‘general function prediction only’ functional class in both classifications. In KOG classification, this was followed by ‘signal transduction mechanisms’, which was completely absent from COG, but ‘replication, recombination and repair’ and ‘amino acid transport and metabolism’ were found in both classifications. A small proportion of DEGs were associated with categories likely to be implicated in growth, such as carbohydrate and lipid (absent from COG) transport and metabolism. Intriguingly, although it may be expected that faster growth would be reflected in a number of DEGs related to the energy production, cell growth and division, DEGs classified as ‘energy production and conversion’, ‘cell cycle control, cell division, chromosome partitioning’, ‘cell wall biogenesis’, and ‘extracellular structures’ were absent from both classifications. This scarcity of functional categories of DEGs becomes even more obvious when the results are compared with liver profiles of fast and slow-growing bighead carp, where all functional categories were represented among the discovered DEGs [23]. This is almost certainly at least partially a reflection of the central regulatory role the liver has in the growth of fish [7], as opposed to very specific functions of intestines. However, importantly for the objectives of our study, pathway enrichment analysis showed that (only) endocytosis pathway was significantly enriched in DEGs, compared to the whole genome background (S1 Fig). This finding indicates that endocytosis efficiency might indeed contribute to the observed differences in the growth rates of two groups.
Specific differentially regulated genes
Among the 106 DEGs (S2 Table), we have selected a subset of 25 genes which might be particularly relevant for the objectives of this study (Table 2). Overall, among the DEGs, there were more downregulated than upregulated genes. To corroborate the results of the transcriptome analysis, the expression of a further subset of ten of these DEGs was studied using qPCR (Fig 3). Correlation between transcriptomic and qPCR results was relatively low (S2 Fig), but this was mostly caused by a number of outliers, whereas some genes exhibited almost perfectly congruent results between the two methods (S3 Table). As discussed before [38], this might be a consequence of different data normalization methods used for qPCR (reference gene) and RNA-seq (RPKM) data analyses [39,40]. Gene expression patterns vary widely between tissues in common carp, which poses a problem for RNA-Seq analyses, which assume that samples are comparable [27,39]. These high individual gene expression differences may also explain the relatively low number of DEGs observed, as a large SD would influence the statistical significance calculations. Furthermore, gene expression analyses in teleost fishes are additionally encumbered by a large number of co-expressed, highly similar paralogs [38]: as a result of a relatively recent (≈8 MYA) genome duplication specific to this taxonomic group, teleost fishes possess a large number of duplicated genes [41]. Although many of the superfluous gene copies have been lost in the course of evolution, in many cases two (or more) paralogs have retained their functions and continue to be expressed [42,43]. In common carp, which is an allotetraploid, possessing a large number of functional paralogs [27,44,45], this can be up to four functional paralogs. As observed in our transcriptomic results as well, sometimes paralogs forming teleost gene families are differentially expressed, which indicates that this functional redundancy may have resulted in the evolution of slightly diversified functions and very complex regulation [5,46,47]. Apart from presenting a problem for gene annotation, these paralogs can also produce apparently inconsistent expression results. For example, in cases where qPCR analysis indicates upregulation, whereas the transcriptomics indicates downregulation, it is possible that the primers designed for qPCR may have non-specifically amplified both (or more) differentially expressed paralogs present in the genome.
Data were normalized to gapdh as the reference gene and presented as a fold change between the Selected and Control groups (mean±SD). Results were analyzed by the t-test, where different letters indicate significant (P < 0.05) differences.
A number of regulated genes could be associated with metabolism, growth, development and proliferation. Most importantly for this study, two upregulated genes were associated with endocytosis pathway: interleukin-2 receptor subunit beta (il2rb; infinite upregulation) and EH domain-containing protein 1 (ehd1 or intersectin-1; +1.6). The il2rb gene is involved in receptor-mediated endocytosis and transduces the mitogenic signals of il2 [48]. This gene has not been associated with animal growth before (to our knowledge), but a very recent study indicated that it might be implicated in jaw diversity in pupfishes [49]. The Ehd1 protein interacts with Insulin-like growth factor 1 (Igf-1) receptor. The level of Igf-1 in serum is known to be positively correlated with growth rate in fishes [3]. Generally, in most vertebrates, growth hormone/insulin-like growth factor (Gh/Igf) axis (a major component of which is Igf-1) plays a key role in the regulation of growth, proliferation and differentiation [6,7,9,21,50]. Particularly interesting for this study is that Gh supports muscle protein synthesis indirectly by enhancing rates of amino acid uptake from the intestine and by stimulating intestinal growth [3,51]. This indicates that the higher growth rate observed in the Selected group is likely to be at least partially attributed to increased endocytosis efficiency.
Two of the upregulated genes are known to be involved in the fatty acid metabolism: the enzyme encoded by phytanoyl-CoA dioxygenase, peroxisomal (phyh) gene (transcriptome = 2.3, qPCr = 2.5) is critical for the normal function of peroxisomes, which play a key role in the breakdown of some fatty acids [48,52]. A novel gene was annotated on the basis of its similarity with zebrafish genes: protein containing a beta-ketoacyl synthase, N-terminal domain. This gene, very highly upregulated in both transcriptome (11.5) and qPCR (710.79) analyses, is likely to be responsible for the chain-elongation step of dissociated (type II) fatty-acid biosynthesis, i.e. the addition of two C atoms to the fatty-acid chain [53].
Several other genes upregulated in the Selected group are likely to be indicators of heightened transcription levels: histone h3.3 gene was extremely highly upregulated in the transcriptome analysis (infinite), but only mildly upregulated (0.77) in the qPCR analysis. This gene is believed to represent an epigenetic imprint of transcriptionally active chromatin [48]. Three more upregulated genes are associated with transcription and RNA processing [54]: rna-binding protein 5 (2.01), n-lysine methyltransferase setd8-a (4.15) and fad synthase region (flad1), which was very highly upregulated in both transcriptome (5.77) and qPCR (11.94) analyses. Intriguingly, two setd8-a paralogs were downregulated in the transcriptomic analysis (-5 and -infinite), which indicates that setd8 paralogs might have slightly different functions and complex regulation.
Several more DEGs are associated with metabolic and developmental roles: tolloid-like protein 2 (tll2) gene encodes a protease which processes procollagen C-propeptides, such as chordin. It is required for the embryonic development, where it influences dorso-ventral patterning and skeletogenesis [48,55,56]. It appears to have a broad range of functions, as it has been linked with behaviour in mice and bipolar disorder in humans [57]. Intriguingly, a knockout of this gene resulted in increased muscle mass in mice [58], which is not in agreement with our observations (positive correlation). This gene was extremely highly upregulated in both analyses: infinitely in transcriptome and 610.70 in qPCR. As tolloids were also associated with gut regeneration in sea cucumber; the authors proposed that Bmp1/Tll axis might be involved in folding of the luminal epithelium and gut looping [59]. This raises some interesting hypotheses about the correlation between tll2 expression, proliferation of intestinal epithelium, and growth in common carp, but they need to be further experimentally tested. Calmodulin was very highly upregulated in both analyses: 7.96 in transcriptome and 5.92 in qPCR. It mediates the control of a large number of enzymes, including protein kinases and phosphatases [48]. Calmodulin plays an important role in the activation of phosphorylase kinase, which ultimately leads to glucose being cleaved from glycogen by glycogen phosphorylase. It also plays an important role in lipid metabolism by activating calcitonin, which is a hormone that lowers blood Ca2+ levels and activates G Protein cascades that lead to the generation of cAMP [48]. In fish, calmodulin has been proposed as a molecular stress indicator [60]. Phospholipid-transporting ATPase IC gene (atp10b), a component of the P4-ATPase flippase complex, which catalyzes the hydrolysis of ATP coupled to the transport of aminophospholipids from the outer to the inner leaflet of various membranes and ensures the maintenance of asymmetric distribution of phospholipids [48], was strongly downregulated (- infinite). Phospholipid translocation also seems to be implicated in vesicle formation and in uptake of lipid signaling molecules [48]. The protein is found in brain and in low levels in testis, but its expression is known to be enhanced in intestines in humans [61,62]. However, the reason for its strong downregulation in the Selected group remains unclear.
Intriguingly, two downregulated genes are associated with vision in animals: rpgrip1-like gene (- infinite) encodes a photoreceptor protein that is a key component of cone and rod photoreceptor cells. Mutations in this gene lead to autosomal recessive congenital blindness [54,63,64]. Ependymin-2-like (epd) was downloaded in the transcriptome analysis (-1.9) and slightly upregulated (0.46) in the qPCR analysis. It belongs to the family of proteins predominant in the cerebrospinal fluid of teleost fishes, associated with neuroplasticity and (optic nerve) regeneration [65]. This particular gene (epd-2), however, is expressed in non-brain tissues in fishes. In intestines, it is overexpressed in colon cancer, and its overexpression might be associated with intestinal regeneration [65]. It has also been associated with aggression levels in teleosts: inactivation of ependymin in subdominant fish resulted in a substantial increase in aggression in parallel with an enhanced competitive ability [66]. Therefore, lower expression of this gene in the Selected group may result in more aggressive behaviour during feeding [67] and ultimately faster growth of the fish in this group. Two more downregulated genes have been associated with neuronal function: ST8 Alpha-N-Acetyl-Neuraminide Alpha-2,8-Sialyltransferase 1 (st8sia1: -6.4) has been associated with multiple sclerosis, as well as with metabolism of membrane-bound sphingolipids, important for cell adhesion and growth of cultured malignant cells [68]. Synaptic vesicle 2-related (svop; -5.9) is involved in neuron formation, maturation, or neuronal function [69]. Another gene associated with multiple functions in the synaptic vesicle cycle [70], itsn1, was upregulated in both transcriptome (1.63) and qPCR (1.94) analyses. Low-density lipoprotein receptor class A domain-containing protein 3 (ldlrad3), an important component of a pathway associated with neurodegenerative diseases, including Alzheimer’s [54,71], was strongly downregulated (-4.0). Similarly, rhoGEF and pleckstrin domain-containing protein 1 (farp1) gene, also relatively strongly downregulated (-2.7), plays a role in the assembly and disassembly of dendritic filopodia, the formation of dendritic spines, regulation of dendrite length and ultimately the formation of synapses [48]. It is not clear whether mRNAs of these genes originate from the enteric nervous system, or whether some of them are expressed in cells of the gastrointestinal wall. Regardless, these results indicate that genetically-driven behavioural differences (higher aggression levels) may also have contributed to the higher growth of the Selected group. However, there are also indications that this population might have a higher propensity for inherited diseases leading to sight impairment, as well higher propensity towards neurodegenerative diseases in general.
A number of regulated genes were associated with the immune system: cd276 antigen homolog (upregulated: infinite) modulates immune responses, trim25 gene (downregulated: -infinite) is involved in innate immune defense against viruses, and cxcl14 (downregulated: -1.5) belongs to the cytokine gene family, which encode secreted proteins involved in immunoregulatory and inflammatory processes [48]. cxcl14 might also be involved in the homeostasis of monocyte-derived macrophages, rather than in inflammation [54]. Nine gene isoforms belonging to the NACHT, LRR and PYD domains-containing protein (nlrp) family were also regulated: three nlrp12 gene isoforms and six nlrp3 gene isoforms. Apart from one nlrp3 isoform (upregulated: 2.43), all other isoforms were downregulated: nlrp3 from -2.01 to -5.74, and nlrp12 from -2.02 to -2.31. Whereas in mammals this family contains only several members, in teleost fish, NLR proteins have expanded into a huge family containing hundreds of genes, which mostly act as innate immune sensors for pathogen-associated stress signals [72]. Therefore, if we presume that studied carps were exposed to similar pathogens, downregulation of these paralogs indicates a possibility of slightly impaired immune responses of the Selected group.
Two downregulated genes are associated with the development of vasculature: mutations in glomulin (-6.2) have been associated with glomuvenous malformations [73], whereas angiogenic factor with G patch and FHA domains 1 gene (aggf1; -1.73) promotes angiogenesis and proliferation of endothelial cells. Another two downregulated genes might be an indication of a higher likelihood of health-related problems in the Selected group: talin-1 (tln1; -1.5) is involved in connections of major cytoskeletal structures to the plasma membrane and cell-cell contacts [48], whereas proteoglycan 4-like (-infinite) prevents protein deposition onto cartilage from synovial fluid by controlling adhesion-dependent synovial growth and inhibiting the adhesion of synovial cells to the cartilage surface [48].
To sum up the results: two upregulated genes are known to be involved in the fatty acid metabolism, several genes are likely to be indicators of heightened transcription levels and proliferation, several are associated with metabolic and developmental roles, several with neuronal functions (including two with vision), several with the immune system, and two downregulated genes with the development of vasculature.
Conclusions
Gene expression analyses indicate that the higher growth rate of the Selected group is likely to be at least partially attributed to increased endocytosis efficiency and genetically-driven behavioural differences (higher aggression levels). Among the detected DEGs, il2rb is the most likely candidate gene to explain the increased growth rate, while phyh is the most likely candidate to explain the increased unsaturated fatty acid content in the muscles of the new Huanghe carp strain. Our results contribute to the understanding of the genetic basis for growth and unsaturated fatty acid content in fish, as well as vertebrates in general.
Materials and methods
Strain creation
Following the national aquaculture development strategy, a new strain of common carp was created at the Nanquan farm of the Freshwater Fishery Research Center (FFRC) of Chinese Fisheries Academy in the period between 2010 and 2016 with the aim of achieving higher growth rate and unsaturated fatty acid muscle content, and lower feeding costs [31]. The basis for the strain creation was a population (n = 1600) of the ‘Yellow River’ (Yuxuan Huanghe) strain specimens introduced from the Fisheries Science Research Institute of the Henan Academy of Fishery Sciences and a local Jian carp population (n = 20). A subpopulation of about 600 Huanghe and 20 Jian specimens exhibiting a comparatively high growth rate was selected in 2010 for the creation of the new strain (age ≈ 17-month-old, average weight = 197.40±75.08 g) via artificial breeding (S3 Fig). The best model of parental pair selection was inferred via BLUP analysis [74] on the basis of their breeding values and inbreeding coefficients (<0.02). ‘Selected’ line was created in the following way: families and individuals in each family were ranked according to their breeding values, then male and female specimens with highest breeding values were selected from the top-ranking families, and their inbreeding coefficient values were calculated. If the inbreeding coefficient value of a given pair was below 0.02 it was chosen for the Selected line family. A total of approximately 60 Selected line families (the number varied among generations) were created in this way, with each family represented by 50 (offspring) specimens produced by a selected parental pair. The control line (average number of families per generation = 20) was generated by random mating of specimens with inbreeding coefficient < 0.02. Their breeding values and the best model for parental pair selection were inferred via a BLUP analysis. When they reached the size of about 10 g (≈ 3 months of age), F1 juveniles were PIT (Passive Integrated Transponder) tagged for subsequent identification, and growth performance data (length and weight) were simultaneously collected. At 8 months of age, growth performance of the F1 offspring was measured again, which was followed by a comprehensive screening for molecular markers, BLUP analysis, and unsaturated fatty acid muscle content analysis, to estimate the breeding value of the broodstock. For the molecular markers we selected five markers (SNPs and SSRs) which we previously found to be associated with body weight in carp: IGF2a4#, IGF2R intron 1, D-LOOP253, COI626, and Koi42 [75]. DNA samples of parents and selected (high-performing) 500 offspring were also collected. F2 parents were selected on the basis of juvenile growth performance and their molecular markers. The best model for parental pair selection was obtained via data integration with the help of SAS/STAT 8.0 (SAS Institute Inc., Cary, NC, USA) and R3.1.14 [76] programs. The breeding value was estimated using ASReml2 [77], DMU [78], Python [79], and two unpublished (but copyrighted) in-house developed programs, DNA compare1.0 and COECAL2.1. This process was repeated for three generations (2010–2016, two years per generation; S3 Fig): one-to-one artificial propagation of the core population was conducted in the early May of 2012, 2014 and 2016 following the standard procedure [31,80]. In 2016, after three generations of selection and breeding, we had a stock of 4000 (80 × 50) selected F3 specimens, which were released into the same pond for the breeding program (not further discussed in this study). In addition to these, we randomly selected five specimens from 48 families (5 × 48 = 240 specimens) for the growth performance test (cage culture experiment) and unsaturated fatty acid muscle content analyses. Among these 48 families, 38 were from the Selected line and 10 were from the control group. When they reached the size of about 10 g (≈ 3 months of age), all 240 juveniles were PIT (Passive Integrated Transponder) tagged for subsequent identification.
Growth performance and tissue sampling
To test the growth performance of the new strain, the aforementioned 240 healthy one-year-old F3 generation specimens were reared in eight 1×1×1.2 m cages (approx. 30 fish per cage) placed in earthen ponds from early September to mid-November 2016 (8 weeks). Selected and Control lines were reared together, and there was no significant difference in the initial weight (at the onset of the cage-rearing experiment) between the eight groups (cages). Light/dark schedule was natural. Fish were fed pelleted feed (S4 Table) three times daily (8:00, 12:00, 16:00) to satiation (until fish would stop taking floating pellets from the surface). The amount of feed was, thus, variable. Hand clapping was used as the feeding signal. The last feeding was on the 56th day, 8:00am, at which point the fish was eight months-old. After starving the fish for 24h, at 8:30am of the 57th day all 240 individuals (mortality rate = 0.0) were tranquilised in 20–30 mg/L MS-222 (Sigma-Aldrich, USA) as described [81]. Selected morphometric traits were recorded: weight (g, 2 kg scale), standard length, body depth, body width (all to mm, using a ruler) (S5 Table). One family from both lines (Control and Selected) was randomly chosen for tissue sampling, followed by random selection of eight specimens from the Selected group and four specimens from the Control group (S1 and S5 Tables). These twelve specimens (5 male + 7 female, S1 Table) were euthanised in buffered MS-222 at 350 mg/L concentration [81]. The fish were then dissected and a caudal portion of the epaxial muscle (> 10 g) sampled using sterile scissors and forceps, and placed in sterile plastic bags. After weighing the entire viscera, the intestine was rinsed three times with sterile phosphate buffered saline (PBS, pH = 7.0) to remove the ingesta, and foregut, midgut and hindgut separated as described [82]. Samples were immediately (<3min) placed in labelled sterile plastic bags, flash-frozen in liquid nitrogen, and stored at -80°C. The handling of animals was conducted in accordance with the guidelines for the care and use of animals for scientific purposes set by the Institutional Animal Care and Use Committee of the Freshwater Fisheries Research Center, Chinese Academy of Fishery Sciences, Wuxi, China and the EU Directive 2010/63/EU for animal experiments. The permit to conduct this study was obtained from the institutional Animal Care and Use Committee of the Freshwater Fisheries Research Center, Chinese Academy of Fishery Sciences.
Analysis of fatty acid composition
To determine the effects of selection for fast growth on fatty acid (FA) composition of carp meat, muscle samples of eight specimens from the Selected group and four specimens from the Control group were collected (described above) and analysed. As fish muscles are mostly composed of long-chain fatty acids [33,83], we focused on the composition of long-chain (C12 to C22) FAs in this experiment. A portion (≈4 g) of each muscle sample was manually crushed in liquid nitrogen using mortar and pestle and sent (on ice) to the State Key Laboratory of Food Science (Jiangnan University, Beijing), where the total lipid extraction and analysis was performed as described before [33,84]. Briefly, lipids were extracted using chloroform/methanol (v/v = 2:1), methylated with 10% (v/v) methanolic HCl at 60°C for 3 h, extracted with n-hexane, and analysed by gas chromatography using Agilent DB-WAXETR column (Agilent Technologies, Santa Clara, CA, US), according to the manufacturer’s protocol. Pentadecanoic acid (15:0) was used as the internal standard. Statistical analysis was conducted in Excel, with the significance threshold set at 0.05.
Transcriptome analysis
RNA isolation, cDNA library construction and Illumina sequencing.
Total RNA for each sample was extracted from the intestine samples, cDNA libraries constructed and sequenced, and transcriptome assembled and annotated by the Beijing Biotech Co. Ltd. (Beijing, China) roughly as described before [85,86]. In total, eight specimens from the Selected group and four specimens from the Control group were used for the transcriptome analyses. To ensure that foregut, midgut and hindgut segments are represented in the transcriptome, RNA was extracted (Trizol, Invitrogen, US) separately from the three segments, and equal amounts (of the three segments belonging to the same specimen) were then pooled together. RNA integrity and concentration were assessed using an Agilent Bioanalyzer 2100 system (Agilent Technologies, CA, USA). cDNA libraries were generated using NEBNext UltraTM RNA Library Prep Kit for Illumina (NEB, USA) following the manufacturer’s recommendations. One μg of RNA per sample was used as input material. Index codes were added to attribute sequences to each sample. Briefly: poly-T oligo-attached magnetic beads (NEB, E7490) were used to purify the mRNA from total RNA. Fragmentation was carried out using divalent cations under elevated temperature in NEBNext First Strand Synthesis Reaction Buffer (5X). First strand cDNA was synthesized using random hexamer primer and M-MuLV Reverse Transcriptase (RNase H). Second strand cDNA synthesis was subsequently performed using DNA Polymerase I and RNase H. Remaining overhangs were converted into blunt ends via exonuclease/polymerase activities. After adenylation of 3’ ends of DNA fragments, NEBNext adaptors with hairpin loop structure were ligated to prepare for hybridisation. In order to select cDNA fragments of preferentially 200–250 bp in length, library fragments were purified with AMPure XP system (Beckman Coulter, Beverly, MA, USA). Following this, 3 μl of USER Enzyme (NEB, USA) was incubated with size-selected, adaptor-ligated cDNA at 37°C for 15 min, followed by 5 min at 95°C. PCR was then performed with Phusion High-Fidelity DNA polymerase, universal PCR primers and Index (X) Primer. Finally, PCR products were purified (AMPure XP system) and library quality assessed on the Agilent Bioanalyzer 2100 system. Clustering of the index-coded samples was performed on a cBot Cluster Generation System using TruSeq PE Cluster Kit v4-cBot-HS (Illumina) according to the manufacturer’s instructions. After cluster generation, the library preparations were sequenced on an Illumina Hiseq 2500 platform and paired-end reads were generated.
Transcriptome assembly, annotation, ontology, and differential gene expression.
Raw data (raw reads) were first processed through in-house perl scripts WipeAadpter.pl and Fastq_filter (Biomarker Technologies, Beijing): adaptor-only reads, reads containing poly-N stretches (>5% total N), and low-quality reads (Q20-value≤20) were filtered out, thereby producing ‘clean reads’. These clean reads of each sample were compared to the designated reference common carp genome downloaded from the CarpBase [87] using TopHat2 [88], based on Bowtie2 algorithm [89]. The numbers of reads produced by RNA-Seq analyses were normalized to FPKM (fragments per kilobase of transcripts per million fragments mapped) to compute gene expression levels [90], and differentially expressed genes (DEGs) between the two groups (S and C) detected using EBseq software [91]. Benjamini-Hochberg procedure [92] was used to control the false discovery rate (FDR). Genes were defined as differentially expressed when they exhibited the following parameters: |fold change|≥ 2 and FDR < 0.01. Genes were queried against a number of databases using BLASTx [93]: RefSeq [94], UniProt [95], GO [35], COG [36], KOG [37], Pfam [96], and KEGG [97]. After using KOBAS 2.0 [98] to obtain the KEGG Orthology results, HMMER [99] web server was employed to obtain the annotation for new genes from the predicted amino acid sequences using Pfam database. Pathway enrichment analysis was conducted using KEGG database as described before [100].
qPCR
This method was used to further investigate the expression of a subset of ten genes significantly differentially expressed between the two groups (eight specimens for the Selected group and four specimens for the Control group). Genes were selected based on their relevance for the objectives of this study and regulation magnitude. Total RNA was extracted prepared as described above from the same samples that were used for the transcriptome sequencing. cDNA libraries were prepared using PrimeScriptTM RT reagent Kit with gDNA Eraser (Takara) following the manufacturer’s protocol. cDNA libraries were diluted five-fold and used as templates for qPCR with the primers listed in Table 3. Three reference genes, gapdh, beta actin and 18S, were tested for the stability of expression among all samples; as reflected in no significant differences between Selection and Control groups (S6 Table), all three genes exhibited a stable expression. On the basis of our previous experiences and published studies [101,102], we selected gapdh as the reference gene. Primers were designed using Primer Premier 5 software (Premier Biosoft, USA) on the basis of sequences of these ten genes obtained from the transcriptome data, and synthesized by the WCcgene company (Shanghai, China). qPCR was performed using ABI VIIA@7 instrument (ABI, USA) and GoTaq qPCR Master Mix (Promega, USA). In brief: the total qPCR mixture reaction volume of 10 μL contained 5 μL GoTaq qPCR Master Mix 2X, 3 μL ddH2O, 0.75 μL of each primer and 0.5 μL of cDNA template. qPCR procedure: pre-incubation at 95°C for 10 min, followed by 40 cycles of 15 s at 95°C, 30 s at 60°C, and 30 s at 72°C. Melt curve analysis was performed at the end. Reactions were performed in triplicate for each sample. gapdh was chosen as the reference gene according to earlier recommendations [103,104]. Expression levels are presented as fold changes relative to the target gene expression in the control group, calculated using the 2–ΔΔCt method [105]. qPCR data were analyzed statistically using Microsoft Excel and SPSS. Student-T test, implemented in SPSS, was applied to test the statistical significance of differences between the two groups, where significance thresholds were set at P < 0.05 (significant) and P < 0.01 (highly significant).
Supporting information
S1 Table. Transcriptome sequencing statistics for the 12 samples.
https://doi.org/10.1371/journal.pone.0206615.s001
(DOCX)
S3 Table. Comparison of gene expression results between transcriptomic and qPCR data.
Log2FC is the result of transcriptome analysis, Reg is regulation, and inf is infinite.
https://doi.org/10.1371/journal.pone.0206615.s003
(DOCX)
S5 Table. Identification details and growth parameters for the 12 specimens selected for RNA-seq and fatty acid composition analyses.
BWe = body weight (g), TL = total length (mm), CF = condition factor, BWi = body width (mm), BH = body height (mm). The last row in each group contains the average values of morphometric characteristics.
https://doi.org/10.1371/journal.pone.0206615.s005
(DOCX)
S6 Table. Ct values of three tested reference genes in all samples.
https://doi.org/10.1371/journal.pone.0206615.s006
(DOCX)
S1 Fig. Endocytosis pathway enrichment analysis in the KEGG database.
Significantly enriched genes are highlighted in red.
https://doi.org/10.1371/journal.pone.0206615.s007
(PNG)
S2 Fig. Correlation between transcriptome and qPCR data.
Transcriptome results are presented on the x-axis (log2FC) and qPCR on the y-axis. The following outliers were removed from the analysis: h3.3 and tll2 (inf) and novel β-ketoacyl (>700) (see S3 Table).
https://doi.org/10.1371/journal.pone.0206615.s008
(PDF)
S3 Fig. New Huanghe carp strain selection workflow.
https://doi.org/10.1371/journal.pone.0206615.s009
(PDF)
Acknowledgments
We thank the students and staff of the Aquatic Genetic Laboratory, Freshwater Fisheries Research Center of Chinese Academy of Fishery Sciences for their kind assistance in the study.
References
- 1. Gui J, Zhu Z. Molecular basis and genetic improvement of economically important traits in aquaculture animals. Chinese Sci Bull. SP Science China Press; 2012;57: 1751–1760.
- 2.
FAO. The State of World Fisheries and Aquaculture, 2014. FAO Fisheries Proceedings. 2014. 92-5-105177-1
- 3. Mommsen TP. Paradigms of growth in fish. Comp Biochem Physiol Part B Biochem Mol Biol. 2001;129: 207–219.
- 4. Sire MF, Vernier J-M. Intestinal absorption of protein in teleost fish. Comp Biochem Physiol Part A Physiol. 1992;103: 771–781.
- 5. Johnston IA, Bower NI, Macqueen DJ. Growth and the regulation of myotomal muscle mass in teleost fish. J Exp Biol. 2011;214: 1617–28. pmid:21525308
- 6. Duan C. The Insulin-like Growth Factor System and Its Biological Actions in Fish. Am Zool. 1997;37: 491–503.
- 7. Reinecke M. Influences of the environment on the endocrine and paracrine fish growth hormone-insulin-like growth factor-I system. J Fish Biol. Blackwell Publishing Ltd; 2010;76: 1233–1254. pmid:20537012
- 8. Boeuf G, Boujard D, Ruyet JP- L. Control of the somatic growth in turbot. J Fish Biol. Blackwell Publishing Ltd; 1999;55: 128–147.
- 9. Vélez EJ, Lutfi E, Azizi S, Perelló M, Salmerón C, Riera-Codina M, et al. Understanding fish muscle growth regulation to optimize aquaculture production. Aquaculture. 2017;467: 28–40.
- 10. Wood JD, Enser M, Fisher AV, Nute GR, Sheard PR, Richardson RI, et al. Fat deposition, fatty acid composition and meat quality: A review. Meat Sci. 2008;78: 343–358. pmid:22062452
- 11. Lunn J, Theobald HE. The health effects of dietary unsaturated fatty acids. Nutr Bull. 2006;31: 178–224.
- 12. Williams CM. Dietary fatty acids and human health. Ann Zootech. 2000;49: 165–180.
- 13. Steffens W. Aquaculture produces wholesome food: cultured fish as a valuable source of n-3 fatty acids. Aquac Int. Springer International Publishing; 2016;24: 787–802.
- 14. Stancheva M, Merdzhanova A. Fatty acid composition of common carp, rainbow trout and grey mullet fish species. Agric Sci Technol. Trakia University; 2011;3: 285–289.
- 15. Martin SAM, Dehler CE, Król E. Transcriptomic responses in the fish intestine. Dev Comp Immunol. 2016;64: 103–117. pmid:26995769
- 16. Betancor MB, Li K, Sprague M, Bardal T, Sayanova O, Usher S, et al. An oil containing EPA and DHA from transgenic Camelina sativa to replace marine fish oil in feeds for Atlantic salmon (Salmo salar L.): Effects on intestinal transcriptome, histology, tissue fatty acid profiles and plasma biochemistry. Soengas JL, editor. PLoS One. Taylor & Francis, CRC Press; 2017;12: e0175415. pmid:28403232
- 17. De Santis C, Taylor JF, Martinez-Rubio L, Boltana S, Tocher DR, Bron J. Influence of Development and Dietary Phospholipid Content and Composition on Intestinal Transcriptome of Atlantic Salmon (Salmo salar). Fernandes JMO, editor. PLoS One. Elsevier; 2015;10: e0140964. pmid:26488165
- 18. Matsumoto Y, Murakami Y, Tsuru S, Ying B- W, Yomo T. Growth rate-coordinated transcriptome reorganization in bacteria. BMC Genomics. 2013;14: 808. pmid:24252326
- 19. Huang G, Han M, Yao W, Wang Y. Transcriptome analysis reveals the regulation of brassinosteroids on petal growth in Gerbera hybrida. PeerJ. PeerJ Inc.; 2017;5: e3382. pmid:28584713
- 20. Puig-Oliveras A, Ramayo-Caldas Y, Corominas J, Estell? J, P?rez-Montarelo D, Hudson NJ, et al. Differences in Muscle Transcriptome among Pigs Phenotypically Extreme for Fatty Acid Composition. Guillou H, editor. PLoS One. 2014;9: e99720. pmid:24926690
- 21. Sun Y, Guo C-Y, Wang D-D, Li XF, Xiao L, Zhang X, et al. Transcriptome analysis reveals the molecular mechanisms underlying growth superiority in a novel grouper hybrid (Epinephelus fuscogutatus♀ × E. lanceolatus♂). BMC Genet. 2016;17: 24. pmid:26785614
- 22. Gao Y, Zhang H, Gao Q, Wang L, Zhang F, Siva VS, et al. Transcriptome Analysis of Artificial Hybrid Pufferfish Jiyan-1 and Its Parental Species: Implications for Pufferfish Heterosis. Yue GH, editor. PLoS One. Public Library of Science; 2013;8: e58453. pmid:23520511
- 23. Fu B, Wang X, Feng X, Yu X, Tong J. Comparative transcriptomic analyses of two bighead carp (Hypophthalmichthys nobilis) groups with different growth rates. Comp Biochem Physiol Part D Genomics Proteomics. 2016;20: 111–117. pmid:27639030
- 24. Ma D, Ma A, Huang Z, Wang G, Wang T, Xia D, et al. Transcriptome Analysis for Identification of Genes Related to Gonad Differentiation, Growth, Immune Response and Marker Discovery in The Turbot (Scophthalmus maximus). Song L, editor. PLoS One. Public Library of Science; 2016;11: e0149414. pmid:26925843
- 25. He L, Pei Y, Jiang Y, Li Y, Liao L, Zhu Z, et al. Global gene expression patterns of grass carp following compensatory growth. BMC Genomics. 2015;16: 184. pmid:25887225
- 26. Tymchuk W, Sakhrani D, Devlin R. Domestication causes large-scale effects on gene expression in rainbow trout: Analysis of muscle, liver and brain transcriptomes. Gen Comp Endocrinol. 2009;164: 175–183. pmid:19481085
- 27. Kolder ICRM, van der Plas-Duivesteijn SJ, Tan G, Wiegertjes GF, Forlenza M, Guler AT, et al. A full-body transcriptome and proteome resource for the European common carp. BMC Genomics. 2016;17: 701. pmid:27590662
- 28. W J., D A., G X., L Z., M R., M M., et al. The hookworm Ancylostoma ceylanicum intestinal transcriptome provides a platform for selecting drug and vaccine candidates. Parasites and Vectors. 2016. p. 518. pmid:27677574
- 29. Sun L, Yang H, Chen M, Ma D, Lin C. RNA-Seq Reveals Dynamic Changes of Gene Expression in Key Stages of Intestine Regeneration in the Sea Cucumber Apostichopus japonicas. PLoS One. 2013;8. pmid:23936330
- 30. Blazie SM, Babb C, Wilky H, Rawls A, Park JG, Mangone M. Comparative RNA-Seq analysis reveals pervasive tissue-specific alternative polyadenylation in Caenorhabditis elegans intestine and muscles. BMC Biol. 2015;13: 4. pmid:25601023
- 31. Su S-Y, Zhang C-F, Dong Z-J, P X, Y X-H. The breeding gain of Huanghe carp (Cyprinus carpio hacmalopterus Temminck et Schlegel) new strain G3- The effect of higher carbohydrate diet on growth and fatty acid profile of huanghe carp new strain. J Yangzhou Univ (Agricultural Life Sci Ed. 2018;39: 63–66.
- 32. Mráz J, Pickova J. Factors influencing fatty acid composition of common carp (Cyprinus carpio) muscle. Neuro Endocrinol Lett. 2011;32 Suppl 2: 3–8.
- 33. Buchtová H, Svobodová Z, Kocour M, Velíšek J. Chemical Composition of Fillets of Mirror Crossbreds Common Carp (Cyprinus carpio L.). Acta Vet Brno. University of Veterinary and Pharmaceutical Sciences; 2010;79: 551–557.
- 34. Mir Z, Paterson LJ, Mir PS. Fatty acid composition and conjugated linoleic acid content of intramuscular fat in crossbred cattle with and without Wagyu genetics fed a barley-based diet. Can J Anim Sci. 2000;80: 195–197.
- 35. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, et al. Gene Ontology: tool for the unification of biology. Nat Genet. 2000;25: 25–29. pmid:10802651
- 36. Tatusov RL, Koonin E V., Lipman DJ. A Genomic Perspective on Protein Families. Science (80-). 1997;278: 631–637.
- 37. Koonin E V, Fedorova ND, Jackson JD, Jacobs AR, Krylov DM, Makarova KS, et al. A comprehensive evolutionary classification of proteins encoded in complete eukaryotic genomes. Genome Biol. 2004;5: R7. pmid:14759257
- 38. Prisingkorn W, Prathomya P, Jakovlić I, Liu H, Zhao Y-HH, Wang W-MM. Transcriptomics, metabolomics and histology indicate that high-carbohydrate diet negatively affects the liver health of blunt snout bream (Megalobrama amblycephala). BMC Genomics. 2017;18: 856. pmid:29121861
- 39. Fonseca NA, Marioni J, Brazma A. RNA-Seq gene profiling—A systematic empirical comparison. PLoS One. 2014;9. pmid:25268973
- 40. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001;25: 402–8. pmid:11846609
- 41. Taylor JS, Van de Peer Y, Braasch I, Meyer A. Comparative genomics provides evidence for an ancient genome duplication event in fish. Philos Trans R Soc B Biol Sci. 2001;356: 1661–1679. pmid:11604130
- 42. Pasquier J, Cabau C, Nguyen T, Jouanno E, Severac D, Braasch I, et al. Gene evolution and gene expression after whole genome duplication in fish: the PhyloFish database. BMC Genomics. 2016;17: 368. pmid:27189481
- 43. Crollius HR, Weissenbach J. Fish genomics and biology. Genome Research. 2005. pp. 1675–1682. pmid:16339365
- 44. David L, Blum S, Feldman MW, Lavi U, Hillel J. Recent duplication of the common carp (Cyprinus carpio L.) genome as revealed by analyses of microsatellite loci. Mol Biol Evol. 2003;20: 1425–34. pmid:12832638
- 45. Xu P, Zhang X, Wang X, Li J, Liu G, Kuang Y, et al. Genome sequence and genetic diversity of the common carp, Cyprinus carpio. Nat Genet. 2014;advance on. http://www.nature.com/ng/journal/vaop/ncurrent/abs/ng.3098.html#supplementary-information pmid:25240282
- 46. Jakovlić I, Liu H, Wang W- M. Identification, origin and evidence for retained functionality of two IκBα paralogs in Megalobrama amblycephala. Dev Comp Immunol. 2016;62: 89–96. pmid:27155355
- 47. Jakovlić I, Wang W- M. Expression of Hox paralog group 13 genes in adult and developing Megalobrama amblycephala. Gene Expr Patterns. 2016;21: 63–68. pmid:27476414
- 48. Consortium U. UniProt: the universal protein knowledgebase. Nucleic Acids Res. Oxford University Press; 2017;45: D158–D169. pmid:27899622
- 49. Lencer ES, Warren WC, Harrison R, McCune AR. The Cyprinodon variegatus genome reveals gene expression changes underlying differences in skull morphology among closely related species. BMC Genomics. 2017;18: 424. pmid:28558659
- 50. Mullis PE. Genetic control of growth. Eur J Endocrinol. 2005;152: 11–31. pmid:15762183
- 51. Farmanfarmaian A, Sun LZ. Growth hormone effects on essential amino acid absorption, muscle amino acid profile, N-retention and nutritional requirements of striped bass hybrids. Genet Anal Biomol Eng [Genet Anal Biomol Eng]. 1999;15: 107–113.
- 52. Zomer AWM, Jansen GA, van der Burg B, Verhoeven NM, Jakobs C, van der Saag PT, et al. Phytanoyl-Coa hydroxylase activity is induced by phytanic acid. Eur J Biochem. Blackwell Science Ltd; 2000;267: 4063–4067. pmid:10866807
- 53. Artimo P, Jonnalagedda M, Arnold K, Baratin D, Csardi G, De Castro E, et al. ExPASy: SIB bioinformatics resource portal. Nucleic Acids Res. 2012;40. pmid:22661580
- 54. Rebhan M, Chalifa-Caspi V, Prilusky J, Lancet D. GeneCards: A novel functional genomics compendium with automated data mining and query reformulation support. Bioinformatics. 1998;14: 656–664. pmid:9789091
- 55. Connors SA, Trout J, Ekker M, Mullins MC. The role of tolloid/mini fin in dorsoventral pattern formation of the zebrafish embryo. Development. 1999;126: 3119–30. pmid:10375503
- 56. Branam AM, Hoffman GG, Pelegri F, Greenspan DS. Zebrafish chordin-like and chordin are functionally redundant in regulating patterning of the dorsoventral axis. Dev Biol. 2010;341: 444–458. pmid:20226780
- 57. de Mooij-van Malsen JG, van Lith HA, Laarakker MC, Brandys MK, Oppelaar H, Collier DA, et al. Cross-species genetics converge to TLL2 for mouse avoidance behavior and human bipolar disorder. Genes, Brain Behav. 2013;12: 653–657. pmid:23777486
- 58. Lee SJ. Genetic analysis of the role of proteolysis in the activation of latent myostatin. PLoS One. 2008;3. pmid:18286185
- 59. Mashanov VS, Zueva OR, Garcia-Arraras JE. Expression of Wnt9, TCTP, and Bmp1/Tll in sea cucumber visceral regeneration. Gene Expr Patterns. NIH Public Access; 2012;12: 24–35. pmid:22079950
- 60. Alves RN, Cordeiro O, Silva TS, Richard N, de Vareilles M, Marino G, et al. Metabolic molecular indicators of chronic stress in gilthead seabream (Sparus aurata) using comparative proteomics. Aquaculture. 2010;299: 57–66.
- 61. Uhlén M, Fagerberg L, Hallström BM, Lindskog C, Oksvold P, Mardinoglu A, et al. Tissue-based map of the human proteome. Science (80-). 2015;347: 1260419–1260419. pmid:25613900
- 62.
The Human Protein Atlas—ATP10B [Internet]. [cited 26 Jun 2017]. Available: http://www.proteinatlas.org/ENSG00000118322-ATP10B/tissue
- 63. Kostic C, Arsenijevic Y. Animal modelling for inherited central vision loss. J Pathol. John Wiley & Sons, Ltd; 2016;238: 300–310. pmid:26387748
- 64. Miyadera K, Brierley I, Aguirre-Hernández J, Mellersh CS, Sargan DR. Multiple Mechanisms Contribute to Leakiness of a Frameshift Mutation in Canine Cone-Rod Dystrophy. PLoS One. 2012;7. pmid:23251588
- 65. Suárez-Castillo EC, García-Arrarás JE. Molecular evolution of the ependymin protein family: a necessary update. BMC Evol Biol. 2007;7: 1–20.
- 66. Sneddon LU, Schmidt R, Fang Y, Cossins AR. Molecular correlates of social dominance: A novel role for ependymin in aggression. PLoS One. 2011;6. pmid:21483679
- 67. Sneddon LU. The bold and the shy: Individual differences in rainbow trout. J Fish Biol. 2003;62: 971–975.
- 68. Husain S, Yildirim-Toruner C, Rubio JP, Field J, Schwalb M, Cook S, et al. Variants of ST8SIA1 are associated with risk of developing multiple sclerosis. PLoS One. 2008;3. pmid:18612409
- 69. Logan MA, Steele MR, Vetter ML. Expression of synaptic vesicle two-related protein SVOP in the developing nervous system of Xenopus laevis. Dev Dyn. 2005;234: 802–807. pmid:16217738
- 70. Pechstein A, Shupliakov O, Haucke V, Heuser JE, Reese TS, Murthy VN, et al. Intersectin 1: a versatile actor in the synaptic vesicle cycle. Biochem Soc Trans. 2010;38: 181–6. pmid:20074056
- 71. Ranganathan S, Noyes NC, Migliorini M, Winkles JA, Battey FD, Hyman BT, et al. LRAD3, A Novel Low-Density Lipoprotein Receptor Family Member That Modulates Amyloid Precursor Protein Trafficking. J Neurosci. 2011;31: 10836–10846. pmid:21795536
- 72. Howe K, Schiffer PH, Zielinski J, Wiehe T, Laird GK, Marioni JC, et al. Structure and evolutionary history of a large family of NLR proteins in the zebrafish. Open Biol. 2016;6: 160009. pmid:27248802
- 73. Brouillard P, Boon LM, Mulliken JB, Enjolras O, Ghassibé M, Warman ML, et al. Mutations in a Novel Factor, Glomulin, Are Responsible for Glomuvenous\rMalformations (“Glomangiomas”). Am J Hum Genet. 2002;70: 866–874. pmid:11845407
- 74.
Henderson CR. Applications of Linear Models in Animal Breeding Models. Guelph: University of Guelph; 1984. https://doi.org/10.1002/9780470316856.ch7
- 75. Chengfeng Z, Su S, Zhu J, Shi L. Candidate parents selection containing multiple genotypes with higher performance and correlated molecular markers’ genetic effects analysis. Acta Hydrobiol Sin. 2017;41: 79–85.
- 76.
Harrell FE. Regression modeling strategies With Applications to Linear Models, Logistic Regression, and Survival Analysis. Springer Series in Statistics. Springer; 2014. https://doi.org/10.1007/978-1-4757-3462-1
- 77.
Gilmour AR, Butler D, Cullis BR, Thompson R, others. ASReml 2 and SAMM 2-mixed models for many occasions. Proceedings of the 8th World Congress on Genetics Applied to Livestock Production, Belo Horizonte, Minas Gerais, Brazil, 13–18 August, 2006. 2006. pp. 12–27.
- 78.
Madsen P, Sørensen P, Su G, Damgaard LH, Thomsen H, Labouriau R, et al. DMU-a package for analyzing multivariate mixed models. 8th World Congress on Genetics Applied to Livestock Production. 2006.
- 79.
Van Rossum G, others. Python 2.7. 10 Language Reference. Samurai Media Limited; 2015.
- 80. Szabó T, Ditrói B, Szabó K, Bokor Z, Urbányi B. Comparison of the Efficiency of Common Carp and Silver Carp Pituitary in the Breeding of Common Carp (Cyprinus carpio) and Northern Pike (Esox lucius). Turkish J Fish Aquat Sci. 2014;14: 841–844.
- 81. Topic Popovic N, Strunjak-Perovic I, Coz-Rakovac R, Barisic J, Jadan M, Persin Berakovic A, et al. Tricaine methane-sulfonate (MS-222) application in fish anaesthesia. J Appl Ichthyol. 2012;28: 553–564.
- 82.
Harder W, Sokoloff S. Anatomy of fishes, part I: text. Stuttgart: Schweizerbartsche Verlagsbuchhandlung; 1975.
- 83. Yeganeh S, Shabanpour B, Hosseini H, Reza Imanpour M, Shabani A. Comparison of Farmed and Wild Common Carp (Cyprinus carpio): Seasonal Variations in Chemical Composition and Fatty Acid Profile. Czech J Food Sci. 2012;30: 503–511.
- 84.
Tang X, Chen H, Chen YQ, Chen W, Garre V, Song Y, et al. Comparison of Biochemical Activities between High and Low Lipid-Producing Strains of Mucor circinelloides: An Explanation for the High Oleaginicity of Strain WJ11. Bahn Y-S, editor. PLoS One. CRC press; 2015;10: e0128396.
- 85. Yue H, Li C, Du H, Zhang S, Wei Q. Sequencing and De Novo Assembly of the Gonadal Transcriptome of the Endangered Chinese Sturgeon (Acipenser sinensis). PLoS One. Public Library of Science; 2015;10: e0127332. pmid:26030930
- 86. Liu X, Xu X, Li B, Wang X, Wang G, Li M. RNA-seq transcriptome analysis of maize inbred carrying nicosulfuron-tolerant and nicosulfuron-susceptible alleles. Int J Mol Sci. 2015;16: 5975–5989. pmid:25782159
- 87. Williams DR, Li W, Hughes MA, Gonzalez SF, Vernon C, Vidal MC, et al. Genomic resources and microarrays for the common carp Cyprinus carpio L. J Fish Biol. Blackwell Publishing Ltd; 2008;72: 2095–2117.
- 88. Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. 2013;14: R36. pmid:23618408
- 89. Langmead B, Trapnell C, Pop M, Salzberg S. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 2009;10: R25. pmid:19261174
- 90. Mortazavi A, Williams BA, McCue K, Schaeffer L, Wold B. Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat Methods. 2008;5: 621–8. pmid:18516045
- 91. Leng N, Dawson JA, Thomson JA, Ruotti V, Rissman AI, Smits BMG, et al. EBSeq: An empirical Bayes hierarchical model for inference in RNA-seq experiments. Bioinformatics. 2013;29: 1035–1043. pmid:23428641
- 92. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Ser B. 1995;57: 289–300.
- 93. Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W, et al. Gapped BLAST and PSI-BLAST: A new generation of protein database search programs. Nucleic Acids Res. 1997;25: 3389–3402. pmid:9254694
- 94. Pruitt KD, Tatusova T, Maglott DR. NCBI reference sequences (RefSeq): A curated non-redundant sequence database of genomes, transcripts and proteins. Nucleic Acids Res. 2007;35. pmid:17130148
- 95. Apweiler R, Bairoch A, Wu CH, Barker WC, Boeckmann B, Ferro S, et al. UniProt: the Universal Protein knowledgebase. Nucleic Acids Res. 2004;32: 115D–119. pmid:14681372
- 96. Finn RD, Coggill P, Eberhardt RY, Eddy SR, Mistry J, Mitchell AL, et al. The Pfam protein families database: Towards a more sustainable future. Nucleic Acids Res. 2016;44: D279–D285. pmid:26673716
- 97. Ogata H, Goto S, Sato K, Fujibuchi W, Bono H, Kanehisa M. KEGG: Kyoto encyclopedia of genes and genomes. Nucleic Acids Research. 1999. pp. 29–34. pmid:9847135
- 98. Xie C, Mao X, Huang J, Ding Y, Wu J, Dong S, et al. KOBAS 2.0: A web server for annotation and identification of enriched pathways and diseases. Nucleic Acids Res. 2011;39. pmid:21715386
- 99. Finn RD, Clements J, Arndt W, Miller BL, Wheeler TJ, Schreiber F, et al. HMMER web server: 2015 Update. Nucleic Acids Res. 2015;43: W30–W38. pmid:25943547
- 100. Chen L, Chu C, Lu J, Kong X, Huang T, Cai YD. Gene ontology and KEGG pathway enrichment analysis of a drug target-based classification system. Wang J, editor. PLoS One. Wiley-Interscience; 2015;10: e0126492. pmid:25951454
- 101. Wang Y, Li Z, Lu Y, Hu G, Lin L, Zeng L, et al. Molecular characterization, tissue distribution and expression, and potential antiviral effects of TRIM32 in the common carp (Cyprinus carpio). Int J Mol Sci. 2016;17: 1693. pmid:27735853
- 102. Özdemir S, Altun S, Arslan H. Imidacloprid exposure cause the histopathological changes, activation of TNF-α iNOS, 8-OHdG biomarkers, and alteration of caspase 3, iNOS, CYP1A, MT1 gene expression levels in common carp (Cyprinus carpio L.). Toxicol Reports. 2018;5: 125–133. pmid:29321977
- 103. MacKenzie S, Ribas L, Pilarczyk M, Capdevila DM, Kadri S, Huntingford FA. Screening for coping style increases the power of gene expression studies. PLoS One. 2009;4. pmid:19390591
- 104. Zhang W, Jia Y, Ji X, Zhang R, Liang T, Du Q, et al. Optimal Reference Genes in Different Tissues, Gender, and Gonad of Yellow River Carp (Cyprinus carpio var) at Various Developmental Periods. Pakistan J Zool. 2016;48: 1615–1622.
- 105. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the2-DDCT method. Methods. 2001;25: 402–408. pmid:11846609