Characterization of TGFβ Signaling Components in Large Yellow Croaker (Larimichthys crocea) and Their Role in Growth and Body Shape Regulation

Abstract

The transforming growth factor β (TGFβ) signaling axis plays a pivotal role in orchestrating a wide array of biological functions, encompassing cellular growth, proliferation, and differentiation. The aim of the present study was to identify the members of TGFβ signaling pathway and their expression patterns in large yellow croaker (Larimichthys crocea) under different culture modes. TGFβ signaling pathway and their expression patterns in fish reared under two different culture modes: Group N (2400 fish in a 120 m³ cage) and Group V (168,000 fish in a 5600 m³ aquaculture vessel). After 120 days, we analyzed 15 fish from each group and found that Group V exhibited faster growth and a slender body shape compared to Group N. Bioinformatics analysis identified 48 TGFβ superfamily members in L. crocea, including 21 ligands, 10 receptors, and 3 Smads. mRNA expression levels indicated that these signaling molecules influence growth rate and body shape through five distinct ligand–receptor–R-Smad pathways, with the INHBB-, Nodal-, and GDF3-ACVR2A-ALK4-Smad2 axis playing a predominant role in regulating these traits.

1. Introduction

The global depletion of fishery resources has driven the transition from traditional extractive fishing to aquaculture. In this context, marine aquaculture has established itself as an essential method for sustainable seafood production [1]. China, the world’s largest producer in this sector, is leading this transformation. Among the various species farmed, the yellow croaker (Larimichthys crocea) stands out, whose production reaches unparalleled scale and throughput [2,3,4,5]. In 2024, production of this species reached 292.615 million tons, representing 13.53% of China’s annual marine fish production [6]. Given this, and with aquaculture expected to become the primary source of aquatic protein by 2050 [7], understanding the biological mechanisms that regulate the growth and morphology of commercially valuable species, such as L. crocea, is becoming increasingly crucial.

Growth-related traits are a priority in breeding programs, as they directly affect productivity. Furthermore, body shape is also a relevant factor in aquaculture, as it influences swimming behavior, feeding, predator avoidance, and, most importantly, consumer acceptance [8]. In this context, the market preference for large, slender yellow croaker specimens has made body shape an important economic characteristic, directly impacting the species’ commercial value [9]. However, factors such as climate change, disease outbreaks, and reduced genetic variability due to inbreeding have been compromising the species’ performance [10].

In this context, understanding the genetic factors involved in growth and body shape characteristics is essential, as these characteristics in L. crocea are complex quantitative traits controlled by multiple genes distributed throughout the genome. Furthermore, recent studies indicate that morphological variability is strongly determined by genetic factors [11,12]. Therefore, identifying the genes responsible for regulating growth and body shape is essential for the genetic improvement of the species. Research in animal models has demonstrated that the transforming growth factor beta (TGFβ) family plays a crucial role in coordinating a wide range of biological processes, including cell growth, proliferation, and differentiation [13,14,15]. Mechanistically, the TGFβ signaling pathway begins with the binding of ligands to membrane receptors, which promotes the activation of type I receptors through phosphorylation by type II receptors. Type I receptors then phosphorylate SMAD proteins, initiating intracellular signal transduction [16,17].

Regarding ligands, the TGFβ superfamily comprises a variety of proteins that bind to specific receptors on the cell surface, thus activating intracellular signaling pathways. These ligands include the TGFβs themselves (TGFB1-5), bone morphogenetic proteins (BMP2-16), growth and differentiation factors (GDF1-15), as well as Nodal, activins (INHBA and INHBB), and inhibins. In turn, receptors are transmembrane proteins classified into two main groups: type I receptors, which include activin-like kinases (ALK1 to ALK7), and type II receptors, which include TGFBR2, ACVR2, ACVR2B, AMHR2, and BMPR2 [18]. At the intracellular level, SMAD proteins, which act in signaling downstream of these receptors, are divided into three categories: R-SMADs (receptor-regulated, such as SMAD2 and SMAD3), Co-SMADs (co-regulatory, such as SMAD4), and I-SMADs (inhibitory, such as SMAD6 and SMAD7). This classification provides a framework for understanding the complex interactions within the TGFβ superfamily [19]. Generally, all cells, both in the embryo and in the adult organism, are capable of responding to these signals [20]. Furthermore, the number and type of TGFβ family members have been extensively evaluated in model organisms ranging from worms and flies to mammals [21,22], demonstrating that most cells respond to at least a subset of their ligands [19].

In addition to its role in growth and development, recent evidence from genome-wide association studies suggests that the TGFβ signaling pathway is also involved in body shape-related traits in L. crocea [9]. However, the specific components of this pathway and their regulatory mechanisms in this species remain largely unexplored. This knowledge gap is due, in part, to the recent completion of the complete genome sequencing and precise mapping of the species, carried out only in 2015 [23,24,25]. Currently, it is known that the TGFβ family in L. crocea comprises 24 ligands, 19 receptors and 3 SMAD proteins. Additionally, some members have even more specific subtypes, such as BMPR1BX1 (XM_010732488.3), BMPR1BX2 (XM_027275036.1) and BMPR1BX3 (XM_027275041.1). Despite this, functional and expression studies that clarify the role of these genes in the body development of the species are still scarce. Therefore, this study aimed to identify and characterize genes belonging to the TGFβ signaling pathway in the yellow croaker genome, as well as analyze their expression profiles. Understanding the role of these genes in regulating growth and body morphology is fundamental for advancing genetic improvement and the sustainability of aquaculture for this commercially valuable species.

2. Materials and Methods

2.1. Experimental Fish

In November 2023, L. crocea with a mean total length of 32.41 ± 1.16 cm and mean body weight of 350.4 ± 34.24 g (Group A) were allocated to two experimental rearing systems. Group N (n = 2400) was stocked in a 120 m³ net-pen cage (density: 20 fish/m³). Group V (n = 168,000) was reared in the “Conson No. 1” Aquaculture Vessel (total water volume: 5600 m³; density: 30 fish/m³) (Figure 1). After a 120-day feeding period, 15 sexually immature fish from each group were randomly selected for further detailed experimental analysis. The study characterized growth and body shape using parameters such as Body Weight (BW), Total Length (TL), Body Length (BL), Body Depth (BD), Caudal Peduncle Length (CPL), Caudal Peduncle Width (CPW), the ratio of Caudal Peduncle Length to Caudal Peduncle Width (CPL/CPW), the ratio of Body Length to Body Height (BL/BD), the ratio of Caudal Peduncle Length to Body Length (CPL/BL), and Fullness.

2.2. Sequence Analysis

The available protein sequences for all TGFβ ligands, receptors, and Smads from L. crocea on the NCBI (National Center for Biotechnology Information) database were downloaded and subjected to comparative analysis with sequences from other species using the BLAST (Basic Local Alignment Search Tool) programs. Multiple sequence alignments were conducted using the ClustalW Multiple Alignment program, accessible at http://www.ebi.ac.uk/clustalw/ (16 January 2025). Based on the amino acid sequences of ligands, receptors, and Smads, individual phylogenetic trees were constructed using the neighbor-joining (NJ) method implemented in MEGA version 7.0 [26]. The robustness of these trees was evaluated through 1000 bootstrap repetitions to ensure reliability.

2.3. mRNA Expression of TGFβ Signaling Pathway Genes

The muscle and liver of L. crocea of different sizes were peeled away carefully, flash-frozen in liquid nitrogen, and stored at −80 °C for subsequent total RNA extraction. Specific primers for TGFβ signaling pathway genes and housekeeping genes were designed based on known L. crocea sequences (

Table 1. Oligonucleotide primers used to amplify the TGFβ signaling pathway genes of L. crocea.

Gene Name/Another Name	Accession No.	Primer Sequence
β-actin/ACTB	FJ936563.1	5′-GACCTGACAGACTACCTCATG-3′
		5′-AGTTGAAGGTGGTCTCGTGGA-3′
TGFB1/TGFβ1	XM_027280465.1	5′-GGTGGTTTTCACAGCAGCAG-3′
		5′-GGGATGCTCGGGGAATAGTG-3′
TGFB2/TGFβ2	XM_010752699.3	5′-GCTGCCTACTGCTCCAGAAA-3′
		5′-TTGGCCTCATAGCCCTTTGG-3′
TGFB2/TGFβ3	XM_010733855.3	5′-AAGACGACCCCAAAGCCAAA-3′
		5′-ACTGATCTCCAGGCCCAGAT-3′
INHBA/Activin A	XM_010747963.3	5′-TCATCAGGAGGACAGAGCGA-3′
		5′-CGGGACTGGCTGTGTTACAT-3′
INHBB/Activin BNodal	XM_010746279.3	5′-GAGGATGGGAGGGTGGAGAT-3′
		5′-CAGATTCTGGTTGCCCTCGT-3′
Nodal	XM_010743837.3	5′-CTTTTGCGCTGCACACATCT-3′
		5′-CATGTAGGTTGGCAGGTGGT-3′
BMP2	XM_010733238.3	5′-CATGGCTTCATGGTGGAGGT-3′
		5′-TCACCCTGACGTGCCTACTA-3′
BMP3	XM_027273955.1	5′-TGCTCGTGCTGCTTTATGGA-3′
		5′-TGGTCGTCTTTCGTGAGGTG-3′
BMP4	XM_010735218.3	5′-CAGCTCTTGGATACGCGACT-3′
		5′-GTGTCTGGTTGAGGTGCAGA-3′
BMP5	XM_010756380.3	5′-CACCTTTGAACCGTGCTGTG-3′
		5′-GATCTCCCTGCGTTCGTGAT-3′
BMP6	XM_010741625.3	5′-GAAACTCCAGGCCAGACACA-3′
		5′-TCCACCACAATCCGACCAAG-3′
BMP7	XM_010755531.3	5′-CACGGCAGCAGAGTTTAGGA-3′
		5′-ACCTGGTAGACGCTGACTCT-3′
BMP8A	XM_019270763.2	5′-TTGAGGACCGCAGGAGAAAC-3′
		5′-TTCATCCATTCACGCCTGCT-3′
BMP10	XM_010738504.3	5′-GGGACAGTTGGATTCTCGCA-3′
		5′-GTTTGGTGGGTGTGACATGC-3′
GDF3	XM_019263909.2	5′-CTCTCTGGACCCTCATGGGA-3′
		5′-CGGGCCAATCAGAACCTCAT-3′
GDF5/BMP14	XM_010754689.3	5′-CACTTCAACGTCAGCTCCCT-3′
		5′-GGGATGAACCTCCTCCTCCT-3′
GDF6/BMP13	XM_010747111.3	5′-CCTGGATTACGAGGCGTACC-3′
		5′-CTGATGGGGCTGAGTTTGGT-3′
GDF8	NM_001303317.1	5′-AGAGTCCGCTCCCTGAAGAT-3′
		5′-CTCTGCTGAAGTGACAGCCA-3′
GDF9	XM_010754774.3	5′-ACAAGCAGATGGCGTTCAGA-3′
		5′-AATCGTACAAGGCGCAGTCA-3′
GDF15	XM_010733382.3	5′-GTGGACAGAATGGACAGCCA-3′
		5′-TGGGCTCCCTGTCTATACCC-3′
AMH/MIS	XM_010730161.3	5′-GGACTCTGCACGGTTTCTGA-3′
		5′-GCAGACTGTGGGAGGTCAAA-3′
ALK2/ACVR1	XM_019259847.2	5′-CACTGTGGAGCTGCCTGTTA-3′
		5′-CTCGTGACGTAAGCCTCTCC-3′
ALK3/BMPR1A	XM_010749871.3	5′-GAGAGAAGGTGGCCGTCAAA-3′
		5′-TGAGGAAGAGCTGCGTGAAG-3′
ALK4/ACVR1B	XM_010747783.3	5′-GGCGTGTTCCTGTTCCAGTA-3′
		5′-ACAAACAGGGGCAAACCAGA-3′
ALK6/BMPR1B	XM_010732488.3	5′-GACATCCCTCCCAACACGAG-3′
		5′-CCCTCCTGAGACACAACGTC-3′
ALK7/ACVR1C	XM_027290679.1	5′-CCTGCCATTGCTCACAGAGA-3′
		5′-TGTCGATGGTGTTGGTCCTG-3′
TBRII/TGFBR2	XM_010735343.3	5′-ATGAAGGTGAGACGGCTGTG-3′
		5′-GGCAGATGGAGGTGATCTCG-3′
ACVR2A/ACTR2A	XM_027287970.1	5′-TGAAGGCAAACGTCCTCTCC-3′
		5′-GGTCAGGTTCGACTTCAGCA-3′
ACVR2B/ACTR2B	XM_019276244.2	5′-CTGGCTGGACGACTTCAACT-3′
		5′-TGGAGGCCGGATTTTCACTC-3′
BMPR2	XM_010740233.3	5′-ACCACCCTGCACAATGAGAG-3′
		5′-CGTGTTCTTAGCCACACCCT-3′
AMHR2/MISR2	XM_027279584.1	5′-GTGGCTGATTTTGGCTGTGG-3′
		5′-GTTCTGGCGAGTCACTTGGA-3′
Smad1/MADR1	XM_010729469.3	5′-AATCGTGTTGGAGAGGCGTT-3′
		5′-CGGGTGTTCTCAATGGTGGA-3′
Smad2/MADR2	XM_019271768.2	5′-TACATCGGAGGGGAGGTGTT-3′
		5′-GTGCAGCAAACTCCTGGTTG-3′
Smad3/MADH3	XM_010735124.3	5′-CCACTACCAGAGGGTGGAGA-3′
		5′-AGGGATGGATGGGCTGTAGT-3′

) using Oligo 7.0. The primers required for this experiment were synthesized by Sangon Biotech (Shanghai, China) Co., Ltd. Total RNA was extracted with TRIzol™ Reagent (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA) and purified of genomic DNA using RNase-free DNase (Takara Bio, Kusatsu, Japan). RNA was reverse-transcribed using the Prime Script RT-PCR Kit (Takara). cDNA was then subjected to quantitative real-time PCR in a PikoReal™ Real-Time PCR System (96-well, Thermo Fisher Scientific, USA) with a 10 μL reaction mix including SYBR Green Master Mix, diluted cDNA, primers, and water. PCR cycling involved 94 °C for 5 min, followed by 30 cycles of 95 °C for 15 s, 60 °C for 30 s, and 72 °C for 1 min, with a final extension at 72 °C for 7 min. PCR products were visualized on an agarose gel, and dissociation curves were analyzed to ensure specificity. mRNA abundance was calculated using the Pfaffl method [27].

Statistical analysis of all data was conducted using GraphPad Prism 5.0 software, with one-way ANOVA employed for assessment. Post hoc multiple comparisons were performed using Tukey’s Honestly Significant Difference (HSD) test to evaluate differences between the means. The results were reported as mean ± SD, and a p-value of less than 0.05 was considered statistically significant.

3. Results

3.1. The Growth and Body Shape of L. crocea Under Two Aquiculture Modes

After a 120-day feeding period, significant differences were observed in body weight, total length, and body length among the three groups of L. crocea with undeveloped gonads, with Group V exhibiting a faster growth rate compared to Group N (

Table 2. The impact of net cages and vessel aquaculture on the growth and body shape of L. crocea. (Values with different superscript lowercase letters (a, b, c) denote statistically significant differences between groups (p < 0.05, [Tukey’s HSD]). Groups sharing the same letter are not significantly different (p ≥ 0.05). Data are presented as mean ± standard deviation (SD; n = 15 per group).

Indexes	Initial Data (in November 2023) for Group A	Finish Data for Group N	Finish Data for Group V
Weight (Wt, g)	350.4 ± 34.24 c	508 ± 29.02 b	535.33 ± 22.24 a
Total Length (TL, cm)	32.41 ± 1.16 c	35.31 ± 1.22 b	37.28 ± 0.92 a
Body Length (BL, cm)	28.29 ± 1.15 c	31.35 ± 0.93 b	33.78 ± 1.24 a
Body Depth (BD, cm)	7.9 ± 0.49 b	9.04 ± 0.35 a	9.06 ± 0.44 a
Caudal Peduncle Length (CPL, cm)	7.49 ± 0.55 c	8.07 ± 0.69 b	9.34 ± 0.42 a
Caudal Peduncle Width (CPW, cm)	2.03 ± 0.13 c	2.19 ± 0.15 b	2.3 ± 0.08 a
Caudal Peduncle Length/Caudal Peduncle Width (CPL/CPW)	3.7 ± 0.28 b	3.7 ± 0.32 b	4.07 ± 0.24 a
Body Length/Body Depth (BL/BD)	3.59 ± 0.21 b	3.47 ± 0.18 b	3.73 ± 0.17 a
Caudal Peduncle Length/Body Length (CPL/BL)	0.26 ± 0.01 b	0.26 ± 0.02 b	0.28 ± 0.01 a
Fullness (g/cm3)	1.55 ± 0.15 a	1.66 ± 0.17 a	1.4 ± 0.15 b

). Additionally, there were significant differences in the ratios of caudal peduncle length to caudal peduncle width (CPL/CPW), body length to body depth (BL/BD), caudal peduncle length to body length (CPL/BL), as well as fullness, between Group V and Group N (Table 2). These indicators collectively demonstrate that large yellow croaker cultured in boats possess a longer body compared to those reared in cages. To provide readers with a more intuitive understanding of the differences in these data, we selected one fish from each of the three groups as representatives (Figure 2).

3.2. Phylogenetic Analysis of Ligand Sequences

The variety and quantity of ligands are illustrated in Figure 3 and Figure 4. The NCBI contains 24 ligand entries, which are categorized into 21 classes. Notably, GDF15 from Larimichthys crocea (XM_010733382.3) does not cluster with other GDF species in the phylogenetic tree but shows a closer relationship to GDF8 (Figure 4). Except for AMH/MIS (Anti-Mullerian Hormone/Muellerian-inhibiting substance), which has four subtypes, each of the remaining 20 classes of ligands has one subtype. In Larimichthys crocea, the definitive ligand types include GDF8 (Growth Differentiation Factor 8), GDF6/BMP13 (Growth Differentiation Factor 6/Bone Morphogenetic Protein 13), GDF5/BMP14 (Growth Differentiation Factor 5/Bone Morphogenetic Protein 14), BMP3 (Bone Morphogenetic Protein 3), GDF9 (Growth Differentiation Factor 9), AMH, BMP10 (Bone Morphogenetic Protein 10), BMP4/BMP2B (Bone Morphogenetic Protein 4/Bone Morphogenetic Protein 2B), BMP2/BMP2A (Bone Morphogenetic Protein 2/Bone Morphogenetic Protein 2A), GDF3 (Growth Differentiation Factor 3), BMP8/BMP8B (Bone Morphogenetic Protein 8/Bone Morphogenetic Protein 8B), BMP5 (Bone Morphogenetic Protein 5), BMP7 (Bone Morphogenetic Protein 7), BMP6 (Bone Morphogenetic Protein 6), NDR2 (Nodal-related 2), INHBB/Activin B (Inhibin Subunit Beta B/Activin Beta-B Chain), INHBA/Activin A (Inhibin Subunit Beta A/Activin Beta-A Chain), TGFB1 (Transforming Growth Factor Beta 1), TGFB2 (Transforming Growth Factor Beta 2), and TGFB3 (Transforming Growth Factor Beta 3).

3.3. Phylogenetic Analysis of Receptors

Upon examining the receptor phylogenetic tree in Figure 5, it is evident that the 19 nucleotide sequence entries for receptors in Larimichthys crocea available in Genebank can be categorized into 10 distinct groups. These include five type I receptors: ALk3/BMPR1A (Activin Receptor-Like Kinase 3/Bone Morphogenetic Protein Receptor Type 1A), ALK6/BMPR1B (Activin Receptor-Like Kinase 6/Bone Morphogenetic Protein Receptor Type 1B), ALK2/ACVR1 (Activin Receptor-Like Kinase 2/Activin A Receptor Type 1), ALK4/ACVR1B (Activin Receptor-Like Kinase 4/Activin A Receptor Type 1B), and ALK7/ACVR1C (Activin Receptor-Like Kinase 7/Activin A Receptor Type 1C). Additionally, there are five type II receptors: ACVR2A/ACTRIIA (Activin A Receptor Type 2A/Activin Receptor Type-2A), ACVR2B/ACTRIIB (Activin A Receptor Type 2B/Activin Receptor Type-2B), TGFBR2/TBRII (Transforming Growth Factor Beta Receptor 2), BMPR2 (Bone Morphogenetic Protein Receptor Type 2), and AMHR2/MISR2 (Anti-Mullerian Hormone Receptor Type 2/Muellerian-inhibiting substance Receptor Type 2). Notably, within some of these categories, different subtypes exist in Larimichthys crocea. For instance, AMHR2 has two subtypes (AMHR2 × 1 and AMHR2 × 2), ACVR2A has five subtypes (ACVR2AX1, ACVR2AX2, ACVR2AX3, ACVR2AX4, and ACVR2AX5), BMPR1A has two subtypes (BMPR1AX1 and BMPR1AX2), and BMPR1B has four subtypes (BMPR1BX1, BMPR1BX2, BMPR1BX3, and BMPR1BX4).

3.4. Phylogenetic Analysis of R-Smads

In the phylogenetic tree depicted in Figure 6, the nucleotide acid sequences of each type of R-Smad grouped together. The five Smad genes identified in L. crocea were categorized into three classes: Smad1, Smad2, and Smad3. Among them, Smad2 comprises three distinct subtypes, namely Smad2 × 1, Smad2 × 2, and Smad2 × 3. In large yellow croaker, Smad5 and Smad8 were not identified, despite their belonging to the R-Smads family and primarily responding to BMP signals.

3.5. mRNA Levels of Ligands

The expression profiles of ligands in the liver of large yellow croaker are illustrated in Figure 7. Specifically, the expression levels of TGFB2, BMP6, and GDF9 were significantly upregulated in both Group N and Group V. In contrast, ligands such as TGFB1, BMP3, BMP4, BMP10, and GDF5 exhibited decreased expression in Group N but increased expression in Group V. Additionally, ligands including TGFB3, INHBA, Nodal, BMP8A, and GDF8 showed no significant change in expression in Group N but were significantly upregulated in Group V. Conversely, GDF3, GDF15, and AMH displayed no significant difference in expression in Group V but were significantly downregulated in Group N. Notably, BMP5, BMP7, and GDF6 exhibited no significant differences in expression levels between Group N and Group V. However, in muscle tissue (Figure 8), we observed a completely different pattern of ligand expression, with one notable exception: GDF9. For instance, the expression level of TGFB2 showed no significant difference in Group N but was significantly upregulated in Group V. Interestingly, BMP4, BMP5, and AMH were the ligands that exhibited significant increases in expression levels in both Group N and Group V. These findings suggest that the expression of ligands in muscle tissue may be regulated by different mechanisms compared to those in liver tissue.

3.6. mRNA Levels of Receptors

The expression levels of ligands in the liver of large yellow croaker are depicted in Figure 9. Specifically, ALK2 and AMHR2 exhibited no significant differences in expression between Group N and Group V. In contrast, the expression of ALK3 and ALK7 decreased in Group N but increased in Group V. Both ALK4 and ACVR2A showed significant decreases in expression in both groups. Among the remaining three receptors (ALK6, TGFBR2, and ACVR2B), there were no significant differences in expression in Group N, but significant increases were observed in Group V. Notably, except for ALK4, the mRNA expression trends of these receptors in muscle tissue differed from those in liver tissue (Figure 9). In muscle, ALK6 showed no significant difference in expression between Group N and Group V. There were no ligands that exhibited decreased expression in Group N and increased expression in Group V. However, the expression of ALK4 significantly decreased in both groups in muscle. Conversely, the expression of ALK7, TGFBR2, and ACVR2B significantly increased in both Group N and Group V in muscle (Figure 10).

3.7. mRNA Levels of R-Smads

Figure 11 and Figure 12 respectively reveal the expression patterns of R-Smads in the liver and muscle tissues of large yellow croaker. Specifically, in these two tissues, no significant changes in Smad3 expression levels were observed between the experimental groups and the control group, nor were there any notable differences in Smad3 expression between the two experimental groups. As for Smad1 and Smad2, their expression patterns in Group N exhibited distinct trends in liver and muscle tissues. In the liver, the expression of these two genes differed from that of the control group, while no similar differences were found in muscle tissue; conversely, those that showed significant differences compared to the control group in muscle did not display differences in the liver. This result indicates that the expression of Smad1 and Smad2 in Group N is regulated in a tissue-specific manner. In Group V, however, the expression of Smad1 and Smad2 appears to be unaffected by the type of tissue used in the experiment, i.e., their expression patterns remained consistent in both tissues, without showing inter-tissue variability. In summary, the expression of Smad3 among R-Smads in large yellow croaker is relatively stable across different treatments and tissues, while the expression of Smad1 and Smad2 exhibits tissue-specific variation patterns under different treatments.

4. Discussion

In contrast to the slender build of wild large yellow croaker, the cultured variety tends to have a more rounded appearance [28]. This morphological difference is relevant, as fish with a more slender body shape demonstrate superior swimming abilities, greater foraging efficiency, and a greater ability to escape predators [29]. Furthermore, consumer preference for individuals with a more slender appearance significantly influences the market value of the species [10]. In this context, a 120-day culture experiment revealed that yellow croaker raised on an industrial vessel exhibited faster growth and a more slender body shape compared to those raised in conventional cages. This faster growth is attributed, in part, to the vessel’s advanced aquaculture system, which performs forced and continuous exchanges between tank water and external seawater, promoting constant renewal and ensuring better environmental conditions. Furthermore, the vessel is equipped with a deep-water intake device, ensuring the supply of water with ideal temperature and salinity for the species’ cultivation, which helps maintain Larimichthys crocea in optimal growth conditions [30]. The authors suggest that the active swimming behavior and schooling tendency of large yellow croakers cultivated in the onboard system, characteristics similar to those of wild populations, are determining factors in the development of a more slender body shape.

Based on numerous studies addressing the specificity of interactions between ligands, receptors, and R-Smads of the TGFβ superfamily [22,31,32,33,34], as well as the author’s previous experience characterizing TGFβ signaling pathways in marine species such as the sea cucumber (Apostichopus japonicus) [35] and the Pacific oyster (Crassostrea gigas) [36], it was possible to identify and analyze corresponding genes in the genome of the yellow croaker. This analysis was based on both records of related nucleotide sequences (Figure 3, Figure 4, Figure 5 and Figure 6) and the mRNA expression levels of these (Figure 7, Figure 8, Figure 9, Figure 10, Figure 11 and Figure 12) in the genome of L.crocea. The ligand–receptor–R-Smad correspondence for L. crocea is summarised in

Table 3. TGFβ superfamily ligand–receptor–R-Smad specificity.

Subfamily	Ligand	Receptor I	Receptor II	R-Smad
TGFβ/activin/Nodal	INHBA/Activin A	ALK4/ACVR1B	ACVR2A/ACTRIIA and ACVR2B/ACTRIIB	Smad/MADR2, 3
	INHBB/Activin B	ALK4/ACVR1B and ALK7/ACVR1C	ACVR2A/ACTRIIA and ACVR2B/ACTRIIB	Smad/MADR2, 3
	Nodal	ALK4/ACVR1B and ALK7/ACVR1C	ACVR2A/ACTRIIA and ACVR2B/ACTRIIB	Smad/MADR2, 3
BMP/GDF/AMH	BMP2	ALK3/BMPR1A/BMPR1A and ALK6/BMPR1B	ACVR2A/ACTRIIA, ACVR2B/ACTRIIB and BMPR2	Smad1/MADR1
	BMP4	ALK3/BMPR1A and ALK6/BMPR1B	ACVR2A/ACTRIIA, ACVR2B/ACTRIIB and BMPR2	Smad1/MADR1
	BMP5	ALK2/ACVR1, ALK3/BMPR1A and ALK6/BMPR1B	ACVR2A/ACTRIIA, ACVR2B/ACTRIIB and BMPR2	Smad1/MADR1
	BMP6	ALK2/ACVR1, ALK3/BMPR1A and ALK6/BMPR1B	ACVR2A/ACTRIIA, ACVR2B/ACTRIIB and BMPR2	Smad1/MADR1
	BMP7	ALK2/ACVR1, ALK3/BMPR1A and ALK6/BMPR1B	ACVR2A/ACTRIIA, ACVR2B/ACTRIIB and BMPR2	Smad1/MADR1
	BMP8A	ALK2/ACVR1, ALK3/BMPR1A and ALK6/BMPR1B	ACVR2A/ACTRIIA, ACVR2B/ACTRIIB and BMPR2	Smad1/MADR1
	GDF3	ALK4/ACVR1B and ALK7/ACVR1C	ACVR2A/ACTRIIA and ACVR2B/ACTRIIB	Smad/MADR2, 3
	GDF5/BMP14	ALK3/BMPR1A and ALK6/BMPR1B	ACVR2A/ACTRIIA, ACVR2B/ACTRIIB and BMPR2	Smad1/MADR1
	GDF6/BMP13	ALK3/BMPR1A and ALK6/BMPR1B	ACVR2A/ACTRIIA, ACVR2B/ACTRIIB and BMPR2	Smad1/MADR1
	GDF8	ALK4/ACVR1B	ACVR2B/ACTRIIB	Smad/MADR2, 3
	GDF9	ALK4/ACVR1B	BMPR2	Smad1/MADR1
	AMH/MIS	ALK2/ACVR1 and ALK3/BMPR1A	AMHR2/MISR2	Smad1/MADR1

. In the TGFβ/activin/Nodal subfamily, the following components were found: ligands (INHBA, INHBB, and Nodal), type I receptors (ALK4 and ALK7), type II receptors (ACVR2A and ACVR2B), and R-Smads (Smad2 and Smad23). On the other hand, analysis of the BMP/GDF/AMH subfamily revealed a greater diversity of components, including ligands (BMP2, BMP4, BMP5, BMP6, BMP7, BMP8A, GDF3, GDF5, GDF6, GDF8, GDF9 and AMH), type I receptors (ALK2, ALK3, ALK4, ALK6 and ALK7), type II receptors (ACVR2A, ACVR2B, BMPR2 and AMHR2) and R-Smads (Smad1, Smad2 and Smad3). In Table 3, we list the family members that present complete signaling pathways, with all essential elements identified. Additionally, the nucleotide and amino acid sequences corresponding to the TGFB1, TGFB2, and TGFB3 genes were detected in the L. crocea genome, and their mRNA expression levels were evaluated by RT-PCR under different culture conditions. However, when comparing the receptor types of TGFB1, TGFB2, and TGFB3 reported in the current literature [32,37,38,39,40,41], it was observed that none of them are present in the yellow croaker genome. Consequently, due to their incomplete signaling pathways, the TGFB1, TGFB2, and TGFB3 members were not included in Table 3.

Based on the promotion and inhibition effects observed in the experimental groups involving ligands, receptors, and R-Smads (Figure 7, Figure 8, Figure 9, Figure 10, Figure 11 and Figure 12), viable signaling pathways responsible for regulating growth and body shape in the tissues of the greater yellow croaker were identified. These pathways are detailed in

Table 4. TGFβ superfamily ligand–receptor–R-Smad specificity in liver.

Subfamily	Ligand	Receptor I	Receptor II	R-Smad
TGFβ/activin/Nodal	INHBA/Activin A	ALK4/ACVR1B	ACVR2A/ACTRIIA and ACVR2B/ACTRIIB	Smad2/MADR2
	INHBB/ActivinB	ALK4/ACVR1B and ALK7/ACVR1C	ACVR2A/ACTRIIA and ACVR2B/ACTRIIB	Smad2/MADR2
	Nodal	ALK4/ACVR1B and ALK7/ACVR1C	ACVR2A/ACTRIIA and ACVR2B/ACTRIIB	Smad2/MADR2
BMP/GDF/MIS	BMP2	ALK3/BMPR1A and ALK6/BMPR1B	ACVR2A/ACTRIIA, ACVR2B/ACTRIIB and BMPR2	Smad1/MADR1
	BMP4	ALK3/BMPR1A and ALK6/BMPR1B	ACVR2A/ACTRIIA, ACVR2B/ACTRIIB and BMPR2	Smad1/MADR1
	BMP6	ALK3/BMPR1A and ALK6/BMPR1B	ACVR2A/ACTRIIA, ACVR2B/ACTRIIB and BMPR2	Smad1/MADR1
	BMP8A	ALK3/BMPR1A and ALK6/BMPR1B	ACVR2A/ACTRIIA, ACVR2B/ACTRIIB and BMPR2	Smad1/MADR1
	GDF3	ALK4/ACVR1B and ALK7/ACVR1C	ACVR2A/ACTRIIA and ACVR2B/ACTRIIB	Smad/MADR2, 3
	GDF5/BMP14	ALK3/BMPR1A and ALK6/BMPR1B	ACVR2A/ACTRIIA, ACVR2B/ACTRIIB and BMPR2	Smad1/MADR1
	GDF8	ALK4/ACVR1B	ACTRIIB and BMPR2	Smad2/MADR2
	GDF9	ALK4/ACVR1B	BMPR2	Smad1/MADR1

(liver) and

Table 5. TGFβ superfamily ligand–receptor–R-Smad specificity in muscle.

Subfamily	Ligand	Receptor I	Receptor II	R-Smad
TGFβ/activin/Nodal	INHBA/Activin A	ALK4/ACVR1B	ACVR2A/ACTRIIA and ACVR2B/ACTRIIB	Smad2/MADR2
	INHBB/ActivinB	ALK4/ACVR1B and ALK7/ACVR1C	ACVR2A/ACTRIIA and ACVR2B/ACTRIIB	Smad2/MADR2
	Nodal	ALK4/ACVR1B and ALK7/ACVR1C	ACVR2A/ACTRIIA and ACVR2B/ACTRIIB	Smad2/MADR2
BMP/GDF/MIS	GDF3	ALK4/ACVR1B and ALK7/ACVR1C	ACVR2A/ACTRIIA and ACVR2B/ACTRIIB	Smad2/MADR2
	GDF8	ALK4/ACVR1B	ACVR2B/ACTRIIB	Smad2/MADR2

(muscle), encompassing a total of 11 pathways in liver and 5 pathways in muscle, all belonging to two subfamilies of the TGFβ superfamily. The reason for the lower number of viable signaling pathways in muscle compared to liver can be explained by differences in Smad1 mRNA expression levels. Specifically, the different rearing environments did not cause significant changes in Smad1 expression in muscle, resulting in a reduced number of viable pathways in this tissue. It is noteworthy that Smad1 is the only R-Smad functionally associated with multiple BMPs identified in the yellow croaker, playing a central role in complex formation with Co-Smad (Smad4) for subsequent translocation to the nucleus and regulation of gene expression [42,43]. Interestingly, the five growth and body size regulation pathways detected in the muscle tissue of the yellow croaker are also present in the liver, suggesting a possible functional sharing between these tissues. However, it remains unclear whether these pathways maintain a similar regulatory pattern in other organs, such as the heart, spleen, kidney, intestine, gonads, and brain. This hypothesis requires further experimental investigation. In muscle group V, a suppression in the expression levels of components of the following three signaling pathways was observed: INHBB-ACVR2A-ALK4-Smad2, Nodal-ACVR2A-ALK4-Smad2, and GDF3-ACVR2A-ALK4-Smad2 (Figure 8, Figure 10 and Figure 12). The INHBB, Nodal, and GDF3 ligands bind to specific receptors to initiate signaling. The reduction in the expression of these ligands may compromise the efficiency of signal transduction [44,45]. The ACVR2A and ALK4 receptors are able to interact with these three ligands, thus activating signal transduction downstream of Smad2 [46]. The decrease in the levels of ACVR2A, ALK4, and Smad2 indicates a possible inhibition of the TGF-β signaling pathway in these cases [47]. On the other hand, the receptors ACVR2B and ALK7, which can also bind to the same ligands, showed elevated levels of expression. This suggests two possibilities: they may be transmitting signals from other, yet unidentified ligands in this context, or they may be acting compensatorily to compensate for the reduced function of ALK4 and ACVR2A under specific conditions [32,48].

5. Conclusions

In summary, this study identified the presence of 34 members of the TGFβ superfamily distributed across two subfamilies in the genome of yellow croaker (L. crocea). These findings suggest that these signaling molecules likely influence the regulation of growth rate and body shape in this species, acting through five distinct ligand–receptor–R-Smad signaling pathways. Among these pathways, the INHBB-, Nodal-, and GDF3-ACVR2A-ALK4-Smad2 signaling axis stands out, which was identified as playing a central role in modulating the aforementioned physiological traits and is, therefore, a potential key target for future interventions in genetic improvement programs and aquaculture management for this species.