Molecular Mechanisms of Growth Differences in Gymnocypris przewalskii and Gymnocypris eckloni through a Comparative Transcriptome Perspective

Abstract

Genetic composition plays a crucial role in the growth rate of species, and transcriptomics provides a potent tool for studying genetic aspects of growth. We explored the growth rates and transcriptomes of the Cyprinids G. przewalskii (GP) and G. eckloni (GE). A total of 500 individuals of G. przewalskii and G. eckloni, matched in size, were separately cultured for 9 months in six cement tanks (each group with three replicates). Growth indices were measured, revealing that the growth rate of GE was greater than that of GP (p < 0.05), while there was no significant difference in survival rates (p > 0.05). Simultaneously, we conducted RNA-Seq on the muscles of both GP and GE. The results indicated a significant difference of gene expression between GP and GE, identifying 5574 differentially expressed genes (DEGs). Quantitative real-time reverse transcription–polymerase chain reaction of 10 DEGs demonstrated consistency in expression profiles with the results from the RNA-Seq analysis. The DEGs were significantly enriched in glycolysis/gluconeogenesis (ko00010), arachidonic acid formation (ko00061), arginine biosynthesis (ko00220), and the MAPK (ko04013), PI3K-Akt (ko04151), mTOR (ko04150), and TGF-β (ko04350) signal pathways, as revealed by Gene Ontology (GO) enrichment and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis. This study also identified some growth-related DEGs, such as IGF2, Noggin, Decorin and others. Notably, the low expression of IGF2 may be a factor contributing to the slower growth of GP than GE.

1. Introduction

The growth rate, an important commercial trait in aquaculture, is influenced by various factors, with genetic composition playing a pivotal role [1]. The rapid advancement of next-generation sequencing technologies has promoted the application of transcriptomics in the aquaculture sector [2,3]. The transcriptome provides a blueprint of actively transcribed genes in an organism, unraveling molecular mechanisms that potentially govern growth performance [4,5]. For instance, a comparative transcriptome analysis of grass carp (Ctenopharyngodon idella) muscle tissue between fast- and slow-growing family groups revealed differentially expressed genes (DEGs) involved in the growth hormone (GH)/insulin-like growth factor (IGF) axis, calcium metabolism, protein and glycogen synthesis, oxygen transport, cytoskeletal and myofibrillar components. These DEGs might be key genes contributing to the growth differences between fast- and slow-growing fish [4]. Similarly, Lin et al. [5] conducted a comparative transcriptome analysis of mixed tissues (brain, liver, and muscle tissue samples) of black porgy (Acanthopagrus schlegelii) with varying growth rates, identifying key genes such as IGF-1, IGF-2, myosin light chain 3-like (Myl3), and growth hormone receptor type 1 (GHR1) involved in the muscle growth process. Thus, the transcriptome is emerging as a potent tool for elucidating growth differences in economically important aquatic organisms [6,7].

The growth trait is one of the most important economic traits of fish, contributing significantly to the development of the aquaculture industry [8]. The GH/IGF axis primarily regulates muscle growth [9,10]. GH and IGFs indirectly stimulate muscle growth through processes such as cell proliferation, differentiation, myogenesis, protein synthesis, and breakdown. Moreover, they directly stimulate muscle growth through hyperplasia and hypertrophy [10,11]. Further studies suggested that IGFs promote growth by activating MAPK/ERK, PI3K/AKT/TOR, and other signaling pathways [9,12]. Existing research underscores the pivotal role of the GH/IGF growth axis in fish muscle growth, as observed in gilthead sea bream (Sparus aurata) [10], fine flounder (Paralichthys adspersus) [13], and rainbow trout (Oncorhynchus mykiss) [14].

Gymnocypris eckloni (GE), a species representative of highly specialized schizothoracine fishes, is primarily found in the Yellow River system [15,16]. This omnivorous fish, residing in the middle layer of water, typically weighs 2–3 kg, or even 5–6 kg [17]. As aquaculture animals, G. eckloni has significance because of its tender and delicious meat, nutrient-rich content, and high economic value and potential for development [17,18]. Another species, Gymnocypris przewalskii (GP), also belonging to the subfamily Schizothoracinae, is a unique indigenous fish and aquatic biological resource in the Qinghai Tibet Plateau (elevation 3200 m) with ecological and economic value [16,19,20]. Despite its excellent characteristics of resistance to salt–alkaline conditions and cold temperatures and low oxygen tolerance, GP exhibits slow growth [21,22]. Both Gymnocypris species are vital economic cold-water fishes in western China. Notably, in our prior observations, GE displayed faster growth than GP in a cold-water fish breeding station in Yue-Xi City, China, especially in the juvenile stage. Surprisingly, no research has been reported on the growth differences between these two Gymnocypris species and their respective growth regulation mechanisms, despite their economic importance.

To comprehend the growth differences between juvenile GE and GP, it is imperative to conduct research. This study investigated the growth performance of these two species over a 9-month period. Additionally, a comparative analysis of the transcriptomes of GE and GP will provide insights into the growth mechanisms of Gymnocypris species. These findings will contribute to understanding the growth mechanisms of Gymnocypris species, offering insights for the advancement in molecular-marker-assisted breeding.

2. Materials and Methods

2.1. Fish

A total of 500 individuals of GP and GE, aged 3 days, were cultured separately in net cages in filtered water (XiDe ZhengYuan Fish Farm in Liangshan Prefecture, Sichuan Province, China). Uniform conditions of the freshwater, temperature (13 °C), light (on at 6:00 and off at 18:00), and same specific amount of feed (Hirear, a special expanded formula feed for fish fry; nutrient compositions’ content was as follows: crude protein ≥ 38.0%, crude fat ≥ 6.0%, crude fiber ≤ 4.0%, crude ash ≤ 17.0%, moisture ≤ 11.0%, phosphorus ≥ 1.2%, lysine ≥ 1.8%; feed twice a day) were maintained for 9 months, with growth indices (weight, length, and survival) measured every 3 months.

2.2. Sampling

For sampling purposes, 3 juveniles were randomly selected from the GP group (8.17 ± 0.61 cm; 7.73 ± 1.17 g) and the GE group (12.28 ± 0.99 cm; 18.93 ± 1.19 g). Muscle samples from the same part of juvenile fishes were collected on ice, the surface dried with absorbent paper, frozen in liquid nitrogen, and stored at −80 °C for a subsequent analysis. All experimental procedures with animals adhered to the guidelines and received approval from the Animal Research and Ethics Committees of Xi Chang University.

2.3. RNA Extraction, Library Construction, and High-Throughput Sequencing

Total RNA was extracted from the muscle tissues using a Trizol reagent (Takara Bio, Otsu, Japan). RNA purity and quantity were assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Shanghai, China). RNA with total RNA ≥ 1 μg, OD260/280, in the range of 1.8–2.2, and Agilent 2100 RIN ≥ 7 were used for library construction. The integrity of the RNA was confirmed with the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). RNA libraries were constructed using the VAHTS Universal V6 RNA-Seq Library Prep Kit (Illumina, Shanghai, China) according to the manufacturer’s instructions. The libraries were sequenced on the Illumina NovaSeq 6000 platform, generating 150-bp-long paired-end reads. OE Biotech Co., Ltd. (Shanghai, China) conducted the transcriptome sequencing and analysis.

2.4. RNA Sequencing Analysis

Raw RNA data in FASTQ format were processed using fastp [23] (Version 20.1). Clean readings were assembled into expressed sequence tag clusters (contigs) after removing adaptor and poor-quality sequences. Trinity [24] (Version 2.4) was employed for de novo assembly using the paired-end approach. The longest transcript was selected as a unigene based on its length and similarity to other analyses. The Swiss-Prot database, using the diamond tool with an e-value < 1 × 10⁻⁵, was used for the functional annotation of the unigenes. The proteins with the highest number of hits to the unigenes were functionally annotated. Unigenes were then mapped to annotate their potential involvement in various metabolic pathways against the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. Following annotation, the number of reads aligned to a unigene in each sample was determined using bowtie2 software [25] (Version 2.3.3.1), and the expression level of each unigene (FPKM) was calculated using eXpress software [26] (Version 1.5.1). DESeq2 software [27] (Version 1.20) identified DEGs between groups using the negative binomial distribution test (NB). The default filter conditions for DEGs were foldChange > 2 or foldChange < 0.5 and p < 0.05. R (https://cloud.oebiotech.cn/task/category/pipeline/, accessed on 10 September 2023) was used to perform a hierarchical cluster analysis of DEGs to show the unigene expression pattern in the various experimental groups and samples. R (software of clusterProfiler and ggplot2) was also used to carry out the KEGG pathway enrichment analysis of DEGs using the hypergeometric distribution. The oebiotech platform (https://cloud.oebiotech.cn/task/category/pipeline/, accessed on 10 September 2023, Shanghai, China) was utilized to create column and bubble diagrams of the significant enrichment pathway.

2.5. Quantitative Real-Time PCR (qRT–PCR) Analysis

The Trizol reagent was employed to extract total RNA from muscle samples following the manufacturer’s instructions. Subsequently, RNA integrity was assessed through 1% agarose gel electrophoresis. cDNA synthesis was performed using the PrimeScript^® RT Reagent Kit with gDNA Eraser (Takara Bio Inc., Shiga, Japan), and the resulting cDNA served as a template for qRT–PCR. The list of primers used in this study is presented in

Table 1. Primers used for RT–PCR assay for the validation of DEGs.

Gene Name	Primer Name	Primer (5′-3′)
FBP1	FBP1 F	AGCATGGTATGATGAAGCACAGTAC
	FBP1 R	GCAGTTGAAGACGGTCCTTAGGT
GAPDH	GAPDH F	GGCGTCGTTCGTGGTCTTCA
	GAPDH R	CGTCCTCGGATCAGCAGAACTC
PFKl	PFKl F	CTCCTCATCCTCCTCATCCTCATCC
	PFKl R	CGAAGCACGAGCACGACATCTATT
IGF2	IGF2 F	GTGCTCACACGCTCTTCCAGTT
	IGF2 R	CGTTCGGCGGCTTCTTTGTTCT
PKM	PKM F	ATGATGAGGAGGAGCGTGAAGAGT
	PKM R	CGAACAGACAGGCGTCCAGAAC
ODC	ODC F	TGGACCAATCTCAAGTTCAGGAAGT
	ODC R	GCAGAGAACCAGAATCACATCACAG
ARG2	ARG2 F	GGAGACCTGACCTTCAAGCATCTG
	ARG2 R	AGCGTGACCTTCTACTGAACCAATC
GADD45G	GADD45G F	CAAGCCTCCTGTGCCACTCAA
	GADD45G R	AGACTCTTCTTCGCCTTCAATCTCA
LOG5	LOG5 F	AAGGCTAATGGCATCGGTGGAA
	LOG5 R	TCCTCGTCCTCCTGAACTGTCTC
LTBP1	LTBP1 F	CAGGATTCAAGGACTCTCAGGATGG
	LTBP1 R	CCTCTGGTGTGACTGGTGGTGTA
EF1α	EF1α F	GTATTACCATTGACATTGC
	EF1α R	CTGAGAAGTACCAGTGAT

Note: FBP1 is fructose-1,6-bisphosphatase 1; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; PFK1, ATP-dependent 6-phosphofructokinase; IGF2, insulin-like growth factor II; PKM, pyruvate kinase; ODC, ornithine decarboxylase; ARG2, arginase 2; GADD45G, growth-arrest- and DNA-damage-inducible protein; LOG5, polyunsaturated fatty acid 5-lipoxygenase; LTBP1, latent-transforming growth factor beta-binding protein 1; EF1α, elongation factor 1 alpha.

. SYBR^®-Premix Ex Taq™ was utilized for qRT–PCR amplifications on an ABI7500 Real-Time PCR Detection System (Applied Biosystems, Waltham, MA, USA) with a 20 µL reaction mixture. The comparative threshold cycle approach (2^−ΔΔCT) analysis of target gene expression levels employed elongation factor 1 alpha (EF1α) as the reference gene [28]. Each experiment included three biological replicates.

2.6. Statistical Analyses

The body length, weight, and survival rates of 15 random-choice juvenile fish in two Gymnocypris groups were measured every 3 months. The growth metrics were calculated using the following formulas:

Survival rate (%) = Final number of animals/Initial number of animals,

Weight again rate (WGR%) = (terminal weight − initial weight)/initial weight.

3. Results

3.1. Growth Trait Analysis

Significant differences were observed in the terminal length, weight, survival rate, and weight rate between the GE and GP groups (p < 0.05;

Table 2. Growth trait measurements of G. przewalskii (GP) and G. eckloni (GE).

Day	Group	Length (cm)	Weight (g)	Weight Gain Rate (%, Month)	Survival Rate (%)
3	GP	1.32 ± 0.14	0.40 ± 0.05	-	-
	CE	1.41 ± 0.17	0.41 ± 0.05	-	-
90	GP	4.81 ± 0.49	1.71 ± 0.51	115.95 ± 32.30 b	92.80
	CE	5.92 ± 0.72	2.80 ± 0.61	228.90 ± 51.30 a	94.30
180	GP	7.60 ± 0.83 b	6.82 ± 1.43 b	107.28 ± 31.16 b	94.29
	CE	11.58 ± 1.35 a	17.83 ± 3.39 a	182.61 ± 33.99 a	95.86
270	GP	11.18 ± 1.56 b	16.33 ± 3.17 b	45.79 ± 10.07	95.77
	CE	16.94 ± 1.95 a	36.39 ± 8.40 a	34.42 ± 7.27	97.12
3–270	GP	-	-	425.44 ± 65.30 b	83.80
	CE	-	-	964.45 ± 150.53 a	87.80

Values (expressed as mean ± SD) with different letters in the same list on same day are significantly different from each other (p < 0.05).

). At 90 days, no significant differences were found in the length, weight, and survival rate (p > 0.05), while GE exhibited faster weight growth than GP (p < 0.05). At 180 days, no significant differences were observed in the survival rate (p > 0.05), but GE showed faster growth in the length, weight, and weight rate compared to GP (p < 0.05). Similarly, at 270 days, no significant differences in the survival rate and weight rate were observed (p > 0.05), but GE demonstrated faster growth in length and weight than GP (p < 0.05).

3.2. Analysis of DEGs between Groups

All gene expression data were used for hierarchical clustering to illustrate relationships between samples from the GP and GE groups. The cluster analysis indicated differences in gene expression between the two groups, with samples of GP and GE clustering separately (Figure 1a). DEGs between GP and GE were identified based on p < 0.05 and absolute log2 (ratio) ≥ 1, revealing 2985 upregulated genes (p < 0.05) and 2589 downregulated genes (p < 0.05) in GP vs. GE (Figure 1b).

The hierarchical cluster analysis of DEGs was conducted using Trinity [24] to illustrate the expression patterns of unigenes across various groups and samples. The results indicated a significant difference in gene expression patterns between the GP and GE groups (p < 0.05; Figure 2).

The top 30 entries from the Gene Ontology (GO) enrichment analysis, with the top 10 entries in each category ordered by the −log10 p-value of GO entries and the number of different genes (number > 2) in the three categories, are presented in Figure 3. The findings revealed that the primary DEGs were enriched in the immune response, extracellular matrix (ECM), ribosome, protein-glutamine gamma-glutamyltransferase of biological processes, cellular components, and molecular functions, respectively. Notably, the term “insulin-like growth factor II binding” in the GP vs. GE comparison may be a key factor in explaining the growth differences between the GP and GE groups.

3.3. KEGG Pathway Analysis of DEGs

The KEGG pathway classification (Figure 4) showed all DEGs categorized into six groups based on their biological functions: human diseases (1266), organismal systems (866), environmental information processing (543), cellular processes (445), metabolism (316), and genetic information processing (175; Figure 4). Pathways related to the metabolism of GP vs. GE were further subdivided into eight subsets, primarily including amino acid metabolism (60), lipid metabolism (79), carbohydrate metabolism (45), the metabolism of co-factors and vitamins (32), glycan biosynthesis and metabolism (28), and energy metabolism (19; Figure 4). Additionally, subsets such as the immune system (269), endocrine system (189), infectious diseases of a bacterial origin (180), cell growth and death (125), and transport and catabolism (164) were the main components of organismal systems, human diseases, and cellular processes, respectively.

The functional enrichment analysis based on the KEGG pathway classification (Figure 5) revealed six pathways directly related to metabolism in GP vs. GE. Specifically, the main metabolic pathways included glycolysis/gluconeogenesis (ko00010), arginine and proline metabolism (ko00330), arachidonic acid metabolism (ko00590), and glycine, serine, and threonine metabolism (ko00260) in GP vs. GE (Figure 5).

Following KEGG pathway enrichment analyses, key metabolic pathways and growth-related signal pathways were identified for a further analysis in GP vs. GE, including glycolysis/gluconeogenesis (ko00010), fatty acid synthesis (ko00061), oxidative phosphorylation (ko00190), arginine biosynthesis (ko00220), glycine, serine, and threonine metabolism (ko00260), the MAPK signaling pathway (ko04010), the mTOR signaling pathway (ko04150), the TGF-β signaling pathway (ko04350), and the PI3K-Akt signaling pathway (ko04151;

Table 3. Metabolic and growth-relative pathways’ differentially expressed genes identified in the muscle of G. przewalskii (GP) vs. G. eckloni (GE) by transcriptomic analysis.

Swiss-Prot Annotation	Swiss-Prot Id	log2FC	p-Value
Metabolism
Glycolysis/gluconeogenesis (ko00010)
Fructose-1,6-bisphosphatase 1, FBP1	Q3SZB7	2.9663	0.0022
ATP-dependent 6-phosphofructokinase, PFK-l	P12382	1.5851	3.00 × 10−18
Fructose-bisphosphate aldolase C, ALDOC	P53448	1.1698	1.73 × 10−26
Fructose-1,6-bisphosphatase isozyme 2, FBP2	Q9N0J6	−1.0720	1.87 × 10−43
Beta-enolase, ENO3	B5DGQ7	−1.1410	3.87 × 10−60
Pyruvate kinase, PKM	Q92122	−1.3675	4.19 × 10−29
Glyceraldehyde-3-phosphate dehydrogenase, GAPDH	Q5XJ10	−1.3808	1.06 × 10−63
Phosphoglycerate mutase 2, PGAM2	P15259	−1.3868	4.98 × 10−73
Phosphoglucomutase-1, PGM-1	Q08DP0	−2.9078	1.56 × 10−7
Fatty acid biosynthesis (ko00061)
Long-chain-fatty-acid-CoA ligase 4, ACSL4	Q9QUJ7	2.3333	1.93 × 10−45
Enoyl-[acyl-carrier-protein] reductase, MECR	Q28GQ2	1.9786	0.0047
Glutaryl-CoA dehydrogenase, GCDH	Q2KHZ9	1.6638	3.80 × 10−124
Long-chain-fatty-acid-CoA ligase 1, ACSL1	Q9JID6	−1.0999	2.52 × 10−9
Long-chain-fatty-acid-CoA ligase, ACBG2	Q7ZYC4	−1.3741	1.14 × 10−6
Oxidative phosphorylation (ko00190)
Cytochrome c oxidase subunit 1, COX1	O78681	1.3706	7.66 × 10−70
V-type proton ATPase 16 kDa proteolipid subunit, ATP6C	P27449	1.3687	2.71 × 10−20
V-type proton ATPase subunit S1, ATP6IP1	P40682	1.0171	3.27 × 10−5
Arginine biosynthesis (ko00220)
Arginase, non-hepatic 2, ARG2	Q91554	3.0893	7.09 × 10−188
Glutamine synthetase, GS	Q4R7U3	2.8844	0.0002
Argininosuccinate synthase, ASS	Q66I24	1.3530	2.09 × 10−18
Alanine aminotransferase 2-like, ALAT2	Q6NYL5	1.0654	8.87 × 10−5
Glycine, serine, and threonine metabolism (ko00260)
Glycine cleavage system H protein, GCSH	P11183	1.9057	1.25 × 10−7
Glycine dehydrogenase (decarboxylating), GLDC	P23378	1.8935	9.32 × 10−43
L-threonine 3-dehydrogenase, TDH	Q2KIR8	1.5801	4.53 × 10−42
2-Amino-3-ketobutyrate coenzyme A ligase, KBL	O75600	1.0194	2.34 × 10−12
Guanidinoacetate N-methyltransferase, GAMT	Q71N41	−1.0939	2.06 × 10−38
5-Aminolevulinate synthase, erythroid-specific, ALAS-E	Q9YHT4	−1.3104	4.77 × 10−55
Phosphoserine phosphatase, PSPH	P78330	−1.4583	1.82 × 10−8
Arginine and proline metabolism (ko00330)
Ornithine decarboxylase, ODC	P09057	2.1095	1.96 × 10−102
Creatine kinase S-type, CKS	P11009	1.2791	0.0054
Prolyl 4-hydroxylase subunit alpha-1, P4HA1	Q1RMU3	1.0175	4.95 × 10−8
Spermine oxidase, SMO	Q70LA7	−1.0509	5.59 × 10−11
Creatine kinase B-type, CKB	P05122	−1.1671	2.80 × 10−17
Arachidonic acid metabolism (ko00590)
Prostacyclin synthase, PTGIS	F1RE08	2.0229	2.97 × 10−35
Prostaglandin G/H synthase 2, COX2	P70682	1.7387	6.57 × 10−94
Prostaglandin E synthase, PTGES	Q95L14	1.7151	1.04 × 10−62
Polyunsaturated fatty acid 5-lipoxygenase, LOG5	P09917	−2.4108	6.26 × 10−8
Growth-relative pathways
MAPK signaling pathway (ko04013)
Growth-arrest- and DNA-damage-inducible protein gamma, GADD45G	Q2KIX1	2.0844	2.02 × 10−55
Growth-arrest- and DNA-damage-inducible protein beta, GADD45B	O75293	1.7536	1.46 × 10−135
Dual-specificity protein phosphatase 1-A, DUSP1-A	Q91790	1.5648	1.18 × 10−174
Growth-arrest- and DNA-damage-inducible protein alpha, GADD45A	P24522	1.0219	2.70 × 10−7
Insulin-like growth factor II, IGF2	Q02816	−1.2402	5.33 × 10−11
Mast/stem cell growth factor receptor, KITA	Q8JFR5	−1.2955	1.96 × 10−9
Ribosomal protein S6 kinase alpha-5, MSK1	Q8C050	−3.9579	0.0126
mTOR signaling pathway (ko04150)
Large neutral amino acid transporter small subunit 5, SLC7A5	Q7YQK4	1.5903	3.22 × 10−7
TBC1 domain family member 7, TBC1D7	F1QRX7	1.2640	1.25 × 10−7
CAP-Gly domain-containing linker protein 1, CLIP1	O42184	1.1088	3.60 × 10−31
Rapamycin-insensitive companion of mTOR Rictor, RICTR	Q6QI06	1.1008	3.20 × 10−11
E3 ubiquitin-protein ligase, RNF152	Q58EC8	−2.1847	0.0064
PI3K-Akt signaling pathway (ko04151)
Integrin alpha-6 light chain, ITGA6	P26007	1.8209	3.80 × 10−91
Integrin alpha-3 light chain, ITA3	Q62470	1.8050	2.42 × 10−21
Collagen alpha-1(VI) chain, COL6A1	P12109	−1.5839	6.26 × 10−65
Collagen alpha-2(I) chain, COL1A2	O93484	−2.1801	3.16 × 10−151
Collagen alpha-1(I) chain, COL1A1	P02457	−2.3289	1.42 × 10−189
TGF-beta signaling pathway (ko04350)
Latent-transforming growth factor beta-binding protein 1, LTBP1	Q00918	−1.0871	0.0030
Decorin	Q29393	−1.3786	8.19 × 10−105
Repulsive guidance molecule A, RGMA	Q8JG54	−1.6540	1.56 × 10−12
Noggin-2, NOGG2	Q9W740	−2.6002	0.0115

). Emphasis was placed on these pathways and the differentially expressed genes they encompass.

3.4. Validation of DEGs by RT–PCR

Randomly selected DEGs in GP vs. GE (ARG2, FBP1, ODC, GADD45G, PFK-1, LTBP1, IGF2, PKM, GAPDH, and LOG5) underwent validation through a quantitative real-time polymerase chain reaction (qRT–PCR). Primer design was based on the mapped sequences (Table 1). Results from qRT–PCR were compared with those obtained from transcriptome sequencing, and demonstrated consistency between the two different methods (Figure 6).

4. Discussion

4.1. Differences in the Growth Rate

The growth rate is a pivotal commercial trait for aquaculture, influenced by various factors [1,8]. G. przewalskii and G. eckloni are economically vital cold-water fishes in China [15,16], closely related and inhabiting similar waters [21,22]. However, our previous observations indicated a noticeable difference in the growth rates of these two species. In this study, G. eckloni exhibited a significantly faster growth rate than G. przewalskii juveniles. From day 3 to day 270, G. eckloni exhibited 2.2 times the growth rate of G. przewalskii, demonstrating a clear growth advantage. This rapid growth is crucial for economic fish performance. Although existing studies suggest that G. przewalskii tends to grow slowly, potential factors such as Qinghai Lake’s extreme hydrological and climatic conditions may contribute to this difference [21]. Additionally, genetics may be an issue here that drives different growth rates among the two species [1].

4.2. Differences in the Metabolism Pathways

Genetic resources are crucial factors influencing growth rates. The transcriptome provides a useful method for exploring genetic information. Relying on the transcriptome can yield insights into growth differences [4,5,6,29]. This study identified a noticeable growth difference, along with an evident transcriptome distinction. A total of 5574 DEGs were found between G. przewalskii and G. eckloni. Moreover, there were significant differences in gene expression patterns between these two fishes. These DEGs, exhibiting diverse expression patterns, may contribute to the growth rate difference observed among G. przewalskii and G. eckloni.

The glycolysis/gluconeogenesis pathway, primarily regulated by glycolytic genes, has been demonstrated in earlier research to contribute to muscle growth stimulation [30,31]. Studies indicate that the activation of glycolysis/gluconeogenesis in female muscle tissues contributes to sexual size dimorphism in flatfish as key genes involved in glycolysis (e.g., gpi, aldo, tpi, gapdh, bpgm, pgk, eno2, pk and ldh) were expressed to a greater extent in female muscle tissues than in male muscle tissues [31]. In our study, most genes related to glycolysis/gluconeogenesis were downregulated in G. przewalskii, including ENO3 (beta-enolase), PKM (pyruvate kinase), GAPDH (glyceraldehyde-3-phosphate dehydrogenase), PGM-1 (phosphoglucomutase-1), and FBP2 (fructose-1,6-bisphosphatase isozyme 2). These genes, implicated in muscle growth, could explain the growth differences between G. przewalskii and G. eckloni. Additionally, the enzyme pyruvate kinase M2 (PKM2) of glycolysis is a significant mediator of growth signals promoting cell proliferation [32]. The low expression of PKM-1 in G. przewalskii may have contributed to slower growth than in G. eckloni.

Oxidative phosphorylation is the primary pathway for energy production under aerobic conditions, critical for life, growth, and development in fishes. Impaired oxidative phosphorylation in hepatic mitochondria has been linked to growth retardation in rats [33]. Our results indicate that three DEGs related to oxidative phosphorylation were upregulated in G. przewalskii. Two of these genes were associated with proton reflux ATP synthase, playing a crucial role in energy generation. Their upregulation may suggest that G. przewalskii has a more vigorous energy metabolism and a greater energy demand for growth than G. eckloni. Arginine synthesis, with high enzyme activities involved in the urea cycle, is relatively high, especially during the early developmental stage of fish, and is crucial for fish growth; the urea cycle consists of five key enzymes [34]. Arginase (ARG) catalyzes the final step of the ornithine–urea cycle (OUC), leading to the conversion of L-arginine to L-ornithine and urea. ARG exists as two predominant isoforms, namely cytosolic ARG1 and mitochondrial ARG2 [35]. We found that ARG2, the mitochondrial isoform, had the highest expression in G. przewalskii. Three genes (GS, ASS, and ARG2) encoding key enzymes of the urea cycle were upregulated in G. przewalskii. The results of this study indicated that G. eckloni had faster growth and development, while G. przewalskii was still in the early stage of growth and development, perhaps in part because G. przewalskii had an upregulated expression of these three genes compared to G. eckloni. The arachidonic acid pathway plays a crucial role in cardiovascular biology, carcinogenesis, and various inflammatory diseases [36]. The expression of polyunsaturated fatty acid 5-lipoxygenase (LOG5) was low in G. przewalskii. LOG5 converts arachidonic acid to inflammatory mediators in the presence of a 5-lipoxygenase-activating protein [37]. LOG5 overexpression in tumor tissues leads to breast cancer [38], but the specific role of LOG5 in fishes is not yet fully understood.

4.3. Differences in the Relative Growth Pathways

In the modulation of skeletal muscle growth in teleost fishes, MAPK, PI3K-Akt, and mTOR are crucial signaling pathways [6,9]. In our study, growth arrest and DNA damage (GADD) were enriched in the MAPK signaling pathway. The GADD-inducible gene family is often upregulated in response to various environmental stresses and drug therapies. GADD45A, the first stress-inducible gene identified to be upregulated by p53, plays a key role. When GADD45A is deleted or repressed, cells exhibit uncontrolled proliferation [39]. The upregulation of GADD45A in G. przewalskii may retard cell proliferation, potentially explaining why G. przewalskii grows more slowly than G. eckloni.

Most notably, IGF2, a DEG in the MAPK pathway, was found in both G. przewalskii and G. eckloni, with lower expression in G. przewalskii. Previous studies have indicated that IGFs activate signaling molecules involved in fish muscle growth [6,11]. IGFs stimulate muscle growth by promoting myogenic cell proliferation, protein synthesis, and hypertrophy [10,11]. Particularly, IGF1 and IGF2 activate MAPK/ERK and PI3K/AKT signaling pathways in fish skeletal muscle [9,12]. IGF2 has a pronounced and direct effect on muscle growth in fishes; for instance, it activates these pathways in the myogenic cells of the gilt-head sea bream (Sparus aurata) [40] and rainbow trout (Oncorhynchus mykiss) [41]. IGF2 is more potent than IGF1 in stimulating muscle growth in fish species [40,42]. Additionally, IGF2 may contribute to the growth of hybrid grouper (Epinephelus fuscogutatus♀ × Epinephelus lanceolatus♂) by enhancing protein and glycogen synthesis [6], aligning with the findings of this study. The downregulated expression of DEGs (such as ENO3, PKM, GAPDH, PGM-1 and FBP2) in the glycolysis/gluconeogenesis pathway might be controlled by the downregulated expression of IGF2. This raises doubts about whether the regulation of G. przewalskii and G. eckloni growth may be directly influenced by IGF2 through glycogen synthesis.

The mTOR signaling pathway is crucial in cell growth, development, and protein synthesis [43]. In this pathway, some DEGs related to cell growth, such as TBCD7, were observed between G. przewalskii and G. eckloni. TBCD7, an important gene in the mTOR signaling pathway, is identified as the third subunit of the tuberous sclerosis complex (TSC), a negative regulator of cell growth [44]. Our study indicates that TBCD7 expression is upregulated in G. przewalskii, potentially contributing to its slowing of growth. The KEGG pathway analysis revealed significant differences in the expression of genes related to growth- and development-related metabolic pathways, such as the PI3K-Akt signaling pathway and TGF-β signaling pathway [45]. The PI3K/Akt signaling pathway, often enriched in the growth differential transcriptome [46], plays a crucial role. Collagen VI, a major extracellular matrix (ECM) protein, maintains the functional integrity of skeletal muscles [47]. Col6a1 is implicated in ECM remodeling during muscle fibrosis. Col6a1/mice and collagen VI-deficient zebrafish display a myopathic phenotype [47]. Mutations in the three collagen VI genes COL6A1, COL6A2 and COL6A3 cause Bethlem myopathy and Ullrich congenital muscular dystrophy (UCMD) in humans [48]. However, the functions of COL6A1, COL6A2 and COL6A3 in fish muscle growth require further investigation. The TGF-β signaling pathway is also involved in growth- and development-related metabolic pathways [45]. Decorin, a myokine and a connective tissue protein, stimulates connective tissue accretion and muscle hypertrophy [49]. It exerts negative control over cell proliferation by blocking TGF-β [50]. Bone morphogenic proteins (BMPs) are growth factors crucial for skeletal development and bone growth and are regulated by Noggin, a soluble BMP antagonist, partially through feedback inhibition [51]. Compared with G. eckloni, both Decorin and Noggin 2 were weakly expressed in G. przewalskii, suggesting their involvement in the negative regulation of fish muscle growth.

5. Conclusions

In summary, G. eckloni exhibits faster growth than G. przewalskii. The RNA-Seq analysis revealed more than 5574 DEGs between the two species, with the KEGG enrichment pathway highlighting molecular functions related to glycolysis/gluconeogenesis, arachidonic acid metabolism, the citrate cycle, and MAPK, PI3K-Akt, and mTOR signal pathways. This study further uncovers growth-related genes and pathways, emphasizing the potential importance of the differential expression of IGF2 in the growth rate difference between G. eckloni and G. przewalskii. This transcriptome analysis enriches the gene resources for both species, contributing to deeper understanding of the growth mechanisms of Gymnocypris.