Comparative Transcriptomic Analysis Reveals Molecular Mechanisms Underlying Scale Adhesion Differences Between Carassius auratus indigentiaus and Carassius auratus gibelio

Abstract

Scale adhesion strength is a key trait in aquaculture, directly influencing disease resistance, survival, and commercial value. The Dongting crucian carp (Carassius auratus indigentiaus, hereafter CaDT) is valued for its rapid growth and superior flesh quality but is characterized by loosely attached scales. In this study, we investigated the morphological and molecular basis underlying scale adhesion by comparing CaDT with the tight-scaled allogynogenetic gibel carp, Zhongke No. 3 (Carassius auratus gibelio, hereafter CaGB). Morphological analysis revealed a significantly lower scale-embedding ratio in CaDT compared to CaGB. To unravel the molecular mechanisms underpinning these phenotypic differences, a comparative transcriptomic analysis was conducted on scale sac, skin, and muscle tissues in CaDT and CaGB. In CaGB, gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses of differentially expressed genes (DEGs) in the critical scale sac tissue showed a significant upregulation of genes involved in ribosomal pathways. Specifically, key epithelial differentiation markers, including keratin 13 (krt13), keratin 15 (krt15), and metabolic genes, enolase 3-like (eno3l), and phosphoglycerate mutase 2 (pgam2) were significantly down-regulated in CaDT, which suggests a compromised epithelial cell differentiation capacity and reduced energetic and biosynthetic activity. Quantitative PCR (qPCR) validation across three tissues showed high concordance with the RNA-seq results, thereby confirming the reliability of the transcriptomic data. The results offer insight into the molecular basis for understanding scale adhesion traits, and provide valuable insights for selective breeding strategies to improve scale retention in aquaculture species.

1. Introduction

Fish scale adhesion is a critical determinant of survival and economic value in aquaculture, and yet remains an underexplored trait at the molecular level. Scale loss during netting and transportation creates portals for pathogen entry, leading to increased infection rates and mortality in severely affected populations. Market devaluation represents another significant impact, as fish with visible scale loss command lower prices due to consumer preferences. Despite its importance, most previous studies have focused on scale development [1,2,3] and biomineralization [4,5], while the molecular and morphological mechanisms governing scale attachment remain poorly understood.

The scale sac, a specialized invagination of the dermis and epidermis, provides the structural framework for scale attachment and growth [6]. The integrity of the scale sac interface determines the mechanical strength of scale adhesion and, consequently, the fish’s resilience to physical handling stress [7]. While previous studies have extensively characterized the advanced and biomedical materials development [8,9,10,11] of scales themselves and their applications [12,13,14], the biological mechanisms anchoring the scale within its pocket were largely unexplored.

Crucian carp (Carassius auratus) represent a cornerstone of global aquaculture, particularly within China’s freshwater aquaculture industry [15]. Modern selective breeding programs have successfully employed techniques such as hybridization and gynogenesis to develop enhanced varieties. A notable success is the allogynogenetic gibel carp, Zhongke No. 3 (Carassius auratus gibelio, CaGB), which exhibits an improved growth rate, superior yield, and enhanced uniformity [16]. In contrast, the Dongting crucian carp (Carassius auratus indigentiaus, CaDT), endemic to Beimin Lake, exhibits rapid growth and excellent flesh quality [17]. However, the loosely adherent scales in CaDT poses a significant challenge, which limits its aquaculture potential.

Given the economic and survival implications of scale adhesion, we performed a comparative study between the tight-scaled CaGB and the loose-scaled CaDT. This study aims to elucidate the morphological and molecular underpinnings of scale adhesion properties. The results provide valuable insights for selective breeding strategies and potential genetic modification approaches to improve scale adhesion in aquaculture species.

2. Materials and Methods

2.1. Experimental Fish

CaDT and CaGB were obtained from Beimin Lake and the Institute of Hydrobiology, Chinese Academy of Sciences (Wuhan, Hubei province, China), respectively, and then temporarily reared in the experimental tanks in Hunan Agricultural University. Ten healthy, one-year-old female fish of comparable body weights (CaDT: 179.44 ± 9.9 g and CaGB: 179.64 ± 11.6 g; p = 0.9830) were randomly selected and placed in tanks, and then temporarily reared for one week in an indoor recirculating aquaculture system at 28 °C.

2.2. Scale Morphology and Embedding Parameter Analysis

Fish were anesthetized with MS-222 before measurement. Subsequently, 6 groups were included in this experiment, and 3 fish were randomly selected from each group. From these selected fish, six scales per region in total were collected across the three body parts (dorsal, lateral line, and ventral region); specifically, the scales were counted starting from those at the posterior margin of the gill cover, i.e., the sixth to eleventh scales (Figure 1A). For each scale, the relevant parameters of scales were measured as Figure 1B.

For each sampled scale, the following parameters were determined:

Total scale length (TSL): the maximum anterior–posterior dimension (Figure 1B).

Total scale width (TSW): the maximum dorsal–ventral dimension, perpendicular to the length axis (Figure 1B).

Embedded scale length (EmSL): the portion of scale within the scale sac (Figure 1B).

Exposed scale length (ESL): the portion of the scale extending beyond the scale sac margin, calculated as TSL minus EmSL.

Embedding ratio (ER): calculated as EmSL/TSL × 100%.

2.3. Sampling

After morphological measurement of the scale-embedding parameters, fish were euthanized using excessive MS-222, followed by severance of the spinal cord. Tissue sampling was performed on 9 fish. Skin, muscle, and scale sac were sampled, making three tissue pools as biological replicates, each from three individuals. Pooled tissue samples were quickly placed in liquid nitrogen and then stored at −80 °C before RNA extraction. The scale sac tissue was carefully dissected under a stereomicroscope (Olympus, Tokyo, Japan) to minimize contamination from adjacent tissues.

2.4. RNA Extraction and Sequencing

Total RNA was extracted from the pooled tissue samples with the Trizol reagent (TransGen Biotech, Beijing, China). VAHTS Universal V10 RNA-seq Library Prep Kit (Vazyme, Nanjing, China) worked for constructing sequencing libraries on DNBSEQ-T7 platform, from 3 μg each of high-quality pooled RNA samples (OD260/280 ≥ 1.9, OD260/230 ≥ 1.8). The sequencing work was conducted by Igenebook Biotechnology Co., Ltd. (Wuhan, China).

2.5. Differentially Expressed Gene Analysis and Functional Enrichment

Firstly, the low-quality reads were filtered out by fastp (v0.21.0) [18], and clean reads were mapped to the reference genome (genebank NO: GCA_023724105.1), using HISAT2 (version 2.1.0) [19]. The expression levels of transcripts were statistically analyzed using featureCounts (v1.6.0). The expression values were normalized to FPKM (Fragments per kilobase of transcript per million Fragments mapped) [20]. Essentially, differential expressed genes (DEGs) were identified using edgeR (v3.36.0) [21], with screening thresholds of false discovery rate (FDR) < 0.05 and |1og2 (fold change)| ≥ 1.

To gain insight into phenotypic differences, GO (http://www.geneontology.org/, accessed on 5 March 2025) and KEGG (https://www.kegg.jp/, accessed on 5 March 2025) enrichment analysis of annotated DEGs were performed by the clusterProfiler package in R software (v4.10.1) for enrichment analysis; the significance of enrichment was assessed using a hypergeometric test. The resulting p-values were then adjusted for multiple testing, using the Benjamini–Hochberg false discovery rate (FDR) correction [22]. GO terms and KEGG pathways with p. adjust ≤ 0.05 were regarded as significant.

2.6. Quantitative PCR (qPCR) Validation of RNA-Seq Data

To verify the reliability of transcriptome sequencing results, we used qPCR to detect the expression levels of related genes. β-actin was used as the internal reference gene and the relative expression of each gene was analyzed using the 2^−ΔΔCT method [23]. qPCR was conducted as described before [24]. Selected genes and primers used for qPCR analysis are listed in Table S1.

2.7. Data Collection and Statistical Analysis

Data are shown as mean ± SD. The figures were drawn using GraphPad Prism software (v9.5.1). Before performing the t-tests, the normality of the data was assessed using the Shapiro–Wilk test, and the homogeneity of variances was evaluated using Levene’s test. All significance analyses were conducted in GraphPad Prism software (v9.5.1) by t-test analysis (repetitive). Statistical significance is represented by asterisks (* p < 0.05, ** p < 0.01, and *** p < 0.001).

3. Results

3.1. Scale Morphology and Embedding Parameter Analysis

The photographs of scales showed distinct differences in scale morphology between CaGB and CaDT across the three body regions (Figure 2A–C). Interestingly, CaDT scales exhibited significantly longer TSL, but narrower TSW, compared to CaGB scales in all body regions (Figure 2D–I) [TSL-dorsal scales (CaDT: 10.53 ± 2.17 mm vs. CaGB: 9.34 ± 0.77 mm, p < 0.001), lateral line scales (CaDT: 12.19 ± 1.68 mm vs. CaGB: 10.79 ± 0.79 mm, p < 0.001), and abdominal scales (CaDT: 11.24 ± 0.91 mm vs. CaGB: 10.36 ± 0.64 mm, p < 0.001); TSW-dorsal scales (CaDT: 8.55 ± 2.17 mm vs. CaGB: 9.34 ± 0.77 mm, p = 0.0022), lateral line scales (CaDT: 8.97 ± 1.25 mm vs. CaGB: 9.65 ± 1.65 mm, p = 0.0041), and abdominal scales (CaDT: 8.14 ± 0.64 mm vs. CaGB: 8.96 ± 1.16 mm, p < 0.001)], resulting in the observed significant differences in width–length ratios between CaDT and CaGB (Figure 2J–L).

The most pronounced difference in width–length ratio was observed in dorsal scales, with CaGB scales reaching 99.23 ± 10.77%, while CaDT scales remained 81.34 ± 13.66% (p < 0.001, Figure 2J,M). Lateral line and ventral scales also showed significantly lower ratios in CaDT (73.67 ± 11.33% and 72.45 ± 10.45%, respectively) than in CaGB (89.45 ± 10.55% and 85.56 ± 14.44%, respectively; p < 0.001 for both; Figure 2K,L,N,O). These results demonstrated that CaDT scales developed a more elongated morphology, compared to the more rounded scales of CaGB.

The scale-embedding analysis revealed that the most dramatic differences were observed in dorsal scales, where CaDT demonstrated a significantly lower ER (38.9 ± 8.97%) than CaGB scales (54 ± 9.63%, p < 0.001; Figure 3A,B). The CaDT lateral line scales demonstrated a significantly lower ER (49.23 ±12.8%) compared to CaGB scales (35.05 ± 12.53%, p < 0.001, Figure 3C,D). Similarly, ventral scales in CaDT showed an ER of 37.78 ± 9.78%, which was significantly lower than the 51.60 ± 8.4% in CaGB (p < 0.001, Figure 3E,F). This evidence established the lower embedding ratios (implying inferior scale sac anchorage) in CaDT as a key morphological correlation of its loose scale adhesion.

3.2. Gene Expression Analysis

In CaDT, a total of 35,028, 34,457, and 32,294 genes were expressed in the scale sac, skin, and muscle tissues, respectively, with 30,382 genes commonly expressed across all three tissues (Figure 4A,C). CaGB showed 31,588, 33,434, and 20,974 genes expressed in the scale sac, skin, and muscle tissues, respectively, with only 20,057 genes commonly expressed in all three tissues (Figure 4B,C). The distribution of gene expression levels in each group was relatively scattered, indicating substantial biological variation within each tissue’s combination.

3.3. GO/KEGG Analysis of DEGs in CaDT Versus CaGB

A total of 3953 DEGs were identified in the scale sac when CaDT was compared to CaGB, with 2404 up-regulated and 1549 down-regulated genes (Figure 5A,B). This was followed by 668 DEGs in the skin (377 up-regulated and 291 down-regulated; Figure 5C,D), and 10,095 DEGs in muscle (6980 up-regulated and 3115 down-regulated genes; Figure 5E,F) in CaDT versus CaGB.

In the scale sac, down-regulated DEGs in CaDT over CaGB were predominantly associated with the ribosomal structure and protein synthesis pathways, while up-regulated DEGs were enriched with cell adhesion and junction-related processes, such as the anchoring junction and adherens junction (Figure S1, Tables S2–S5). A similar trend was observed in skin tissue, where down-regulated genes were linked to ribosomal functions, and up-regulated genes were involved in cell junction assembly and developmental processes (Figure S1, Tables S6–S9). In muscle tissue, down-regulated DEGs in CaDT were enriched in metabolic and cardiovascular pathways, including glycolysis/gluconeogenesis and hypertrophic cardiomyopathy. In contrast, up-regulated DEGs were significantly associated with cell growth and signaling pathways’ focal adhesion, the PPAR signaling pathway, and other metabolic processes (Figure S1, Tables S10–S13). This comprehensive analysis underscores that the phenotypic divergence in scale adhesion is underpinned by distinct, tissue-specific molecular reprogramming.

3.4. Scale Sac-Specific Expressed Genes Analysis

Given that the scale ER showed pronounced differences between CaDT and CaGB, we hypothesized that the scale sac tissue plays a crucial role in determining scale attachment properties. Therefore, we conducted a focused analysis of genes specifically expressed in the scale sac to identify the molecular mechanisms underlying differential scale adhesion.

Function enrichment of scale sac-specific genes revealed both conserved and divergent features between CaDT and CaGB. Scale sac-specific expressed genes in CaDT were enriched with light perception, neuropeptide signaling, and ligand-gated ion channel complexes (Figure 6A). In contrast, those in CaGB were enriched with a broader sensory profile, encompassing both photic and mechanical stimulus perception. Additionally, CaGB showed stronger involvement in neuronal development pathways and fibroblast growth factor receptor signaling (Figure 6B). This comprehensive profile establishes the scale sac as a specialized microenvironment with well-developed but distinctly focused neuro-sensory capabilities.

Strikingly, the most significantly enriched pathways in CaDT and CaGB are cell adhesion molecules and cornified envelope formation, confirming that it is indispensable on a core molecular basis for the construction of a functional scale sac. Metabolically, shared metabolic pathways, including steroid hormone biosynthesis, taurine and hypotaurine metabolism, and beta-alanine metabolism, and viral infection-related pathways are identified in CaDT and CaGB (Figure 6C,D). Scale sac-specific expressed genes in CaGB displayed distinctive enrichment with xenobiotics metabolism by cytochrome P450, vitamin digestion and absorption, and butanoate metabolism.

3.5. Validation of RNA-Seq Results by qPCR

Thirty genes, including 10 genes from scale sac tissue (krt13, tβ, eef1b2, krt15, vmo1l, zfp395, angptl4, mcl1, hdlbpa, and myh9l), 10 genes from muscle tissue (c-fos, xirp1, hsp90α1, rad, and prelid3b, ckm, pvalbbl, eno3l, pgam2, and hbal), and 10 genes from skin tissue (serpine1, aldocb, lul, fstl3l, haus1, pparα, pgam2, pfkma, htra1blt and npntl) were selected for qPCR validation. qPCR results exhibited excellent concordance with RNA-seq analysis across all validated genes, indicating the high reliability of the transcriptomic data (Figure 7). All 30 genes showed consistent directionality of change between RNA-seq and qPCR, with 30 of 30 genes reaching statistical significance (p < 0.05) in both methods.

4. Discussion

This study provides a comprehensive morphological and molecular characterization of scale adhesion differences in crucian carp. CaDT exhibited poorer scale attachment, with a significantly lower scale-embedding ratio compared to CaGB, underscoring the critical role of the scale sac in determining the scale mechanical attachment strength. Morphologically, CaDT scales were rounder, whereas CaGB scales were more elongated. This difference is likely to be linked to their divergent scale-embedding rates and adhesion capabilities, though this potential connection merits further investigation.

Transcriptome analysis suggests that scale adhesion properties are the result of coordinated molecular reprogramming of multiple biological systems, rather than individual genetic changes. These results indicate a clear structural and molecular basis for the inferior mechanical adhesion properties in CaDT. Keratin proteins form the intermediate filament network, which provides mechanical stability for epithelial cells and serves as attachment points for cell–cell junctions [25,26]. Downregulation of epithelial differentiation markers, particularly krt15 (log2FC = −5.55) and krt13 (log2FC = −292 7.53), in CaDT over CaGB suggests that keratins play a critical role in scale attachment. The overall reduction in keratins in CaDT may contribute to diminished epithelial maturation and organization within the scale sac, resulting in a compromised cellular framework for scale adhesion. Glycolytic enzymes (e.g., eno3l, pgam2) and oxidative phosphorylation components are essential to meet the metabolic demands of enhanced tissue maintenance [27]. Current downregulation in CaDT of eno3l and pgam2 possibly reflects the deficiency in the substantial energetic and biosynthetic investment required to sustain a reinforced adhesion apparatus.

The significant downregulation in the scale and muscle of ribosome pathways indicates a systemic reduction in basal protein synthesis capacity in CaDT [28,29]. This suggests that ribosomal integrity is fundamental to maintaining structural stability in scale-associated tissues. Compromised scale adhesion in CaDT is likely closely linked to its systemic deficiency in ribosome biogenesis and protein synthesis capacity.

Our focused analysis of scale sac-specific transcriptomes underscores the specialized role of this tissue in scale attachment. CaDT and CaGB showed mostly conserved expression patterns in the scale sac. This suggests that the key phenotypic divergence in scale adhesion strength may not stem from the presence or absence of cell adhesion molecules or the cornified envelope formation pathway, but rather from differences in neuro-sensory and metabolic pathways. This evidence indicates that neurosensory-related pathways might indirectly regulate the adhesion of skin-derived structures [30,31,32].

Notably, the enrichment of the mucin type O-glycan biosynthesis pathway, specifically in CaDT, may suggest a distinct adaptive strategy compared to CaGB. Additionally, scale sac-specific genes in CaDT and CaGB were both enriched in viral infection-related pathways, indicating that CaDT might be as susceptible to Carassius auratus herpesvirus (CaHV) as CaGB, and may be a potential subject for research on the hemorrhagic disease in Carassius auratus [33].

From an aquaculture perspective, the study identified several candidate genes, such as krt13, krt15, eno3l, and pgam2, providing potential molecular targets for genetic improvement programs. These may serve as promising markers for marker-assisted selection or genetic modification aimed at enhancing scale adhesion. The observed divergence in scale adhesion between CaDT and CaGB likely reflects the adaptive evolutionary outcomes of their distinct natural histories and ecological niches. Future studies correlating wild population distributions and environmental stability with scale adhesion strength will be essential for proposing ecologically based evolutionary hypotheses.

5. Conclusions

In conclusion, scale adhesion is a complex trait and is regulated by a combination of morphology and multiple pathways. Moreover, we identified several potential trait-associated genes, aiming to lay the foundation for the molecular basis of the scale traits and provide targets for marker-assisted breeding for superior traits.