Effects of High Dietary Starch Levels on the Growth Performance, Liver Function, and Metabolome of Largemouth Bass (Micropterus salmoides)

Abstract

In this study, we conducted a 16-week feeding trial to investigate the effects of a high-cassava starch diet on growth performance, liver function, and metabolism in largemouth bass (Micropterus salmoides). We formulated five diets containing varying levels of cassava starch: 12%, 9%, 6%, 3%, and 0% (termed M12, M9, M6, M3, and M0, respectively). We distributed these diets among largemouth bass with the initial body weight of 83.33 ± 0.55 g via an in-pond “raceway” aquaculture system. Our findings suggest that high level (12%) of cassava starch dietary inclusion adversely affected growth performance metrics such as weight gain rate and specific growth rate, along with feed utilization efficiency indicators, including protein efficiency, protein deposition rate, and the apparent digestibility of dry matter and protein. This negative impact was accompanied by a decrease in intestinal amylase activity. Through further transcriptomic analysis, we identified several key genes associated with carbohydrate metabolism, which underwent changes influencing liver function. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis revealed the involvement of these differentially expressed genes (DEGs) in the tricarboxylic acid cycle (TCA cycle). Comparative metabolomics analysis further indicated that the M9 group showed significant enrichment in pathways related to amino acid metabolism and alterations in the levels of metabolites involved in carbohydrate metabolism. In conclusion, our study demonstrates that incorporating up to 9% cassava starch in the diet can enhance growth performance in largemouth bass by stimulating digestive enzyme activities and promoting glucose utilization.

1. Introduction

In recent years, sugars, as cost-effective and widely available energy sources, have been increasingly incorporated into aquatic animal feeds to provide not only energy but also spare proteins for growth. However, the metabolic capacity of fish to utilize these sugars efficiently is limited compared with terrestrial mammals. This limitation is attributed to factors such as reduced insulin receptors and the absence or low activity of key enzymes like hexokinase (HK) and low glucokinase (GK), which are essential for glucose metabolism [1]. Studies have shown that compared with simple sugars, complex carbohydrates such as starch are better suited for fish nutrition due to a structure that facilitates absorption and utilization [2,3]. As the most important carbohydrate in feed ingredients, starch has become an indispensable ingredient in aquatic feeds. However, many fish species, especially carnivorous ones, display glucose intolerance. For example, rainbow trout (Oncorhynchus mykiss) develop persistent hyperglycemia and suffer from liver damage and immunosuppression when their diet exceeds 20% carbohydrates [4]. Therefore, in recent years, research on the application of starch in the aquatic feed industry has emerged.

Largemouth bass has become one of the leading freshwater aquaculture species in China due to its taste, nutritional benefits, rapid growth, adaptability, omnivorous diet, low disease susceptibility, and minimal intermuscular bones. At present, although studies have evaluated the optimal feed starch sources and levels for largemouth bass, revealing enhanced growth with 0–13% corn starch [5] and optimal results with 10% wheat starch, the effects of cereal starch on the growth and digestion of largemouth bass remains underexplored [6]. Compared with cereal starch, cassava starch, characterized by a lower amylose-to-amylopectin ratio, suggests a potentially higher digestibility for fish due to its Lin high amylopectin content [7]. Cassava starch has been successfully applied in feeds across livestock, poultry, and aquaculture, coupled with its advantages in viscosity, permeability, and film-forming properties, which underscore its potential. Moreover, it has lower crude protein and crude ash content than corn starch, which are irreplaceable physicochemical properties [8]. In recent years, the application effects of cassava and cassava starch have been explored in aquatic feeds. It has been reported that adding cassava to the feed of Nile tilapia (Oreochromis niloticus) [9] and grass carp (Ctenopharyngodon idella) [10] improves nutrient absorption and muscle quality without compromising growth. However, investigations into cassava starch’s implications for largemouth bass, especially regarding liver metabolism, are notably lacking.

Therefore, this study aims to fill this gap by assessing the effects of different levels of cassava starch on the growth performance, digestibility, and sucrose metabolism of largemouth bass, thereby contributing to a better understanding of the health implications of dietary starch in largemouth bass aquaculture.

2. Materials and Methods

2.1. Experimental Diets

In the present study, five experimental diets were formulated to contain 12.71%, 11.6%, 10.70%, 9.83%, and 8.91% starch (named as M12, M9, M6, M3, and M0), respectively. Wheat gluten was used to homogenize feed starch levels to 12%. Fish meal was used as the main protein source, and soybean oil was used as the lipid source. Ingredients and detailed compositions of experimental diets are shown in Table 1. After being ground through a 247 µm mesh, the feed ingredients were mixed through a commercial food mixer. Pellet feed with a particle size of 10 mm was prepared and then naturally air-dried and stored at 4 °C until used.

2.2. Fish Management

Largemouth bass were obtained from a breeding farm in Huzhou, Zhejiang, China. These fish were fed with a commercial diet and acclimated for 2 weeks in circulating water tanks in the laboratory. After adapting to the experimental conditions, 1800 individuals (average initial weight: 83.33 ± 0.55 g) with healthy physiques and no injuries were randomly divided into five groups with five replicates. Each replicate had 120 individuals cultured in an in-pond “raceway” aquaculture system (length × wide × height: 3 m × 0.6 m × 2.5 m, water depth: 1.5 m). All fish were fed the diet to apparent satiation twice daily (8:00 and 17:00) for 112 days (13 August–2 December, Huzhou, Zhejiang Province). During the breeding period, the water temperature ranged from 22.0 to 32.5 °C, dissolved oxygen was more than 5.36 mg/L, ammonia-N was less than 0.25 mg/L, and nitrite nitrogen was less than 0.16 mg/L. The photoperiod of culture was 12 h light and 12 h dark. The study was approved by the Institutional Animal Care and Use Ethics Committee of Agriculture Ministry Key Laboratory of Healthy Freshwater Aquaculture, Zhejiang Institute of Freshwater Fisheries (Huzhou, China) (AEEIZJF-20210101).

2.3. Sample Collection

At the end of the trial, the largemouth bass were fasted for 24 h, and then they were weighed and counted after being anesthetized with 0.01% MS-222 (Sigma, St Louis, MO, USA). The survival rate (SR), final body weight (FBW), weight gain rate (WGR), specific growth rate (SGR), feed conversion ratio (FCR), protein efficiency (PER), and protein retention efficiency (PRE) were calculated in each test pond. Individual body weight, body length, viscera, liver, and intestine weight were sequentially recorded from twenty randomly selected fish (half male and female) in each pond to calculate condition factor (CF), viscerosomatic index (VSI), hepatosomatic index (HSI), enterosomatic index (ESI), and visceral adipose index (VAI). In the above dissected fish, five per tank were used for liver and intestine enzyme activity analysis and glycogen content detection of liver. Additionally, six fish (half male and female) were pooled for whole-body composition analysis and muscle glycogen content. These two detections were also performed in pooled dorsal muscles of male and females respectively. Based on growth performance, liver tissues from 18 fish (6 in each pond) were collected from groups M12, M9, and M0, frozen in liquid nitrogen, and placed in a refrigerator at −80 °C for transcriptome sequencing.

2.4. Chemical Analysis

All chemical composition analyses of diets, the whole body, muscle, and the liver were performed by standard methods [11]. The Kjeldahl method (N × 6.25) was used to measure crude protein content. Moisture was determined after samples were dried at 105 °C till constant weight. Crude lipid was determined by the Soxhlet extraction method, and crude ash was determined by combustion in a muffle furnace at 550 °C for 16 h. Starch was analyzed by colorimetric using 3,5-dinitrosalicylic acid at a wavelength of 540 nm (TU-1900, PERSEE, Beijing, China). Total intestinal protein concentration and trypsin, lipase, and amylase activities were measured according to the instructions of a commercial kit (Jian Cheng Bioengineering Institute, Nanjing, China). All assays were performed in triplicate.

2.5. Hepatic Transcriptome Analysis

The total RNA of liver samples was isolated using Trizol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s instructions. The concentration, purity, and integrity of the RNA were examined using the Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA). Six extracted RNA samples were pooled as one sample to minimize the variation among individuals. Then, cDNA libraries were constructed with Illumina’s standard protocol for RNAseq library according to the procedures described [12]. Finally, the constructed libraries were sequenced on the Illumina HiSeq™ 4000 platform (Illumina, San Diego, CA, USA) by Genedenovo Biotechnology Co., Ltd. (Guangzhou, China). The sequencing data were submitted to NCBI with accession no. PRJNA948576.

Data processing procedures were conducted as follows. In short, the sequence raw reads were filtered, low-quality reads were discarded, and the clean reads were mapped to the reference genome of largemouth bass (assembly GCF_014851395.1 ASM1485139v1) using Hisat2 v2.1.0 with default parameters [13]. The expression levels were normalized to fragments per kilobase of transcript per million mapped reads (FPKM) for further analysis. Differentially expressed genes (DEGs) analysis was performed by the DESeq2 R package v16.1.1 [14] with the criteria of FDR < 0.05 and log2|(fold change)| > 1. Finally, KEGG categories were enriched based on the KEGG database (http://www.genome.jp/kegg/, accessed on 14 April 2023).

2.6. LC-MS/MS Non-Target Metabonomics

In the present study, non-target metabonomics data were collected by Guangzhou Genedenovo Biotechnology Co., Ltd. (Guangzhou, China) and analyzed on the OmicShare bioinformatics platform (v1). All samples were weighted and treated according to the method described in Ning et al. [15]. Quality control (QC) of each sample was conducted by mixing equal volumes of aliquots of the supernatants. LC-MS/MS analysis was performed by UHPLC (1290 Infinity LC, Agilent Technologies) coupled with QTOF (TripleTOF 6600, SCIEX, Shanghai, China) according to the manufacturer’s instructions.

Principal component analysis (PCA) and orthogonal projection to latent structures-discriminant analysis (OPLS-DA) were carried out to distinguish metabolites between the three groups. The variable important in projection (VIP) value of differential metabolites was also calculated for difference identification. Only a metabolite with VIP ≥ 1 was reserved. A metabolite with the value of p < 0.05 was considered as a differential abundance metabolite (DAM). All the metabolites were matched against the KEGG database (Kyoto Encyclopedia of Genes and Genomes, http://www.genome.jp/kegg/, accessed on 14 April 2023) and the annotation information was analyzed.

2.7. Calculations

Formulas used in the study are as follows.

Survival rate (SR, %) = final fish number/initial fish number × 100;

Weight gain rate (WGR, %) = (final body weight − initial body weight)/initial body weight × 100;

Specific growth rate (SGR, %/d) = [Ln (final body weight) − Ln (initial body weight)] × 100/days;

Feed conversion rate (FCR) = dry feed intake/(final body weight − initial body weight);

Protein efficiency ratio (PER) = total weight gain/protein intake;

Protein retention efficiency (PRE, %) = (final protein weight − initial protein weight)/protein intake × 100;

Condition factor (CF, g/cm³) = body weight/body length³;

Viscerosomatic index (VSI, %) = internal organ weight/body weight × 100;

Hepatosomatic index (HSI, %) = liver weight/body weight × 100;

Enterosomatic index (ESI, %) = intestine weight/body weight × 100;

Gonadosomatic index (GSI, %) = gonad weight/body weight × 100.

2.8. Statistical Analysis

All data were presented as the mean value ± standard deviation (SD). A one-way ANOVA was performed for the data analysis of five groups in SPSS (v25.0) software in the present study, and Student’s t-test and Tukey’s HSD were used for significance analysis. p < 0.05 was considered significantly different. The graphics were drawn using GraphPad Prism 9.0 (GraphPad Software Inc., San Diego, CA, USA).

3. Results

3.1. Growth Performance and Feed Utilization

The results of growth performance and feed utilization are shown in Table 2. The FCR and PER had no statistically significant difference in all five groups (p > 0.05). The M9 and M6 group exhibited significantly higher FBW, WGR, and SGR compared with the high-cassava starch (M12) group (p < 0.05). In the M9 group, the FBW was 3.14 times the initial weight after 111 days of cultivation. There was no significant difference in FBW after adding wheat gluten to diets (group M9 ~ M0) (p > 0.05). In addition, the PRE in the M9 group was notably higher than those in the other four groups (p > 0.05).

3.2. Morphometric Parameters and Chemical Composition

Morphological parameters were measured in the five groups of male and female individuals, respectively (Table 3). The indices of ESI and GSI had no differences among groups, whether male or female. Similar results were observed with CF of males. In females, CF showed a downward and then upward trend with the supplementation of wheat gluten in diets. The M6 and M3 groups had lower values of CF than the other three groups (p < 0.05). Both VSI and HSI were higher in the M12 and M6 group, independent of gender (p < 0.05). Dietary inclusion of different levels of cassava starch had no significant effect on the moisture and crude protein contents of male largemouth bass (p > 0.05) (Table 4). Similar results were observed for the crude protein and lipid contents of female fish. In males, the M9 and M3 groups had higher levels of crude lipid, yet the M12 and M0 groups had more crude ash compared with other groups, suggesting no apparent pattern. In muscle, the crude ash content of female individuals in the M12 group was significantly higher than that in the other four groups. Remarkable variations in moisture, crude protein, and crude lipid in male muscle were detected, with the maximum values distributed in the M3, M0, and M12 groups, respectively. Males in the M9 group had the highest crude protein and crude lipid content in the liver (p < 0.05).

3.3. Intestinal Digestive and Absorption Enzyme Activity

Dietary inclusion of different levels of cassava starch had no significant effect on the activity of intestinal trypsin and lipase in largemouth bass (p > 0.05, Table 5). The activities of intestinal amylase in the M12 group were significantly higher than those in the M6 and M3 groups (p < 0.05).

3.4. Detection and Functional Enrichment Analysis of DEGs in Liver

In the present study, largemouth bass from the M12, M9, and M0 groups were used to detect the difference at the transcriptional level. Nine libraries were constructed using the RNA from livers. A total of 23,578,187 raw sequencing reads were obtained. After trimming and quality control, the high-quality reads were mapped to the referenced genome, with a high mapping rate ranging from 96.01% to 97.46%. In total, 27,881 annotation genes of largemouth bass were identified. Expression profiles of genes from nine libraries were exhibited in the violin plot (Figure 1a). The mRNA expression showed high repeatability in M12 and M9, with correlation coefficients greater than 0.980 (Figure 1b). However, M0 had weak correlation with the other two groups; it was also greater than 0.911. M12-1 and M0-2 were outlays from the principal component analysis (PCA) analysis.

Under the conditions of FDR < 0.05 and |log2FC| ≥ 1, 157, 53, and 94 differentially expressed genes were identified in the comparisons of M12 vs. M0, M12 vs. M9, and M9 vs. M0, respectively (Figure 1d). To deeply understand the biological processes in which the DEGs were involved, KEGG pathway enrichment analysis was carried out. DEGs were divided into six categories, of which a total of 199 pathways were significantly enriched. In the category of metabolism, carbohydrate metabolism, amino acid metabolism, and nucleotide metabolism were the top involved pathways. In the category of organismal systems, the endocrine system was most enriched (Figure 2a). Pathways associated with the digestive system (pancreatic secretion and protein digestion and absorption) were exclusively over-represented, with high cassava starch level addition (Figure 2b). Pathways associated with carbohydrate metabolism (glycolysis/gluconeogenesis) were over-represented at the M9 stage (Figure 3).

3.5. Liver Metabolome

Analyses of the liver samples from M12, M9, and M0 were conducted in both positive (POS) and negative (NEG) ionization modes. Samples were densely overlapped in PCA analysis, suggesting a high reliability of sequencing datasets. As a result, a total of 17,845 and 11,286 metabolites were identified in POS and NEG mode, respectively. The clustered relationship and expression patterns of metabolites from these three groups are displayed in Figure 4, showing reliability for subsequent analysis. Multivariate statistical analysis was performed using the OPLS-DA model to maximize capture of the DAMs among groups (Figure 5). The cross-validation scores of OPLS-DA analysis for M0, M9, and M12 are displayed in Figure 5. Two-by-two comparison reveals that M0, M9, and M12 are clearly classified into independent clusters representing different metabolic profiles. In comparisons of M12 with the other two groups, both Q2s were more than 0.5, showing great prediction ability for this model.

Then, with the threshold of the OPLS-DA variable important in projection (VIP) score ≥ 1 and p-value < 0.05, significantly different metabolites (SDMs) between any two groups were carried out. As shown in Figure 6a, a total of 960, 787, and 531 different metabolites were identified between the comparisons of M12 vs. M0, M12 vs. M9, and M9 vs. M0, respectively. M12 had more DAMs with the other two groups. Between M12 and M0, the significantly different metabolites were notably enriched “Glycine, serine and threonine metabolism” (p < 0.05). Similar results can be found between M12 and M0. The SDMs between M9 and M0 were mainly related to “choline metabolism in cancer”.

4. Discussion

Evidence has demonstrated that high dietary carbohydrate levels impair growth performance in aquatic animals [16]. However, our study revealed that cassava starch supplementation showed a beneficial effect on weight gain and feed utilization of largemouth bass as the supplementation level increased, suggesting cassava starch’s potential to reduce fishmeal usage in aquafeeds while promoting growth performance. This result aligns with findings in Nile tilapia (Oreochromis niloticus) [17], where cassava starch improved growth, feed utilization, and nutrient composition, and in African catfish (Clurius guriepiuu), where a 49-day study by Olurin et al. [18] observed similar growth enhancement. These results collectively indicate cassava starch’s efficacy in fish growth promotion. However, with an increased proportion of cassava starch added to the diet, growth performance showed a non-significant decrease in snakehead fish (Channa striata) [19], paralleling the observed trends. This may be attributed to the biotechnological treatment of cassava starch, which reduced the content of certain anti-nutritional factors like concentrated tannins and flax bitter glycosides. Compared with terrestrial animals, the utilization of carbohydrates in fish is relatively low, and the carbohydrate level of feed for marine and freshwater fish should not exceed 20% and 40%, respectively. In the present study, a group with 9% cassava starch addition showed better growth performance.

Digestive efficiency, particularly the digestion of starch, is the most important factor in aquaculture nutrition, with amylase activity being a key indicator of cassava starch utilization in fish feeds. In this study, increasing cassava starch supplementation initially led to a decrease in the activity of intestinal amylase, followed by an increase, which is consistent with trends reported in snakehead (Channa argus) studies [20], indicating that an optimal level of cassava starch supplementation can enhance the secretion of amylase and enhance digestive efficiency in largemouth bass. In addition, it is important to note that the levels of starch addition were not the same at the highest amylase activity, suggesting differences in the ability of largemouth bass to utilize starch. This variability may be influenced by the size of starch granules and the ratio of straight-chain to branched-chain starch [21]. Previous studies reported that cassava starch, with a higher proportion of branched-chain molecules (20:80 ratio), is more easily hydrolyzed by enzymes compared with other starches, such as corn starch and wheat gluten, which was evidenced by the superior digestion of cassava starch [22,23]. Therefore, the largemouth bass displayed a stronger digestion ability for cassava starch than for wheat gluten in the present study. However, trypsin and lipase activities gradually decreased with the increase in cassava starch concentration, aligning with variations in the apparent digestibility of crude protein and lipids. Thus, it can be speculated that the digestive enzyme activity has a close relationship with the apparent digestibility of feed nutrients, ultimately affecting the content of crude protein and lipids, as well as other morphometric parameters.

In aquatic animals, the liver is involved in various biological processes, including lipid synthesis, starch degradation, and nutrient transport [24]. The inclusion of different levels of starch in the diet can alter carbohydrate metabolism in the liver. Previous studies have confirmed that high levels of starch in the diet upregulated carbohydrate biosynthesis-related genes and downregulated steroid biosynthesis-related genes in spotted sea bass (Lateolabrax maculatus) [25]. In the current study, a total of 157 DEGs were identified between M12 and M0, 53 DEGs between M12 and M9, and 94 DEGs between M0 and M9, showing a significant impact of cassava starch on liver carbohydrate deposition. KEGG enrichment analysis further highlighted that 35 genes were enriched in several pathways closely related to carbohydrate metabolism, including glycolysis/gluconeogenesis, with a notable difference between the M12 and M9 groups. Among these DEGs, there was nearly 30 percent in common in the comparisons involving the M12 group and the other two groups, suggesting a greater capacity for cassava starch deposition in the M12 group. Previous studies have shown that the balance between glucose storage and production depends on key enzymes associated with glycolysis, glycogenolysis, and gluconeogenesis [26]. In the present study, the formulated diet contained different levels of starch, which does not exist in forage fish and their feeding habits. Despite largemouth bass’s reported low efficiency in dietary starch utilization [5], our results indicate that high cassava starch diets significantly upregulated genes associated with the tricarboxylic acid cycle (TCA cycle), including GK and pyruvate kinase (PK), enhancing hepatic glucose metabolism and energy production [27].

In the present study, we explored the hepatic metabolomics profiling of largemouth bass fed with three levels of cassava starch diets using GC-MS techniques. Our analysis revealed significant alterations in metabolites primarily associated with the metabolism of glycine, serine, and threonine, suggesting that the cassava starch supplementations can bolster growth performance through enhanced amino acid metabolic pathways. Amino acids are the building blocks of proteins that are crucial for diet quality. Some amino acids, such as glycine, serine, aspartic acid, and norleucine, were elevated in the livers of largemouth bass in the M12 and M9 groups. While glycine is not classified as essential in fish diets, it has been shown to support the growth of fish and the efficiency of nutrient absorption in rainbow trout (O. mykiss) [28]. Moreover, glycine, serine, and threonine are abundant in collagen and elastin, which are essential proteins in connective tissue health [29,30]. Therefore, sufficient supply of the three amino acids is crucial for maximum collagen synthesis and optimal fish health [31]. These amino acids, along with vital vitamin B, play roles in regulating the function of the nervous system [32]. Additionally, serine and threonine regulate the body’s immune system, strengthen the body’s immunity, enhance antiviral and antioxidant functions, and maintain the body’s normal activity and homeostasis level [33]. The observed increase in these amino acids and improvements in digestive enzyme activities provide compelling evidence that optimal cassava starch supplementation can improve amino acid metabolism, thereby enhancing largemouth bass growth.

5. Conclusions

In summary, the optimal inclusion rate of cassava starch in the diet for largemouth bass should not exceed 9% based on evaluations using morphometric parameters and feed conversion ratios. Excessive dietary cassava starch may damage hepatic metabolism, reducing feed nutrient digestibility and hindering growth performance in largemouth bass. Therefore, this study’s exploration of varying cassava starch levels provides valuable insights for determining appropriate starch inclusion rates, offering a basis for optimizing commercial feed formulations for largemouth bass aquaculture.