Regulatory effects of soy isoflavones on ovarian development in female Chinese mitten crabs (Eriocheir sinensis)

Abstract

Soy isoflavones are phytoestrogens that exhibit both estrogenic and/or antiestrogenic effects. This research investigated the potential of soy isoflavones as functional feed additives to promote ovarian development in female Chinese mitten crabs (Eriocheir sinensis). One hundred ninety-two crabs (101.52 ± 4.57 g) were randomly assigned to four groups (six replicates per group and eight crabs per replicate), and fed diets supplemented with 0.00, 32.51, 70.83, or 369.03 mg/kg soy isoflavones for 11 weeks. Compared to the control group, supplementation with 32.51 mg/kg soy isoflavones significantly increased the gonadosomatic index, hemolymph vitellogenin content, and the vtg mRNA levels in the hepatopancreas and ovary (P < 0.05). Moreover, supplementation with 32.51 and 70.83 mg/kg soy isoflavones significantly promoted yolk granule formation (P < 0.05). At the molecular level, soy isoflavones modulated estradiol levels and activated the estrogen-related receptor signaling. They also upregulated the expression levels of esrrb, hsd3b2, and hsd17b6 genes (P < 0.05), compared to the control group. Additionally, they increased estradiol synthesis through activating the cyclic adenosine monophosphate/PKA/CREB protein signaling pathway. High dose supplementation (369.03 mg/kg) did not significantly affect ovarian development (P > 0.05). Therefore, soy isoflavones exhibit a U-shaped effect on ovarian development of E. sinensis, with 32.51 mg/kg being an effective dose for promoting ovarian maturation.

1. Introduction

Soy isoflavones are phytoestrogens contained in soybeans that are structurally similar to the estrogen 17β-estradiol (E₂). By binding to estrogen receptors (ERs), soy isoflavones can exhibit estrogenic and/or antiestrogenic effects to influence reproductive development, growth, and immune responses (Cederroth and Nef, 2009). These properties suggest their potential application valuable in the field of reproductive health in aquaculture. Soy isoflavones exhibit both agonistic and antagonistic effects on the processes of sex differentiation and reproductive development in aquatic animals. Studies demonstrated that isoflavones induce feminization and sex reversal such as Japanese eel (Anguilla japonica) (Inaba et al., 2022) and Nile tilapia (Oreochromis niloticus) (El-Sayed et al., 2012). Furthermore, they disrupt sex differentiation in Russian sturgeon (Acipenser gueldenstaedtii) (Fajkowska et al., 2021) and negatively impact fertilization and hatching in goldfish (Carassius auratus) (Bagheri et al., 2013). However, the effects of soy isoflavones on the gonadal development of aquatic invertebrates (especially crustaceans) remain unclear.

Ovarian development in crustaceans is closely linked to yolk protein accumulation, particularly vitellin. Vitellogenin (VTG) is the precursor of vitellin, serves as a key biomarker for ovarian development (Maria et al., 2004). The biosynthesis of VTG is regulated by the interplay of multiple hormones, including methyl farnesoate, vitellogenesis-inhibiting hormone, hyperglycemic hormone, and others (Subramoniam, 2011). Among these, E₂ plays a critical role in regulating VTG synthesis. Both endogenous and exogenous E₂ significantly promote ovarian development and vitellogenesis in crustaceans (Pan, 2018). Reports have confirmed its effects in Kuruma prawn (Marsupenaeus japonicus) (Yano and Hoshino, 2006), Freshwater crayfish (Cherax albidus) (Coccia et al., 2010), Freshwater edible crab (Oziothelphusa senex senex) (Swetha et al., 2016), Tiger shrimp (Penaeus monodon) (Merlin et al., 2015), and Swimming crab (Portunus trituberculatus) (Liu et al., 2018). Crustaceans can also convert cholesterol into estrogens via enzymatic processes in the hepatopancreas and ovaries (Janer and Porte, 2007; Warrier et al., 2001). In the cytoplasm or nucleus, E₂ binds to specific ERs that then form homo or heterodimers. These complexes then bind to the promoter regions of estrogen responsive genes on specific estrogen response element sequences (Gruber et al., 2004). This action enhances the transcription of vtg (The full names of all gene and protein abbreviations are provided in Table S1) gene, facilitating its synthesis (Klinge et al., 2004). Estrogen synthesis occurs within theca and granulosa cells through the activation of aromatase, driven by the synergistic actions of follicle-stimulating hormone (FSH) and luteinizing hormone (Liu et al., 2021). The cyp19a1 gene encodes aromatase enzyme that catalyzes the demethylation and aromatization of testosterone and androstenedione into estrone and E₂ (Cui et al., 2013). This reaction represents a critical step in estrogen biosynthesis. Importantly, the effects of soy isoflavones on estrogen synthesis and VTG formation in crustaceans remain poorly understood.

The Chinese mitten crab (Eriocheir sinensis) is distributed across Asia, Europe, and North America, while its aquaculture is predominantly centered in China, accounting for an annual production of nearly 890,000 tons in 2023 (Ministry of Agriculture of the People's Republic of China, 2024). The ovaries are valued as a preferred edible part due to high nutritional content, which includes the polyunsaturated fatty acids, essential amino acids, and unique flavors (Bu et al., 2023). In commercial aquaculture, farmers commonly utilize fresh frozen fish during the fattening period. While effective for growth, this practice deteriorates water quality and increase disease susceptibility. To address these issues, artificial compound feeds have been developed. However, feeding feeds often experience delayed ovarian maturation and other related issues (Wu et al., 2007). Therefore, developing a functional feed additive to enhance ovarian development in E. sinensis is essential, as it directly impacts both the commercial value and offspring health.

Therefore, this study employed E. sinensis as a model to evaluate the potential of soy isoflavones as functional feed additive, and explore their regulatory mechanisms in ovarian development. These results provide valuable insights for optimizing feed formulations and managing ovarian development in E. sinensis.

2. Materials and methods

2.1. Animal ethics statement

All animal experimental procedures were conducted in accordance with the Guidelines for the Care and Use of Laboratory Animals from East China Normal University and with approval from the Animal Ethics Committee of East China Normal University (approval No. F20201002).

2.2. Diet formulation

The optimal supplemental level of soy isoflavones was determined on the previous research (He et al., 2024). Soybean meal accounted for 10%–20% in fattening feed by weight and contained 2.0 to 2.5 mg/g soy isoflavones (Chen et al., 2025; Shao et al., 2013; Xiao et al., 2011; Zhang et al., 2015, 2024). Based on this estimation, the background level of soy isoflavones in the feed was 200 to 500 mg/kg. Accordingly, the basal diet was supplemented with 0, 40, 80, and 400 mg/kg soy isoflavones (purity ≥ 98%; Xian Tiankang Biotechnology Co., Ltd., Xian, Shaanxi, China), respectively. In each diet, the analyzed concentrations of soy isoflavones by high–performance liquid chromatography (LC-20 CE, Shimadzu Corporation, Kyoto, Japan) were 0.00, 32.51, 70.83, and 369.03 mg/kg. The highest supplementation level (369.03 mg/kg) fell within the medium range of estimated background levels (200-500 mg/kg). The diet without soy isoflavones served as the control (C), and the experimental groups were designated as SL (32.51 mg/kg), SM (70.83 mg/kg), and SH (369.03 mg/kg), respectively. All feed ingredients were milled and sifted through a 60-mesh strainer, weighed, finely ground, and thoroughly mixed. Soy isoflavones and choline chloride were separately dissolved in oil and water, and then thoroughly blended with the mixture. Pellets (2 mm in diameter) were produced with an F-26 II pelletizer (South China University of Technology, Guangzhou, Guangdong, China). Pellets were ventilated indoors at 25 °C to approximately 10% moisture content and subsequently packaged in separate sealed bags for storage at −20 °C (Table 1).

Table 1.

Ingredient formulation and chemical composition of the experimental diets (dry matter basis, %).

Items	Groups1
	C	SL	SM	SH
Ingredients
Fish meal	35.00	35.00	35.00	35.00
Casein	13.50	13.50	13.50	13.50
Gelatin	4.50	4.50	4.50	4.50
α-Starch	18.00	18.00	18.00	18.00
Fish oil: soybean oil (1:1)	8.00	8.00	8.00	8.00
Vitamin premix2	4.00	4.00	4.00	4.00
Mineral premix3	2.00	2.00	2.00	2.00
Choline chloride	0.50	0.50	0.50	0.50
Cholesterol	0.50	0.50	0.50	0.50
Betaine	3.00	3.00	3.00	3.00
Soybean lecithin	2.00	2.00	2.00	2.00
Cellulose	6.900	6.896	6.892	6.860
Antioxidant	0.10	0.10	0.10	0.10
Carboxymethyl cellulose	2.00	2.00	2.00	2.00
Soy isoflavones, mg/kg	0.00	40.00	80.00	400.00
Total	100.00	100.00	100.00	100.00
Proximate composition4
Organic matter	91.32	91.40	90.90	91.32
Moisture	11.07	10.92	11.79	11.55
Crude protein	42.57	42.66	42.05	42.16
Crude lipid	13.06	12.54	12.50	12.73
Soy isoflavones, mg/kg	0.00	32.51	70.83	369.03

¹C, control diet; SL, SM, and SH, control diet with 32.51, 70.83, and 369.03 mg/kg soy isoflavones, respectively.

²Vitamin premix (per 100 g premix): folic acid, 0.025 g; retinol acetate, 0.043 g; α-tocopherol acetate, 0.5 g; biotin, 0.005 g; riboflavin, 0.0625 g; niacin, 0.3 g; pyridoxine hydrochloride, 0.225 g; thiamine hydrochloride, 0.15 g; para-aminobenzoic acid, 0.1 g; ascorbic acid, 0.5 g; cholecalciferol, 0.0075 g; menadione, 0.05 g; inositol, 1 g; Ca pantothenate, 0.3 g. All ingredients are filled with α-cellulose to 100 g, which is sourced from Jinan Hailong Biotech Co., Ltd., Jinan, Shandong, China.

³Mineral premix (per 100 g premix): CoCl₂·6H₂O, 0.14 g; KH₂PO₄, 21.5 g; KCl, 2.8 g; MnSO₄·6H₂O, 0.143 g; Ca(H₂PO₄)₂, 26.5 g; CaCO₃, ZnSO₄·7H₂O, 0.476 g; AlCl₃·H₂O, 0.024 g; 10.5 g; MgSO₄·7H₂O, 10.0 g; KI, 0.023 g; CuCl₂·2H₂O, 0.015 g; NaH₂PO₄, 10.0 g; calcium lactate, 16.50 g; Fe-citrate, 1 g. All ingredients are diluted with α-cellulose to 100 g, which is sourced from Jinan Hailong Biotech Co., Ltd., Jinan, Shandong, China.

⁴Organic matter was calculated value, and moisture, crude protein, crude lipid, and soy isoflavones were analyzed values.

Water stability of the diets was assessed to confirm their integrity upon immersion. According to the China National Formula Feeding Standard of Chinese mitten crab (2022), water stability was determined by the leaching rate that must remain below 10% after 30 min immersion. In this study, the leaching rate was determined after 15 min, 30min, 1 h, and 2 h immersion. The results demonstrated that all measured values were below the 10% (Table 2), confirming compliance with feed quality requirements. The specific methodology was as follows: precisely weighed 10.00 g feed samples were placed in pre-weighed nylon mesh bags (20 cm × 15 cm, 0.85 mm pore size) and immersed in a constant-temperature water bath (26 ± 2 °C) for 15 min, 30 min, 1 h, and 2 h. During immersion, uniform oscillation was applied to simulate natural aquatic conditions. Subsequently, feed samples were removed and dried to constant weight at 105 °C (M₁). Untreated control samples were similarly processed to constant weight (M₂). Each group included three replicates.

$$\text{Leaching}\text{rate}(\%)=100\times \left[M_2\left(\mathrm{g}\right)–M_1\left(\mathrm{g}\right)\right]/M_2\left(\mathrm{g}\right)\text{.}$$

Table 2.

Effects of immersion duration on the leaching rates of experimental diets (%).

Time	Groups1				SEM	P-value
	C	SL	SM	SH		ANOVA	Linear	Quadratic
15 min	3.91	4.08	3.88	3.58	0.124	0.610	0.310	0.391
30 min	5.07	5.23	5.17	5.04	0.106	0.932	0.886	0.800
1 h	6.35	6.29	6.28	6.48	0.130	0.962	0.751	0.865
2 h	8.02	7.95	7.98	8.14	0.146	0.979	0.766	0.902

SEM = standard error of the mean.

¹C, control diet; SL, SM, and SH, control diet with 32.51, 70.83, and 369.03 mg/kg soy isoflavones, respectively; n = 3.

The proximate composition of the diet was determined according to 2005. Moisture content was determined via drying the diets in an oven (645, Thermo Fisher Scientific Inc., Wilmington, DE, USA) at 105 °C until a constant mass was reached (method 934.01). Crude protein was analyzed via the Kjeldahl method (method 988.05) with a Kjeldahl apparatus (KjeltecTM 8200，FOSS Analytical A/S, Hillerød, Zealand, Denmark), and crude lipids via the chloroform/methanol method (method 983.23) extraction followed by drying in a vacuum oven (DZF-6030TH, Shanghai Jinghong Experimental Equipment Co., Ltd., Shanghai, China) to constant weight. Ash content was determined via a muffle furnace (F6030CM-33, Thermo Fisher Scientific Inc., Wilmington, DE, USA) at 550 °C for 6 h (method 924.05). Organic matter (% dry matter) was calculated as the difference between dry sample weight and ash weight.

2.3. Feeding trial

The feeding trial was conducted at the Experimental Base of the Shanghai Fisheries Research Institute (Shanghai, China). The same batch of second-year adult female E. sinensis was purchased from Nantong Haida Biotechnology Co., Ltd. (Nantong, Jiangsu, China) in late August following reproductive molting. Crabs were acclimated for seven days in a concrete tank (9 m × 2 m). The substandard or shipping damaged individuals were culled, and fed a basal diet to stabilize physiological status and adapt indoor experimental environment. During this period, 20 crabs with similar size (102.10 ± 3.74 g) were randomly selected and dissected for morphological and histological observations of the ovaries. The gonadosomatic index (GSI) was 0.29 ± 0.02, and the ovarian development stage classified as stage Ⅰ according to Wu et al. (2017).

After the acclimation period, 192 healthy crabs (101.52 ± 4.57 g) were randomly distributed into four groups and assigned to 24 tanks (300 L, bottom external dimensions: 71.0 cm × 48.5 cm, height: 61.5 cm), and six replicates per group and eight crabs per replicate (n = 6). Each tank was added with 200 L natural river water, and an airstone was installed to maintain dissolved oxygen levels through continuous aeration. Six arched plastic tubes were positioned in tank to mitigate fighting behavior. Crabs were fed at 3% of body weight at 07:00, 14:30, and 20:30 every day for 11 weeks. As it was a static water system, daily tank maintenance included removing leftover feed and feces, collecting and weighing dead crabs, and 1/3 of the water in each tank was replaced daily to maintain water quality.

Natural river water was collected at 06:00 from a fixed location each time, and allowed to settle in concrete pond (10 m × 5.5 m × 2 m) under dark conditions for four days. It was then treated with sand filtration and ultraviolet sterilization to remove suspended particles and microorganisms. Subsequently, water was transferred to the reservoir (9 m × 2 m) and continuously aerated for two days to remove residual chlorine and increase dissolved oxygen levels. At the same time, the water temperature was maintained at 24 to 26 °C using a constant temperature equipment (CW2500A-2, Xingcheng Mechanical and Electrical Aquatic Products Co., Ltd., Chaozhou, Guangdong, China). Finally, each batch of river water was tested water quality parameters prior to use. As follows: pH 7.3 to 8.1, dissolved oxygen levels above 7.0 mg/L and nitrite concentrations below 0.3 mg/L, ammonia-N concentrations below 0.05 mg/L. All tanks were supplied with water from the same reservoir processed as described above, ensuring consistent water quality conditions. During the feeding trial, the photoperiod was maintained at 12 h of light and 12 h of dark. Daily water temperature for each tank was provided in Table S2. The range was 23 to 28 °C, which represented the overall variation recorded throughout the feeding trial period, rather than the sharp daily fluctuations. Tank temperature (23-28 °C) fell entirely within the optimal range for gonadal development in adult crabs (22-30 °C) (Shen et al., 2023). Additionally, all experimental groups were maintained under identical environmental conditions and experienced the same temperature variation pattern. Therefore, this temperature variation was within their normal physiological buffering capacity and is unlikely to decisively affect the results.

2.4. Sample collection

After feeding trial, the numbers and weights were recorded. After fasting for 24 h, crabs were anesthetized on slurry ice and six crabs in each group were randomly selected to draw the hemolymph. The extracted hemolymph was then centrifuged at 4 °C and 3300 × g for 10 min, after which resulting supernatant was harvested and stored at −80 °C. The ovaries and hepatopancreas were quickly dissected and separated to determine the GSI and hepatopancreas index (HSI). A small sample of ovary was excised and immersed in the Bouin’s reagent for histological processing. The residual hepatopancreas and ovaries were immediately immersed in liquid nitrogen, and then stored at −80 °C.

2.5. Calculation

$$\text{Weight}\text{gain}\left(\text{WG},\%\right)=100\times \left[\text{Final}\text{body}\text{weight}\left(\mathrm{g}\right)–\text{Initial}\text{body}\text{weight}\left(\mathrm{g}\right)\right]/\text{Initial}\text{body}\text{weight}\left(\mathrm{g}\right)\text{;}$$

$$\text{HSI}(\%)=100\times \text{Hepatopancreas}\text{wet}\text{weight}\left(\mathrm{g}\right)/\text{Body}\text{wet}\text{weight}\left(\mathrm{g}\right)\text{;}$$

$$\text{GSI}(\%)=100\times \text{Gonad}\text{wet}\text{weight}\left(\mathrm{g}\right)/\text{Body}\text{wet}\text{weight}\left(\mathrm{g}\right)\text{;}$$

$$\text{Survival}(\%)=100\times \left(\text{Final}\text{crab}\text{number}/\text{Initial}\text{crab}\text{number}\right)\text{;}$$

$$\text{Feed}\text{intake}\left(\mathrm{g}/\mathrm{d}\text{per}\text{crab}\right)=\left[\text{Total}\text{feed}\text{weight}\left(\mathrm{g}\right)/\text{Final}\text{crab}\text{numbers}\right]/77\left(\mathrm{d}\right)\text{;}$$

$$\text{Feed}\text{conversion}\text{ratio}=\text{Total}\text{feed}\text{weight}\left(\mathrm{g}\right)/\left[\text{Final}\text{crab}\text{weight}\left(\mathrm{g}\right)–\text{Initial}\text{crab}\text{weight}\left(\mathrm{g}\right)\right]\text{.}$$

2.6. Biochemical analysis

Using kits from the Jiancheng Biological Engineering Institute (Nanjing, Jiangsu, China) to analyze the total cholesterol (T-CHO; A111-1-1) and triglyceride (TG; A110-1-1) contents in the hemolymph, hepatopancreas and ovaries. The low-density lipoprotein cholesterol (LDL-C; A113-1-1) and high-density lipoprotein cholesterol (HDL-C; A112-1-1) contents were also analyzed in the hemolymph. The contents of TG and T-CHO were analyzed by the glycerophosphate oxidase–peroxidase method. The color intensity of the quinone compounds was directly proportional to content, and absorbance was measured at 500 and 510 nm respectively using a microplate spectrophotometer (Epoch2, BioTek Instruments, Inc., Winooski, VT, USA). The contents of LDL-C and HDL-C were enzymatically hydrolyzed and oxidized to generate hydrogen peroxide, which subsequently reacted with a chromogenic agent to form red benzoquinone pigment, and absorbance was measured at 600 and 550 nm, respectively.

Using ELISA kits from the Shanghai Enzyme Biotechnology Company (Shanghai, China) analyzed the E₂ (YJ263589), VTG (YJ265896), testosterone (YJ285002), progesterone (PROG; YJ265890), and FSH (YJ265895) contents in the hemolymph, the aromatase (YJ262010) and cyclic adenosine monophosphate (cAMP; YJ256021) contents in the ovary, and the gonadal inhibiting hormone (GIH; YJ290712) content in the eyestalk. Target substance concentrations were determined by comparing the sample and standard sample absorbances on the standard curve. First, a standard curve was established using the absorbance and concentrations of the standard samples. Sample concentrations were then calculated by substituting their absorbance values into the standard curve equation and multiplying by the dilution factor.

2.7. Analysis of gene expression (quantitative real-time PCR [qPCR])

The total RNA of the ovary and hepatopancreas was isolated via RNAiso Plus (9109, Takara Biotechnology Co., Ltd., Dalian, Liaoning, China). The total RNA quality and concentration were estimated via 1% agarose gel electrophoresis and a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific Inc., Wilmington, DE, USA). Only samples exhibited clear electrophoretic bands and meeting the 1.8 to 2.0 standard were reverse transcribed via PrimeScript RT Reagent Kit (RR037B, Takara Biotechnology Co., Ltd., Dalian, Liaoning, China). The PCR system and procedure were described in the prior study (He et al., 2024), and RT–qPCR was conducted via a CFX96 Real-Time PCR system (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The relative expression levels of the following genes were detected: vtg, vgr, esrrb, star, hsd3b2, and hsd17b6. The actb and rps27 served as housekeeping genes, vtg gene in the hepatopancreas and ovary shared same transcript (Table 3). The relative expression levels of mRNA were quantified via the 2^−ΔΔCt method (Livak and Schmittgen, 2001).

Table 3.

Information of the primers used for quantitative real-time PCR.

Genes	Position	Primer sequences (5′–3′)	References
vtg	Forward	AAGGTCCGCAGCAAGCAGAT	Bu et al. (2023)
	Reverse	GGCGAGGCACGAGGTAGAAT
vgr	Forward	GCAACGCCTTCCTTCTGGTA	Bu et al. (2023)
	Reverse	GGCACGGTGTTCGCTATCAT
esrrb	Forward	CTACTATGAGGTGAGCGGCG	GenBank: XM_050859816.1
	Reverse	CATGTTGCTTGGCAGCGTTA
star	Forward	TACTATGGCCACGGGGAAGA	GenBank: XM_050864147.1
	Reverse	CCAGTCAGCGACATCACAGT
hsd3b2	Forward	GTCTGGTCGCATATCGGGTT	GenBank: XM_050860983.1
	Reverse	CGATTCGGGCCAAGAGTAGG
hsd17b6	Forward	GAACTGGGGCAAGAACGAGA	GenBank: XM_050840838.1
	Reverse	GCGTACCCTAGCCCTCTACT
actb	Forward	TCGTGCGAGACATCAAGGAAA	Lin et al. (2021)
	Reverse	AGGAAGGAAGGCTGGAAGAGTG
rps27	Forward	CCCCCAAGAAGATCAAGCACA	Lin et al. (2021)
	Reverse	CAGATGGCAGCGACCACAGTA

2.8. Western blotting

Radioimmunoprecipitation assay lysis buffer (P0013B, Beyotime Biotechnology Co., Ltd., Shanghai, China) and phenylmethanesulfonyl fluoride protease inhibitor (ST506, Beyotime Biotechnology Co., Ltd., Shanghai, China) were mixed at a ratio of 100:1 to form a mixed lysis mixture. Ovarian tissue and lysate were mixed and homogenized at a ratio of 1:15 (weight: volume). After being lysed on ice for 30 min, the mixture was centrifuged (12,000 × g, 4 °C, 10 min) to obtain clarified supernatant. Total protein concentration was analyzed via a bicinchoninic acid (BCA) protein assay kit (P0010S, Beyotime Biotechnology Co., Ltd., Shanghai, China). The supernatant and 5 × Sodium Dodecyl Sulfate buffer (P0015L, Beyotime Biotechnology Co., Ltd., Shanghai, China) were subsequently diluted and mixed at a ratio of 4:1, boiled in a water bath (100 °C, 10 min) to fully denature the protein and then stored at −80 °C for future use. A 10% Polyacrylamide Gel Electrophoresis gel (PG212, Epizyme Biomedical Technology Co., Ltd., Shanghai, China) was prepared, then protein marker (10-200 kDa, G2058, Wuhan Servicebio Technology Co., Ltd., Wuhan, Hubei, China) and 30 μg of protein sample were added to the sample well. The proteins were isolated via a Bio-Rad Mini-PROTEAN Tetra Cell electrophoresis apparatus for 1.0 mm gels (Bio-Rad Laboratories, Inc., Hercules, CA, USA) for 90 min and transferred to a nitrocellulose membrane (HATF00010, Merck Millipore, Billerica, MA, USA). Then, it was blocked with rapid blocking buffer ( PS108P, Epizyme Biomedical Technology Co., Ltd., Shanghai, China) for 15 min. Samples were subsequently incubated with primary antibodies at 4 °C for 12 h: CREB (1:1000; ET1601-15, HuaAn Biotechnology Co., Ltd., Hangzhou, Zhejiang, China), p-CREB (1:1000; ET7107-93, HuaAn Biotechnology Co., Ltd., Hangzhou, Zhejiang, China), PKA (1:1000; P010098, Epizyme Biomedical Technology Co., Ltd., Shanghai, China), p-PKA (1:1000; R015300, Epizyme Biomedical Technology Co., Ltd., Shanghai, China). Using Tris-Buffered Saline with Tween 20 to wash the samples for 30 min, they were incubated with the corresponding secondary antibodies (1:10,000; ab175775 and ab175773, Abcam Plc, Cambridge, UK) for 1 h. Then, the images were scanned and photographed using an Odyssey CLx Imager (LI-COR, Inc., Lincoln, NE, USA) fluorescence imaging system. Subsequently, the membrane was stripped with stripping buffer (P0025, Beyotime Biotechnology Co., Ltd., Shanghai, China) for 10 min, and reblotted with GAPDH (1:3000; GB15002-100, Wuhan Servicebio Technology Co., Ltd., Wuhan, Hubei, China). The immunoreactive band intensity was measured via ImageJ software (v1.8.0, National Institutes of Health, Bethesda, MD, USA).

2.9. Histological analysis

Following 24 h immersion fixation in the Bouin’s solution, the ovarian tissues were dehydrated, cleared with ethanol and xylene solution step by step prior to paraffin embedding. Each sample slide contained three consecutive sections (5 μm in thickness). These sections were prepared using a microtome and stained with hematoxylin and eosin. Six fields per slide were randomly observed under a microscope (BX51, Olympus Corporation, Tokyo, Japan), and the nucleus to cytoplasmic ratio (N/C ratio), average number and volume of yolk granules per oocyte were calculated via ImageJ software.

$$\mathrm{N}/\mathrm{C}\text{ratio}=\text{Nucleus}\text{volume}\left(\mathrm{\mu }{\mathrm{m}}^3\right)/\text{Cell}\text{volume}\left(\mathrm{\mu }{\mathrm{m}}^3\right)\text{;}$$

$$\text{Volume}\left(\mathrm{\mu }{\mathrm{m}}^3\right)=0.523\times \text{Maximum}{\text{width}}^2\left(\mathrm{\mu }\mathrm{m}\right)\times \text{Maximum}\text{length}\left(\mathrm{\mu }\mathrm{m}\right)\text{.}$$

2.10. Statistical analysis

Data are expressed as the means and standard errors of the means (SEM). First, all the data were tested for normality (Shapiro–Wilk test) and homogeneity (Levene’s test) of variance, and then one-way ANOVA and Duncan’s multiple range tests were used for differences between groups. P < 0.05 represented significant differences. Additionally, orthogonal polynomial contrasts were performed to determine whether the results were linear and/or quadratic. All analyses were performed via SPSS version 23.0 (SPSS Inc., Chicago, IL, USA).

The ANOVA employed the following statistical model:

$$Y_{ij}=\mu +{\alpha }_i+{\epsilon }_{ij}\text{,}$$

where Y_ij is the j-th observation within the i-th group; μ is the grand mean; α_i is the effect of the i-th treatment (soy isoflavone levels); ε_ij is the random error.

3. Results

3.1. Growth performance and feed utilization

The highest GSI was observed in the SL group, with GSI in the SM group being significantly greater than that in the SH group (P = 0.004; Table 4). As shown in Fig. 1 and Table 5, the N/C ratio in the SL and SM groups was significantly lower than that in the C and SH groups (P < 0.001). Furthermore, the number and volume of yolk granules were significantly greater in the SL and SM groups than in the C group (P < 0.001). An initial decrease followed by an increase in the HSI was observed across groups, but no significant differences that was similar to the WG, feed intake and feed conversion ratio (P > 0.05; Table 4).

Table 4.

The growth performance and feed utilization of female Eriocheirsinensis that fed different soy isoflavones levels diets .

Items	Groups1				SEM	P-value
	C	SL	SM	SH		ANOVA	Linear	Quadratic
IBW, g	103.08	99.75	101.67	102.76	0.595	0.187	0.866	0.179
FBW, g	113.00	109.75	112.00	112.25	0.733	0.474	0.921	0.523
WG, %	9.63	10.02	10.13	9.26	0.352	0.843	0.759	0.666
HSI, %	5.06	4.56	4.86	5.35	0.152	0.330	0.401	0.192
GSI, %	8.88bc	10.75a	9.70ab	8.47c	0.268	0.004	0.469	0.004
FI, g/d per crab	2.28	2.14	2.23	2.18	0.064	0.876	0.691	0.865
FCR	2.68	2.72	2.51	2.57	0.043	0.296	0.180	0.406
Survival, %	81.25	87.50	85.42	81.25	2.304	0.730	0.921	0.541

IBW = initial body weight; FBW = final body weight; WG = weight gain; HSI = hepatopancreas index; GSI = gonadosomatic index; FI = feed intake; FCR = feed conversion ratio; SEM = standard error of the mean.

Values in the same row with different superscript letters are significantly different (P < 0.05).

¹C, control diet; SL, SM, and SH, control diet with 32.51, 70.83, and 369.03 mg/kg soy isoflavones, respectively; n = 6.

Fig. 1.

The hematoxylin and eosin staining of ovarian tissues. The scale represents 100 μm (20×). C, control diet; SL, SM, and SH, control diet with 32.51, 70.83, and 369.03 mg/kg soy isoflavones, respectively.

Table 5.

The ovarian histological analysis of female Eriocheirsinensis that fed different soy isoflavones levels diets.

Items	Groups1				SEM	P-value
	C	SL	SM	SH		ANOVA	Linear	Quadratic
N/C ratio	0.0037a	0.0008c	0.0017b	0.0032a	0.00036	<0.001	0.877	0.001
Yolk granule numbers	312.67c	438.33a	353.67b	321.00c	15.739	<0.001	0.692	0.018
Volume of yolk granule, μm3	637.29b	842.09a	876.38a	644.06b	33.600	<0.001	0.866	0.001

N/C ratio = the nuclear to cytoplasmic ratio; SEM = standard error of the mean.

Values in the same row with different superscript letters are significantly different (P < 0.05).

¹C, control diet; SL, SM, and SH, control diet with 32.51, 70.83, and 369.03 mg/kg soy isoflavones, respectively; n = 6.

3.2. Reproductive hormones and enzymatic activity parameters

The levels of VTG and aromatase in the SL and SM groups was significantly greater than that in the C and SH groups (P < 0.05), whereas testosterone content was significantly lower (P < 0.001; Table 6). The 17β-E₂ and FSH contents were significantly greater in the SL and SM groups than in the SH group (P < 0.01). The levels of PROG and cAMP in the SL group were significantly greater than that in the C group (P < 0.05), whereas the GIH content was significantly lower (P < 0.001), with GIH reaching its maximum value in the SH group.

Table 6.

The reproductive hormones and enzymatic activity parameters of female Eriocheirsinensis that fed different soy isoflavones levels diets.

Items	Groups1				SEM	P-value
	C	SL	SM	SH		ANOVA	Linear	Quadratic
VTG, ng/mL	554.50c	641.26a	602.33b	545.82c	10.545	<0.001	0.510	<0.001
17β-E2, pmol/L	67.88bc	75.46a	70.21b	64.13c	1.211	<0.001	0.132	0.001
Testosterone, ng/mL	16.43a	13.37b	14.82b	16.77a	0.412	<0.001	0.520	0.001
FSH, U/L	12.558bc	13.87a	13.37ab	12.01c	0.225	0.003	0.303	0.001
PROG, ng/mL	16.43b	17.94a	17.16ab	16.24b	0.248	0.040	0.556	0.027
GIH, μg/mL	232.83b	205.55c	217.58bc	266.06a	6.618	<0.001	0.056	<0.001
Aromatase, IU/L	115.62c	136.39a	123.98b	110.79c	2.80	0.001	0.170	0.001
cAMP, nmol/mL	11.52b	12.38a	11.93ab	11.50b	0.143	0.027	0.540	0.034

VTG = vitellogenin; 17β-E₂ = 17β-estradiol; FSH = follicle-stimulating hormone; PROG = progesterone; GIH = gonadal inhibiting hormone; cAMP = cyclic adenosine monophosphate.

Values in the same row with different superscript letters are significantly different (P < 0.05).

¹C, control diet; SL, SM, and SH, control diet with 32.51, 70.83, and 369.03 mg/kg soy isoflavones, respectively; n = 6.

3.3. Gene expression

Dietary supplementation with 32.51 and 70.83 mg/kg soy isoflavones significantly induced the expression of vtg genes in the hepatopancreas and ovaries (P < 0.001), with a more prominent regulatory effect in the hepatopancreas (Fig. 2 A and B). Concurrently, the expression of genes such as esrrb, star, hsd3b2, and hsd17b6 in the ovaries was significantly upregulated (P < 0.05; Fig. 2D–G). The expression of vgr gene was significantly upregulated only in the SL group (P = 0.039; Fig. 2C).

Fig. 2.

Effects of dietary soy isoflavones on the gene expression of female Eriocheirsinensis. (A) The mRNA levels of vtg in the hepatopancreas. The mRNA levels of vtg (B), vgr (C), esrrb (D), star (E), hsd3b2 (F), and hsd17b6 (G) in the ovary. C, control diet; SL, SM, and SH, control diet with 32.51, 70.83, and 369.03 mg/kg soy isoflavones, respectively. Different lowercase letters above columns represent significant differences among treatments at P < 0.05.

3.4. Biochemical parameters

Compared with the control diet, dietary supplementation with 32.51 mg/kg soy isoflavones increased significantly the TG, T-CHO, LDL-C, and HDL-C contents in the hemolymph (P < 0.05), whereas the T-CHO content in the hemolymph of the SH group decreased significantly (P = 0.001; Table 7). Additionally, the T-CHO content in the ovaries increased significantly in the SL group (P = 0.002). The content of TG in the hepatopancreas increased linearly and significantly with increasing levels of soy isoflavone supplementation (P < 0.001). The level of T-CHO in the SL group showed significantly higher than the other groups, whereas the SH group had significantly lower levels (P < 0.001). The content of TG in the ovaries was not significantly affected by supplementation with soy isoflavones (P > 0.05).

Table 7.

Biochemical profiles of female Eriocheirsinensis that fed different soy isoflavones levels diets.

Items	Groups1				SEM	P-value
	C	SL	SM	SH		ANOVA	Linear	Quadratic
Hemolymph, mmol/L
TG	0.08b	0.14a	0.08b	0.07b	0.009	0.007	0.255	0.173
T-CHO	0.38b	0.55a	0.332bc	0.23c	0.036	0.001	0.032	0.008
LDL-C	0.11b	0.15a	0.12ab	0.11b	0.007	0.046	0.307	0.078
HDL-C	0.07b	0.10a	0.09ab	0.07b	0.004	0.046	0.921	0.028
Hepatopancreas, mmol/g
TG	0.40d	0.62c	0.77b	0.96a	0.058	<0.001	<0.001	<0.001
T-CHO	0.10b	0.14a	0.10b	0.05c	0.008	<0.001	0.019	<0.001
Ovary, mmol/g
TG	0.11	0.11	0.10	0.10	0.003	0.731	0.397	0.664
T-CHO	90.59bc	100.69a	97.06ab	82.97c	2.117	0.002	0.185	0.001

TG = triglyceride; T-CHO = total cholesterol; LDL-C = low-density lipoprotein cholesterol; HDL-C = high-density lipoprotein cholesterol; SEM = standard error of the mean.

Values in the same row with different superscript letters are significantly different (P < 0.05).

¹C, control diet; SL, SM, and SH, control diet with 32.51, 70.83, and 369.03 mg/kg soy isoflavones, respectively; n = 6.

3.5. cAMP/PKA/CREB signaling pathway in the ovaries

The p-CREB protein levels in the SL and SM groups were significantly greater than those in the SH group (P = 0.019), but showed no significant difference from the control group (P > 0.05; Fig. 3 A and B). Except for those in the 369.03 mg/kg group, the p-PKA protein level of the other two groups were significantly greater than those in the C group, with the highest level observed in the SL group (P < 0.001; Fig. 3C).

Fig. 3.

Effects of dietary soy isoflavones on the cAMP/PKA/CREB signaling pathway in the ovary of female Eriocheirsinensis. (A) The protein expression levels of CREB, p-CREB, PKA, and p-PKA. (B) Normalized protein expression of p-CREB relative to CREB. (C) Normalized protein expression of p-PKA relative to PKA. C, control diet; SL, SM, and SH, control diet with 32.51, 70.83, and 369.03 mg/kg soy isoflavones, respectively. Different lowercase letters above columns represent significant differences among treatments at P < 0.05.

3.6. Mantel analysis

As shown in Fig. 4, the GSI was significantly correlated with the contents of VTG and testosterone in the hemolymph, the expression level of hsd17b6 gene and p-PKA/PKA protein ratio in the ovary (r ≥ 0.25, P ≤ 0.001). Additionally, the GSI also showed significant correlations with the following indicators (P ≤ 0.05): the contents of E₂, TG, T-CHO, and HDL-C in the hemolymph; the content of T-CHO in the hepatopancreas; as well as yolk granule deposition (N/C ratio, the number and volume of yolk granules, the expression level of vtg in the hepatopancreas and ovary, and the expression level of vgr in the ovary), steroid hormone synthesis (the expression level of esrrb and hsd3b2, and the content of aromatase) and p-CREB/CREB protein ratio in the ovary.

Fig. 4.

A Mantel test was conducted between the GSI and other indices. The size and color of the squares in the matrix represent the Pearson coefficient value and positive and negative correlations, respectively. The color of the lines outside the matrix represents statistical significance, and the thickness of the lines represents the strength of the correlation between the GSI and these variables. HSI = hepatopancreas index; GSI = Gonadosomatic index; N/C ratio = nucleus to cytoplasmic ratio; VTG = vitellogenin; 17β-E₂ = 17β-estradiol; T = testosterone; FSH = follicle-stimulating hormone; PROG = progesterone; GIH = gonadal inhibiting hormone; cAMP = cyclic adenosine monophosphate; TG = triglyceride; T-CHO = total cholesterol; LDL-C = low-density lipoprotein cholesterol; HDL-C = low/high-density lipoprotein cholesterol.

4. Discussion

In this study, soy isoflavones (32.51 and 70.83 mg/kg) increased the GSI and VTG levels of crabs. This finding is consistent with findings in female Channel catfish (Ictalurus punctatus) (Kelly and Green, 2006), female Rainbow trout (Oncorhynchus mykiss) (Bennetau-Pelissero et al., 2001), and Medaka (Oryzias latipes) (Scholz et al., 2004), suggesting that isoflavones can stimulate the VTG synthesis. Contrary findings have also been reported, indicating that isoflavones may reduce the GSI or produce no significant effect in female aquatic animals (And et al., 1999; Bagheri et al., 2013; Bennetau-Pelissero et al., 2001). Such inconsistencies could stem from species, sex, developmental stage, and the dose and composition of isoflavones (Inaba et al., 2022).

As an omnivorous crustacean, E. sinensis consumes phytogenic foods under natural conditions such as hornwort (Ceratophyllum demersum), western waterweed (Elodea nuttallii), and tape grass (Vallisneria spiralis). These aquatic plants contain various polyphenols, including phenolic acids, terpenes, and flavones (Gao et al., 2016). This speculates that E. sinensis may have a stronger tolerance to phytoestrogens or a more efficient biotransformation process. Additionally, all soy isoflavones doses used in this study fell within safe ranges. Crustaceans may enhance isoflavones excretion efficiency through metabolic adaptive capacities shaped by long-term evolution. In contrast, carnivorous aquatic animals metabolize phytoestrogens slowly, leading to their accumulation in tissues (Fajkowska et al., 2021). Although the high dose of soy isoflavones did not have obvious adverse effects on gonadal development, the significant quadratic effect indicates that further increases in dose could lead to adverse reactions.

Dietary supplementation with soy isoflavones did not significantly affect the WG. This may be attributed to crabs that after reproductive molting, which shifted energy allocation from growth to gonadal development (He et al., 2014). The N/C ratio of oocytes gradually decreases as ovarian development progresses (Wu et al., 2017). Histological observations of the oocytes revealed that soy isoflavones decreased significantly the N/C ratio and increased the number and volume of yolk granules. These phenotypic changes suggest that an appropriate dose of soy isoflavones can increase yolk granule deposition in E. sinensis.

This study demonstrated a more pronounced enhancement of exogenous the VTG biosynthesis by soy isoflavones. Specifically, the vtg gene upregulation in the hepatopancreas was more significant compared to the ovaries. During the ovarian development of E. sinensis, the E₂ concentration in the hemolymph, hepatopancreas and ovaries increases significantly from stages Ⅰ to Ⅲ and then gradually decreases until stage Ⅴ (Pan, 2018). In this study, the GSI values ranged from 8.47 to 10.75, with ovarian development occurring at stages Ⅳ to Ⅴ. Soy isoflavones enhanced significantly the VTG and E₂ levels in the hemolymph. Therefore, these findings indicate that soy isoflavones promotes yolk granules deposition by enhancing the exogenous VTG synthesis and endogenous E₂. Correlation analysis revealed a significant positive association between the VTG and E₂ levels, and the upregulation of vgr gene further supported this hypothesis.

Soy isoflavones increased the TG content of the hepatopancreas in a dose-dependent manner, and significantly elevated the TG levels of the hemolymph. In contrast, the TG content in the ovary remained unchanged. This could be because the TG is degraded into fatty acids, and soy isoflavones more effectively promote β-oxidation to provide energy for ovarian development. Previous research revealed that soy isoflavones have the ability to enhance lipid oxidation and transport (He et al., 2024). Therefore, soy isoflavones can increase the TG oxidation for energy, synergistically stimulating ovarian development in E. sinensis.

Compared with estrogen, soy isoflavones exhibit a lower affinity for ERs, consequently resulting in weaker efficacy to regulate the target genes transcription (Nuzaiba et al., 2020). Although arthropods have lost ERs during evolution, a structurally and sequentially similar estrogen-related receptors (ERR) has been identified in crustaceans (Park et al., 2017; Pan, 2018; Thornton and J., 2003). In this study, a low dose of soy isoflavones significantly upregulated esrrb gene expression, whereas high doses showed no significant difference from the control group. As mentioned earlier, this may be due to the low E₂ levels, allowing soy isoflavones to exert estrogenic effects. However, 369.03 mg/kg high dose caused a mixture of both effects, leading to mutual cancellation and exhibiting a bidirectional regulatory effect. This phenomenon is also consistent with previous results (Nuzaiba et al., 2022).

Additionally, soy isoflavones significantly increased the PROG level and decreased the GIH level in hemolymph, synergistically regulating the VTG synthesis. In terms of cholesterol metabolism, dietary supplementation with 32.51 mg/kg soy isoflavones significantly increased the T-CHO levels in all tissues, as well as the LDL-C and HDL-C levels in the hemolymph. It also upregulated the expression of genes related to key enzymes for steroid synthesis (hsd3b2 and hsd17b6) and cholesterol transport (star) in the ovary. This indicates that soy isoflavones coordinated the cholesterol transport metabolism and steroid hormone synthesis processes, and exerted regulatory effects on ovarian development through the E₂/ERR signaling pathway.

The FSH acts as the main trigger for aromatase expression within ovarian granulosa cells, with the cAMP acting as the primary second messenger. As a key responsive factor of the cAMP/PKA signaling pathway, CREB requires phosphorylation to activate aromatase gene transcription (Guo et al., 2022; Somers et al., 1999). This study found that 32.51 mg/kg soy isoflavones increased significantly the FSH content, and both the 32.51 and 70.83 mg/kg doses enhanced significantly aromatase activity. Moreover, ELISA and Western blot results revealed that soy isoflavones increased significantly the cAMP level, p-PKA and p-CREB protein levels. Therefore, it is hypothesized that an appropriate dose of soy isoflavones may stimulate FSH secretion via neuroendocrine regulation, subsequently activating the cAMP/PKA/CREB signaling pathway to modulate the E₂ synthesis and secretion, thereby enhancing ovarian development in E. sinensis. However, the mechanisms through which soy isoflavones stimulate the FSH secretion require further investigation. A dose of 32.51 mg/kg was identified as an effective level of soy isoflavones. However, further investigations with more dose gradients (e.g., five or more levels) are required to determine the optimal supplementation range.

5. Conclusions

A dose of 32.51 mg/kg soy isoflavones could regulate the endocrine status and steroid hormone synthesis in E. sinensis via the E₂/ERR and cAMP/PKA/CREB signaling pathways, inducing the VTG synthesis and promoting ovarian development (Fig. 5). In contrast, the 369.03 mg/kg soy isoflavones did not cause any adverse effects, but the significant quadratic effect suggests that increasing the dose further may not be advisable. This indicates that the actual biological effects of soy isoflavones may be underestimated when soybean meal is used as a protein source in fattening feed. In summary, this study confirms that soy isoflavones can serve as a functional feed additive to promote ovarian development. These results provide a feasible strategy for improving reproductive development in crustaceans, and provide a theoretical basis for the scientific application of phytoestrogens in aquafeeds. However, further research is needed to quantify the optimal supplementation range and validate exact mechanisms via cellular and tissue models.

Fig. 5.

Schematic illustration of the potential mechanism by which soy isoflavones affect ovarian development in adult female Eriocheirsinensis. Optimal levels of soy isoflavones (32.51 mg/kg) promoted the VTG synthesis in the hepatopancreas and ovaries by binding to ERR and regulating the endocrine status. On the other hand, the synthesis of steroid hormone (E₂) is regulated through the cAMP/PKA/CREB signaling pathway, which jointly promotes ovarian development. Conversely, excessive soy isoflavones may competitively bind to the ERR with E₂, resulting in a slowing of vtg transcription and translation rates. Although no adverse effects were observed in this study, this competitive binding could hinder ovarian development. E₂ = estradiol; ERR = estrogen-related receptors; ERE = estrogen response element; VTG = vitellogenin; FSH = follicle-stimulating hormone; PROG = progesterone; GIH = gonadal inhibiting hormone; cAMP = cyclic adenosine monophosphate; TG = triglyceride; T-CHO = total cholesterol; LDL-C = low-density lipoprotein cholesterol; HDL-C = low/high-density lipoprotein cholesterol.

CRediT authorship contribution statement

Long He: Writing – review & editing, Writing – original draft, Methodology, Investigation, Data curation, Conceptualization. Jinping Li: Investigation, Formal analysis, Data curation. Dexiang Cao: Resources. Zhijun Liu: Resources. Chuanjie Qin: Writing – review & editing, Resources, Conceptualization. Xiaodan Wang: Methodology, Investigation, Conceptualization. Jianguang Qin: Writing – review & editing, Methodology, Investigation. Erchao Li: Writing – review & editing, Supervision, Resources, Conceptualization. Liqiao Chen: Supervision, Resources, Methodology, Investigation, Conceptualization.

Declaration of competing interest

We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work, and there is no professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the content of this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article: