Transcriptome profiling of banana shrimp (Fenneropenaeus merguiensis) ovaries and testes: Insights into FoxL2

Abstract

The banana shrimp is found in the Pacific and Indian Oceans. Female shrimp are preferred for consumption because they are larger than males. Understanding the mechanism of sex differentiation is important for developing techniques to increase the number of female shrimp for economic benefits. This study investigates the reproductive development of F. merguiensis using transcriptome analysis. Sxl2, dsx, AGH, FEM-1, and Nrg-X2 were classified as essential genes for testes development during the juvenile stage. Several genes were required for both juvenile and adult male development. Additionally, the expression of several genes was shown to be required for juvenile and adult ovarian development, including SOP1, SOP2, Ptgr1, EST, Vgr, Vmol1, and TR-beta A. Interestingly, high levels of FoxL2 expression were observed in the testes, in contrast to previous studies in humans and other mammals. The binding of FoxL2 to the Vtg promoter was demonstrated in silico with the highest relative binding score (RS = 0.89) using the JASPAR program. Knock-down of the FoxL2 gene with dsRNA significantly suppressed FoxL2 at 2, 4, and 6 d. As a result, Vtg expression increased when compared with the control at 2, 4, and 6 d, indicating that FoxL2 plays an important role in Vtg expression in the ovary. Our findings highlight the role of FoxL2 in banana shrimp reproduction and provide valuable information on the genes associated with the F. merguiensis reproductive system.

Introduction

The banana shrimp (Fenneropenaeus merguiensis) is a marine crustacean belonging to the family Penaeidae. It is native to the Pacific and Indian Oceans. Female shrimp are preferred over males for commercial purposes because they grow faster and have a larger body size than males [1, 2]. Therefore, the production of female shrimp is economically important. Understanding the mechanisms underlying sex differentiation in banana shrimp is essential for developing technologies to increase the abundance of female shrimp on the market [3].

Transcriptome analysis is a molecular biology technique used to study gene expression levels in biological samples. It has several benefits; it assists in identifying differentially expressed genes [4], understanding biological processes [5], the functional annotation of genes [6], developing diagnostic tools [7], and identifying alternative splicing events [8]. Recently, transcriptome analysis has been utilized to study the differences in gene expression levels in the reproductive systems of various shrimp species, such as brine shrimp (Artemia franciscana) [9], banana shrimp (F. merguiensis) [10], Pacific white shrimp (Litopenaeus vannamei) [6], Japanese mantis shrimp (Oratosquilla oratoria) [11], and littoral shrimp (Palaemon serratus) [4]. Forkhead box protein L2 (FoxL2) is a transcription factor that is involved in sex determination in female vertebrates [12, 13]. FoxL2 is preferentially expressed in the ovary and plays a crucial role in ovarian differentiation and maintenance by repressing testis-specific genes [14].

The FoxL2 gene plays a role in regulating the proliferation and differentiation of cells during ovarian development in humans [12, 13]. In mammals, FoxL2 plays a vital role in the activation of the follicle-stimulating hormone (FSH) and is expressed in the ovaries, eyelids, pituitary gland, and follicular (granulosa) cells [12, 15–19]. FoxL2 deficiency results in an increase in the expression of the SRY-box transcription factor 9 (Sox-9), doublesex and mab-3-related transcription factor 1 (Dmrt1) genes. In contrast, the expression levels of the wingless-type MMTV integration site family members 4 (Wnt4) and R-spondin 1 (R-Spo1) decreased in mice as a result of FoxL2 deficiency [19–21]. Although these mice possess a female genotype, FoxL2 deficiency leads to the development of a male organism [19–21]. In mammals and fish, FoxL2 plays a role in activating the cytochrome P450, family 19, subfamily A, member 1 (CYP19A1), which participates in the synthesis of estrogen and aromatase [22–26]. The expression of male differentiation genes, including steroidogenic factor-1 (Sf1), gonadal soma-derived factor (Gsdf), and Dmrt1, increased after FoxL2 and CYP19A1 were knocked down. This led to male differentiation [27].

In the olive flounder (Paralichthys olivaceus), the FoxL2 gene has been reported to control the expression of the CYP19A gene in collaboration with the nuclear receptor, subfamily 5, group A, member 2 (Nr5a2). This collaboration leads to the suppression of Dmrt1. However, increasing FoxL2 expression in the male olive flounder did not reverse its sex to female. This finding indicates that increased expression of FoxL2 alone is insufficient for inducing sex reversal [28]. Three homeologs of FoxL2: Cgfoxl2a-B, Cgfoxl2b-A, and Cgfoxl2b-B, have been reported in gibel carp (Carassius gibelio). Cgfoxl2a-B deficiency results in the arrest of complete sex reversal or ovarian development. Furthermore, the complete disruption of Cgfoxl2b-A or Cgfoxl2b-B resulted in germ cell depletion [29].

In crustaceans, FoxL2 of the Pacific white shrimp was highly expressed in the testis [6], but no significant expression was observed in the ovary samples by transcriptome analysis [10]. These findings indicate variations in the expression of FoxL2 between vertebrates and shrimp. FoxL2 knock-down using RNAi in Scylla paramamosain increased vitellogenin (Vtg) expression in the ovary. In addition, the overexpression of FoxL2 in Eriocheir sinensis resulted in a reduction of Vtg expression [30]. FoxL2 binds DEAD (Asp-Glu-Ala-Asp) box RNA helicase 20 (DDX20) along with the fushi tarazu factor 1 (Ftz-F1), which is likely the mechanism behind the observed reduction in Vtg expression levels [30].

In this study, we aimed to explore the differential expression levels of genes in the reproductive system. A transcriptome analysis was conducted on testes and ovary specimens of juvenile and adult banana shrimp to identify genes involved in shrimp reproduction. The results of this analysis were further validated using quantitative real-time PCR (qPCR) on selected genes of interest. High levels of FoxL2 transcripts were detected in the juvenile and adult testes of banana shrimp. This is in contrast to humans and other mammals, where FoxL2 is highly expressed in the ovary [12, 15–19]. Therefore, the role of FoxL2 in banana shrimp was also investigated.

Materials & methods

Sample collection

A juvenile F. merguiensis was obtained 3 days post-larvae (PL3) that was produced using a wild-caught broodstock in a farm in the Nakhon Si Thammarat province of Thailand. The PL3 shrimp were reared in seawater with a salinity of 30 parts per thousand (ppt) at 28–30 °C in a concrete pond at the aquaculture building, Faculty of Science, Prince of Songkla University. They were fed Artemia four times a day until ~PL10. From PL10 to PL15, they were fed a crushed commercial shrimp feed mixed with Artemia four times a day. From PL15 to two months post-larvae, they were fed Artemia mixed with shrimp pellets three times daily, and thereafter, they were fed with shrimp pellets. When the shrimp appeared an external sex organ, the shrimp’s length was measured and used as the juvenile shrimp. The adults F. merguiensis were wild-caught (13–14 cm body length, 14–20 g weight) shrimp before they became broodstock shrimp. They were obtained from the same farm mentioned above and were reared in a concrete tank at the aquaculture building as described above. These shrimps used for the transcriptome experiment were sacrificed by knocking in the ice-cold water for 5–10 min and decapitation, then separated the ovaries and testes for the investigation.

The shrimp used for the Foxl2 dsRNA experiment was the wild-caught adult shrimp with the length of a proximate size (15.61 ± 1.04 cm) and weight (28.48 ± 5.59 g) of the broodstock shrimp. The shrimp were reared for 2 wk to achieve the undeveloped ovary before the experiment. The shrimp in this experiment were anesthetized by placing them on ice for 30 sec. Then the shrimp was quickly injected with the Foxl2 dsRNA and softly put the shrimp back into the water tank. The Foxl2 dsRNA post-injection shrimps for 2, 4, and 6 d were sacrificed as described above for separating the ovaries and testes for the investigation.

Animal ethics statement

All animal experimental procedures were performed under the relevant guidelines and regulations and were approved by the Institutional Animal Care and Use Committee, Prince of Songkla University.

Animal selection for RNA sequencing

The shrimp were randomly killed by soaking in ice for 5 min. The ovary and testis sections were performed H&E staining for histological screening, according to Chimnual and colleagues [31]. Shrimp were defined as juveniles at the first appearance of external gonads and when their body length measured 7–8 cm (3.0–3.8 g) for females containing ovaries with abundant oogonia and few oocytes. The males were 6–7 cm long (2.5–3.6 g), having testes with many spermatogonia and without spermatid cells (Fig 1).

Fig 1. H&E staining section of the testis and ovarian samples used for transcriptome analysis.

Transverse section (5μm thick) of the testis and ovarian samples at the same stage as the sample used for transcriptome sequencing. Sg: spermatogonial stem cells; PSc: primary Spermatocyte; SSc: secondary Spermatocyte St: spermatids; Sz: spermatozoa; Se: Sertoli cell; Og: Oogonia; Oc: oocytes; Fc: follicular cells.

The adults F. merguiensis (13–14 cm body length, 14–20 g weight) were randomly prepared for histological screening as described above. The adult samples were testes containing somatic stem cells, spermatogonia, spermatocyte, spermatid cells, and a more pronounced loop division. The ovary of adult banana shrimp contained oogonia, oocytes, and previtellogenic oocytes < 65 μm in size. The nucleus of each oocyte is composed of prominent granular nucleoli (Fig 1). The three shrimp were sacrificed, dissected for the ovaries and testes, and pooled for RNA extraction and sequencing.

RNA preparation

Total RNA was extracted from the pooled sample (three testes and three ovaries from juvenile and mature shrimp) using TRIzol^® reagent (Thermo Fisher Scientific Inc., CA, USA) according to the manufacturer’s protocol. Briefly, the shrimp tissue (50–100 mg) was homogenized in TRIzol^® reagent (1 mL), chloroform (0.2 mL) was added, and it was incubated for 3 min. The sample was spun at 12,000 x g at 4 °C for 15 min, and the aqueous phase was transferred into a new tube. Isopropanol (0.5 mL) was added to the aqueous phase and spun at 12,000 x g at 4 °C for 10 min. The pellet was washed with cold 75% ethanol (1 mL) and spun at 7,500 x g at 4 °C for 5 min. The RNA pellet was air-dried and suspended in 50 μL RNase-free water. After the total RNA was determined, RNA quality and quantity were analyzed using a 2100 Bioanalyser (Agilent, CA, US).

RNA sequencing and data processing

RNA sequencing and data processing were carried out using BGI (BGI, Shenzhen, China). The BGISEQ-500 transcriptome library construction protocol (BGI, Shenzhen, China) was used to construct a transcriptome library from total RNA. Briefly, mRNA fragments were converted into double-stranded cDNA (dscDNA) by reverse transcription using the N6 random primer. The dscDNA was then subjected to end-repair and 3′ adenylation. Adaptors were then ligated to the 3′ adenylated ends of the cDNA fragments. The purified products were used to perform PCR amplification to increase the number of cDNA templates. The PCR product was heat-denatured to convert dsDNA to ssDNA. The ssDNA was cyclized using splint oligo and DNA ligase. Sequencing was performed using the BGISEQ-500 platform (BGI, Shenzhen, China).

Raw reads were then filtered, and de-novo assembly, functional annotation, and differentially expressed gene (DEG) detection were performed using BGI. Briefly, adaptor sequences were removed from the reads; and reads with more than 5% unknown bases (N) and low-quality reads (reads were considered low-quality if 20% of that read, or more, scored less than 15 for quality) were removed from the transcript data to obtain clean reads. De-novo assembly was performed on the clean reads using Trinity (version: 2.06), and transcripts were grouped into unigene clusters using TGICL (version: 2.06). For gene annotation, BLAST (version: 2.2.23) and Diamond (version: 0.8.31) software were used to align the unigenes to the non-redundant nucleotide, non-redundant protein, KOG, KEGG, and SwissProt databases. Software Blast2GO (version: 2.5.0) was used for GO annotation. InterProScan5 (version: 5.11–51.0) was used for InterPro annotation. PossionDis software was used to detect DEGs between the testes and ovaries samples. When comparing the ovaries and testes of juveniles and adults, genes were considered significantly differentially expressed if the false discovery rate p-value (FDR p-value) ≤ 0.001 and the absolute value of a log₂fold-change (ovary expression/testis expression) was ≥ 1.

Validation of DEGs of the transcript data by quantitative real-time PCR analysis

Quantitative real-time PCR was performed to validate the RNA sequencing gene expression data. To analyze gene expression, we selected four unigenes that are involved in sex differentiation and sex development. The primers for doublesex (dsx), FoxL2, Vtg, and ovarian peritrophin 1 (SOP1) were designed (Table 1). Three technical sample replicates were used to quantify gene expression from the ovaries and testes of F. merguiensis in the juvenile and adult stages. RNA was converted to cDNA in a total reaction volume of 20 μL. This included 1 μg of RNA (1 μL) and 100 ng of the random primer (1 μL). Samples were incubated at 70 °C for 5 min and then placed on ice for 5 min. The 5x reverse transcription buffer (4 μL), 4 U/μL AMV (0.5 μL), and 10 mM dNTP (1.25 μL) were added to the reaction; distilled water was added to reach 20 μL, and samples were incubated at 37 °C for 1 hr. The qPCR reaction of 12.5 μL contained 700 ng of cDNA (1 μL), 0.4 μL of 10 μM of the forward and reverse primer, 2X FastStart Universal SYBR Green Master (6.25 μL) [Roche, Mannheim, Germany], and distilled water was added to reach 12.5 μL. PCR was performed using the Mx3000PTM (Stratagene, CA, USA) under the following conditions: initial denaturation at 95 °C for 5 min, 40 cycles of amplification [94 °C for 30 sec, (specific annealing temperature for each primer, Table 1) for 30 sec, 72 °C for 45 sec]. β- actin was used as an internal control. The relative expression level of the selected unigene was calculated based on the 2^−ΔΔCT method [32].

Table 1. Primer sequences for quantitative real-time PCR analysis used to verify RNA sequencing gene expression data.

Primer name	Primer sequences (5′→3′)	Annealing temperature
F-dsx	AACGCTGAGGGAGTTTGTTG	61 °C
R-dsx	CCTGAAGTTGTTGCTGTTGC
F-FoxL2	GCTACAGCTTAGCGAAATCT	54 °C
R-FoxL2	GTCTTCGTGGTTGGGGTCTA
F-FoxL2 Full length	CCGGATCCATGACTTCCCTGGA	60 °C
R-FoxL2 Full length	GGGTCGACCTATATTTTCGAATC
F-FEM1	CCCCATTTGTACTTTCCCAAGC	60 °C
R-FEM1	TGCTAGGCACTGTAAGGAAGTG
F-Tra-2c	TACACAACCGAGAGACAGCTTC	60 °C
R-Tra-2c	ACCCATGTATATGCCAGGAGTG
F-Sxl2	TCATCAACTACCTGCCACAGAC	60 °C
R-Sxl2	ACCTTGATGCGTTTGTGCTG
F-SOP1	TTATGTTGTGTTGGCCCTGG	55 °C
R-SOP1	ACCAAGGTCTTGCAGTTGGC
F-Vtg	ATCTCACCTGGATCAGCCCT	57 °C
R-Vtg	GAGGACTCGGAGATGAAGCG
F-β-actin	CAGATCATGTTTGAGACCTTC	55 °C
R-β-actin	GATGTCCACGTCCACTTCAT

Sequence and phylogenetic analysis of FmFoxL2

FmFoxL2 cDNA sequence (GenBank OQ870552) was derived from the transcriptomic data. The FmFoxL2 protein sequence was translated FmFoxL2 cDNA sequence using bioinformatics tools. Nucleotide sequence similarity was analyzed using the NCBI BLAST tool (http://blast.ncbi.nlm.nih.gov/Blast.cgi). Multiple sequence alignments were performed using Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/), and a phylogenetic tree was constructed using the neighbor-joining method in MEGA 11.0 (http://www.megasoftware.net/), with bootstrap values calculated from 1000 replicates.

Functional domains of the protein were predicted using the ExPASy PROSITE tool (https://prosite.expasy.org/). FoxL2 domains were predicted using SMART (http://smart.embl-heidelberg.de/). The molecular size and theoretical isoelectric point of the protein were determined using the ExPASy Compute pI/Mw tool (https://web.expasy.org/compute_pi/). Signal peptides in the protein sequence were predicted using the SignalP 6.0 software (https://services.healthtech.dtu.dk/service.php?SignalP-6.0). N-glycosylation and phosphorylation sites were indicated using the NetNglyc 1.0 and NetPhos 3.1 servers (https://services.healthtech.dtu.dk), respectively.

Prediction of the FmFoxL2 protein Vtg promoter binding site

A three-dimensional structure of the FoxL2 protein was constructed using the SWISS-MODEL software (https://swissmodel.expasy.org/interactive). Human FOXL2 with a PDB ID of 7vou.1.C was used as a template. The Vtg binding site of the FoxL2 protein was analyzed using the JASPAR online software (http://jaspar.genereg.net/). According to no promotor part of FmVtg and no closely related shrimp Vtg promotor part were available on the NCBI database. Therefore, the mud crab Vtg promotor part was used to predict the banana shrimp FoxL2 Vtg promoter binding site, as reported previously [33].

Knock-down of FoxL2 in vivo by double-stranded RNA

The FoxL2 gene was knocked down to study the role of FoxL2 in the ovarian development of female banana shrimp, based on a report that focused on S. paramamosain [33]. FoxL2 of F. merguiensis was designed from position 600–801, which is in the FoxL2 domain. The FoxL2 fragment was amplified using the forward and reverse primers specified for FoxL2 in Table 1, cloned into pGEM^®-T Easy, and sequenced to verify the FoxL2 fragment. dsRNA-FoxL2 was synthesized using the T7 RiboMAX^™ Expression RNAi System kit (Promega, WI, USA). The integrity of dsRNA was analyzed by 1.2% agarose gel electrophoresis. The concentration of dsRNA was quantified using a NanoDrop 2000.

The dsRNA was diluted to 0.75 μg/μL with phosphate buffer saline (PBS), and 3 μg of dsRNA per gram of shrimp was injected into the shrimp as in previous studies [34, 35]. Female shrimp measuring 15 ± 0.5 cm in length and 26 ± 2.8 g in weight were used in this experiment. One group was injected with dsFoxL2 (three shrimp per group), and the other was injected with PBS (the negative control group). Repeat injections of FoxL2 dsRNA were administered 2, 4, and 6 d after the initial injection. After injection at 2, 4 d, and 6 d, the ovaries of three shrimp from each group were collected as described above. Each ovary was divided into two parts (left lobe and right lobe). The first part of the ovary was inspected the ovarian development by the histological method as previously described in the section “Animals selection for RNA sequencing”. The other part of the ovary was used to analyze the expression levels of Vtg and FoxL2, genes using the same method described in the section “Confirmation of the transcript data by quantitative real-time PCR analysis”.

Statistical analysis

Each experiment was performed in triplicate, and the results were presented as means ± standard deviations (SD). T-test was used to analyze the differences between the samples in each experiment. Differences were considered statistically significant if the p-value < 0.05. Statistical analyses were performed using GraphPad Prism software (version: 9.3.0).

Results

Quality of RNA sequencing and assembly analysis

Four cDNA libraries were constructed from the ovaries and testes of F. merguiensis at the juvenile and adult stages. The BGISEQ-500 sequencing platform generated reads for the juvenile testes and ovaries and the adult testes and ovaries at 77.13, 77.13, 71.87, and 73.62 million reads (MR), respectively (Table 2). After removing the adaptor sequences and filtering out low-quality and unknown base (N) reads, high-quality clean reads of the transcript data were achieved for the juvenile testes (69.25 MR) and ovary (70.39 MR) and the adult testes (66.31 MR) and ovaries (65.39 MR). The number of unigenes in the testes was greater than in the ovaries. The total number of DEGs was higher in the juvenile testes than in the adult testes. In contrast, the DEGs in the ovaries were more abundant in the adult ovaries than in the juvenile ovaries. In addition, the DEGs were approximately 19,846 and 22,815, overlapped between the testes and ovaries of the juvenile and adult shrimp, respectively (Fig 2). The unigenes’ mean length and N50 length were reasonable for both the testes and ovaries samples (Table 2). The RNA-seq data from this study was submitted to the NCBI database with the following project numbers: PRJNA961319 for testes and PRJNA997123 for ovaries, respectively.

Table 2. Summary of the generated transcript data of testes and ovaries from the F. merguiensis juvenile and adult stage.

The values in the brackets¹ and brackets² represent the percentage of clean read and unigenes, respectively.

Samples	Juvenile		Adults
	Testes	Ovaries	Testes	Ovaries
Total number of raw reads (MR)	77.13	77.13	71.87	73.62
Total number of clean reads (MR)	69.25(89.79%)1	70.39(91.26%)1	66.31(92.26%)1	65.39(88.82%)1
Total number of unigenes	38,216	24,503	77,809	43,211
The mean length of unigenes (bp)	1,159	979	725	769
N50 length of unigenes (bp)	2,295	1,786	1,426	1,511
Total number of DEGs (unigenes)	13,838(36.2%)2	1,741 (9.5%)2	4,081(5.2%)2	13,574 (31.4%)2

Fig 2. Circular diagrams showing DEGs number of testis and ovary samples.

The DEGs of the testis and ovary samples from juvenile and adult banana shrimp are present in a circular chart. The overlap area was the repetitive DEGs between the testis and ovary of each stage.

Unigene annotation

The numeric depiction of unigenes in Table 2 and Fig 2 illustrates the number of unigenes obtained through the analysis of the PossionDis software. The testes and ovaries transcriptome data from the juvenile and adult stages were categorized into biological, cellular, and molecular processes using GO (Fig 3A–3C). A large number of genes were expressed in the biological processes and cellular processes of the testes and ovaries in both the juvenile and adult stages. Fewer genes were expressed in the molecular process compared to the other processes. The number of genes expressed in the juvenile stage for each GO term was significantly lower in the ovaries when compared to that of the testes. Conversely, the number of genes expressed for each GO term in the adult ovaries samples was higher than that in the testes (Fig 3D). It is important to clarify that a unigene is categorized when it associates with one or more interconnected GOs.

Fig 3. The DEGs between the testes and ovaries of F. merguiensis were classified using Gene Ontology (GO) terms.

The GO terms are (A) biology process, (B) cellular component, and (C) molecular function. X-axis: The number of DEGs; Y-axis: Ontology sub-categories of genes. (D) Summary of the number of DEGs in each GO term. X-axis: DEGs in GO Terms, and Y-axis: The number of DEGs. The blue bars show the number of DEGs in the juvenile shrimp gonad. The red bars show the number of DEGs in the adult shrimp gonad.

Analysis of DEGs

In the juvenile stage, the number of DEGs in the testes (13,838 unigenes) was greater than that of the ovary (1,741 unigenes). In comparison, the number of DEGs in the adult testes (4,081 unigenes) was less than that of the ovaries (13,574 unigenes) (Table 2). When comparing DEGs in the testes and ovaries, unigenes with a false discovery rate p-value (FDR p-value) ≤ 0.001 and the absolute value of a log₂fold-change (ovaries expression/testes expression) ≥ 1 were considered significantly differentially expressed. This suggests higher levels of gene expression during the development of the testes in the juvenile stage when compared to that of ovarian development. In contrast, more genes were expressed in the ovaries of adult shrimp when compared to those in the testes, likely because female adult shrimp develop a mature ovary. Therefore, this study classified DEGs into two groups: DEGs in the testes of the juvenile and adult shrimp and DEGs in the ovaries of the juvenile and adult shrimp.

DEGs in the testis of the juvenile and adult shrimp

DEGs in the testis were separated into two groups. The first group was the DEGs for juvenile testes development, and five highly DEGs in this group included dsx, Sex-lethal 2 (Sxl2), androgenic gland hormone-like protein (AGH), Protein fem-1 homolog (FEM-1), and neuroglian-like isoform X2 (Nrg-X2) (Table 3). These DEGs were identified as essential for early testis development in the juvenile stage. This was demonstrated by the higher negative ratio observed in the transcript DEGs of the juvenile ovary/testis when compared to that of the adult stage. For example, the log₂fold changes of the dsx and Sxl2 transcripts for ovary/testis were -6.98 and -6.57 and increased to -2.50 and -1.93 in the adult stage, respectively. These genes are mainly involved in sex differentiation. dsx from Fenneropenaeus chinensis was reported to play a crucial role in male sexual differentiation in crustaceans [36]. In contrast, the Sxl gene is involved in sex determination and sexual differentiation in the female development of Drosophila sp. [37]. Another two genes are known as key regulators of male sexual differentiation, the AGH and FEM-1 genes. AGH is the insulin-like androgenic gland hormone (IAG) and plays a crucial role in regulating male sexual differentiation in the Chinese shrimp, F. chinensis [38], and the giant freshwater shrimp, Macrobrachium rosenbergii [39]. The FEM-1 gene has been shown to be specifically expressed in spermatogonia and was suggested to play a role in gamete formation in male gametes of both Caenorhabditis elegans and L. vannamei [40]. Nrg-X2 is an isoform of the neuroglian (Nrg) gene family that involves forming septate junctions in various tissue of insects. During spermatogenesis, septate junctions are crucial permeability barriers at an early stage for appropriate sperm development. In addition, the knock-down of septate junctions disrupted the integrity of the permeability barrier, finally leading to sterility [41, 42].

Table 3. Differentially expressed genes (DEGs) in the testes of juvenile and adult banana shrimp.

Gene annotation	Log2fold-change		FDR p-value
	Ovary/Testis of Juvenile	Ovary/Testis of Adults	Ovary /Testis of Juvenile	Ovary/Testis of Adults
DEGs for juvenile testes development
Crustacean hyperglycemic hormone 1 (CHH1)	-11.24	NA	0.00	NA
SRY-box transcription factor 5 (Sox5)	-8.26	NA	4.35E-73	NA
Doublesex (dsx)	-6.98	-2.50	4.74E-41	4.63E-11
Sex-lethal 2 (Sxl2)	-6.57	-1.93	0.00	9.99E-144
Neuroglian-like isoform X2 (Nrg-X2)	-5.69	-3.82	0.00	0.00
Androgenic gland hormone-like protein (AGH)	-5.21	-1.63	1.82E-01	3.95E-01
Neuroglian-like isoform X1 (Nrg-X1)	-5.04	NA	2.42E-144	NA
Protein fem-1 homolog (FEM-1)	-4.92	-1.21	1.22E-195	1.18E-54
Retinoid X receptor 2	-3.17	0.37a, b	1.13E-05	8.08E-01
Transformer-2a (Tra-2a)	-2.96	-0.70	9.44E-113	1.66E-24
Kinesin-like protein KIFC1 (Kifc1)	-2.18	-0.70a	4.90E-147	1.90E-01
Beta-catenin (β-catenin)	-2.05	-0.76b	5.60E-288	8.98E-14
Nuclear progesterone receptor (PGR)	-1.79	-0.04a	0.00	1.50E-01
Male-specific lethal 3 (Msl3)	-1.61	0.13a, b	5.03E-29	1.61E-01
Cathepsin D (Ctsd)	-1.53	0.88b	4.35E-239	2.07E-217
DEGs for both juvenile and adult testes development
Disrupted meiotic cDNA (Dmc1)	-8.51	-9.91	0.00	0.00
Heat shock protein 70 (Hsp70)	-7.84	-7.44	0.00	0.00
Hematopoietic prostaglandin D synthase (Hpgds)	-7.83	-7.47	1.61E-07	2.75E-10
Forkhead box L2 (FoxL2)	-7.52	-7.14	3.17E-50	0.00
Follistatin-related protein 5 (Fst5)	-6.44	-7.81	9.20E-165	3.00E-82
SRY-box transcription factor 14B (Sox-14B)	-5.73	-7.55	2.72E-60	1.08E-05
Lutropin-choriogonadotropic hormone receptor-like (LHCGR)	-5.13	-7.75	1.39E-09	4.69E-09
Farnesoic acid-O-methyl transferase (FAOMeT)	-4.58	-4.35	0.00	0.00
Juvenile hormone epoxide hydrolase (JHEH)	-4.11	-2.42	0.00	9.98E-242
Ecdysone receptor (EcR)	-4.09	-7.03	2.90E-08	2.84E-38
Piwi-1	-3.82	-3.90	7.28E-92	2.30E-101
Protein fem-1 homolog B (FEM-1b)	-3.74	-2.55	2.59E-09	1.75E-20
Transformer-2 (Tra-2)	-3.74	-3.52	2.15E-253	3.89E-17
Tudor and KH domain-containing protein (TDRKH)	-3.59	-3.43	8.42E-47	1.13E-165
Vasa	-3.09	-1.62	0.00	0.00
Piwi-2	-2.77	-2.02	0.00	0.00
Fushi tarazu-factor 1 (Ftz-F1)	-2.10	-2.10	2.22E-29	9.81E-43

NA indicates that no transcriptome data were found.

^a gene expression is not different between the testis and ovary because FDR > 0.001.

^b gene expression is not different between the testis and ovary because the absolute value of log₂fold-change is less than 1.

The second group of genes showed high levels of differential expression in both the juvenile and adult testes (Table 3). These DEGs were required for both juvenile and adult testes development, for example, disrupted meiotic cDNA (Dmc1), hematopoietic prostaglandin D synthase (Hpgds), FoxL2, and follistatin-related protein 5 (Fst5). Dmc1, a DNA recombinase and RecA/Rad51 superfamily member plays a crucial role in meiotic recombination [43]. In the testis samples of Penaeus monodon [44] and L. vannamei [45], Dmc1 expression was significantly upregulated. Hpgds, or PGDS, is an enzyme that catalyzes producing PGD2 in the peripheral tissues [46]. PGDS is expressed in tissues such as the testis and ovaries [47, 48]. PGDS was highly expressed in the hepatopancreas, and the testis regulates spermatogenesis in the Chinese mitten crab Eriocheir sinensis [49]. FoxL2 synthesized estrogen and aromatase in mammals and fish [22–26] and was involved in Vtg expression by binding to the Vtg promoter [33]. Therefore, low expression levels of FoxL2 are directly related to the increase in Vtg expression observed in E. sinensis [30] and S. paramamosain [33]. Follistatin (Fst) is a vital regulatory protein of the transforming growth factor-beta (TGF-β) superfamily in vertebrates [50].

DEGs in the ovary between juvenile and adult shrimp ovary

A total of nine genes were found to be highly differentially expressed in the juvenile and adult ovaries. These genes included ovarian peritrophin (SOP), SOP1, ovarian peritrophin 2 (SOP2), prostaglandin reductase 1 (Ptgr1), vitelline membrane outer layer protein 1 (Vmo1), estrogen sulfotransferase (EST), vitellogenin receptor (Vgr), thyroid hormone receptor beta-A (TR-beta-A), and transformer isoforms C (Tra-2c) (Table 4). Most of the genes in this group are involved in structural formation; SOP is a key component of the jelly layer and cortical rods in shrimp and plays a crucial role in protecting eggs during spawning [51]. Ptgr1 is involved in polyunsaturated lipid proliferation during vitellogenesis in freshwater crayfish [52, 53]. Vmo1 is a protein that separates the yolk from the egg, is located in the outer layer of the egg white vitelline membrane, and protects the embryo against bacterial infection [54]. In addition, VmoI was reported to be produced in the hepatopancreas and transported into oocytes during the vitellogenesis of L vannamei [55]. Vgr is a plasma membrane-bound protein that specifically binds and mediates the transport of the protein Vtg into oocytes [56]. Vtg, Transformer-2 (Tra-2) was the only regulator gene in this group that was found to be differentially expressed. It has been reported to play a crucial role in sex differentiation and development in Drosophila melanogaster by directing the sex-specific alternative splicing of the DSX pre-mRNA in conjunction with the transformer (Tra) protein. Three isoforms, Tra-2a, Tra-2b, and Tra-2c, have been reported. Tra-2c has been shown to play a role in female determination in Chinese shrimp [3]. Other DEGs that were required during the adult stage included progestin membrane receptor component 1 (Pgmrc1), juvenile hormone esterase-like protein 1 (JHE1), and Chorion peroxidase-like (Pxt).

Table 4. Differentially expressed genes in the ovaries of juvenile and adult shrimp.

Gene annotation	Log2fold-change		FDR p-value
	Ovary /Testis of Juvenile	Ovary/Testis of Adults	Ovary /Testis of Juvenile	Ovary/Testis of Adults
DEGs for both juvenile and adult ovary development
Ovarian peritrophin 1 (SOP1)	20.53	15.39	0.00	0.00
Ovarian peritrophin 2 (SOP2)	19.64	15.09	0.00	0.00
Prostaglandin reductase 1 (Ptgr1)	18.05	12.02	0.00	0.00
Ovarian peritrophin (SOP)	14.40	8.50	0.00	0.00
Vitelline membrane outer layer protein 1 (Vmo1)	8.39	5.53	3.10E-107	8.05E-134
Transformer-2c (Tra-2c)	2.20	1.03	3.03E-09	4.85E-10
Estrogen sulfotransferase (EST)	10.35	12.56	3.24E-75	0.00
Vitellogenin receptor (Vgr)	8.62	13.67	0.00	0.00
Thyroid hormone receptor beta-A (TR-beta-A)	7.60	10.74	2.33E-17	2.93E-270
Vitellogenin (Vtg)	2.58a	10.24	1.06E-01	0.00
Juvenile hormone esterase-like protein 1 (JHE1)	-3.91a	6.63	9.63E-02	1.26E-12
Chorion peroxidase-like (Pxt)	-7.25	5.32	2.75E-09	1.03E-04
Wingless-type MMTV integration site family, member 4 (Wnt4)	NA	4.86	NA	9.41E-08
Octopamine receptor beta-2R (Octbeta2R)	NA	3.32	NA	1.24E-32
Argonaute 1 (AGO1)	-1.36a	2.87	9.82E-02	9.10E-07
SRY-box transcription factor 9 (Sox-9)	NA	2.31	NA	1.36E-11
Neuroparsin (NP)	-2.21a	1.87	1.10E-03	7.55E-05
Profilin	-2.72	1.71	8.39E-19	0.00
Prostaglandin E synthase 2 (Pges2)	0.51b	1.41	4.96E-09	1.20E-88
Protein fem-1 homolog A (FEM-1a)	-0.84b	1.37	8.97E-06	3.20E-06

NA indicates that no transcriptome data were found.

^a gene expression is not different between the testis and ovary because FDR > 0.001.

^b gene expression is not different between the testis and ovary because the absolute value of log₂fold-change is less than 1.

Validation of transcriptome analysis using qPCR

We used qPCR to validate the transcriptome data. The expressions of dsx and FoxL2 were selected as representative DEGs for testis development. Additionally, SOP1 and Vtg were used to validate the transcriptome data in relation to ovarian development. The genes selected for qPCR were based on early and late gonad development requirements.

The expression of dsx in the testis and ovary of banana shrimp was evaluated in Fig 4A and 4B. The dsx expression level was higher in the banana shrimp testis than that of the ovary and showed lower expression levels in adult testis compared to that of juveniles. These findings suggest that dsx plays a more important role in male sexual development. qPCR analysis revealed that FoxL2 gene expression was relatively elevated in the testis when compared to that of the ovary and exhibited higher expression levels in adult testis when compared to that of juveniles (Fig 4C and 4D).

Fig 4. Validation of transcriptome data using qPCR.

Fold-change in gene expression using transcriptome data and qPCR. Fold changes represent gene expression in the ovary and testis of shrimp in the juvenile and adult stages (n = 3). (A) and (B) show dsx expressions. (C) and (D) indicate FoxL2 expression. (E) and (F) show Vtg expression levels. (G) and (H) show SOP1 expression. Different letters (a, b) indicate a significant expression between sample groups at a p-value < 0.05.

Vtg plays a role in ovary development in female banana shrimp. Vtg was not expressed in the testis samples (both juvenile and adult) but at high levels in adult ovary samples (Fig 4E and 4F). The gene SOP1 was also found to be more important in adult ovary development, as higher expression levels were observed in adult ovary samples when compared to those of testis in both qPCR and transcriptome data (Fig 4G and 4H). The expression levels of dsx, FoxL2, Vtg, and SOP1 obtained from qPCR showed the same pattern as the transcriptome data.

Sequence and phylogenetic analysis of FoxL2

Since FoxL2 was more highly expressed in the testis than the ovary in shrimp, and FoxL2 is involved in cell proliferation and differentiation during human ovarian development [12, 13]. Therefore, the full-length cDNA of the banana shrimp FoxL2 was analyzed from transcriptome data. It has 1,497 base pairs, encodes 492 amino acids with 53.2 kDa, and has an isoelectric point (pI) of 6.81. The FoxL2 sequence contains two glycosylation sites and 87 phosphorylation sites (Fig 5). The forkhead domain is spaned from 180 to 279 (Fig 5). FoxL2 belongs to the winged helix/forkhead transcription factor family, which consists of a conserved DNA-binding domain known as the forkhead box [57]. The FoxL2 identities of eight different decapod species (Homarus americanus, M. rosenbergii, E. sinensis, S. paramamosain, Portunus trituberculatus, P. monodon, Procambarus clarkii, Cherax quadricarinatus, and F. merguiensis) were compared. The FoxL2 domain of F. merguiensis is the same size (96 amino acids) as P.monodon, and the FoxL2 domain area is in the same position (Fig 6A). S. paramamosain has the longest FoxL2 domain size (97 amino acids). In addition, E. sinensis has the smallest domain size (~84 amino acids), while other species share a common domain size, including F. merguiensis (~96 amino acids).

Fig 5. Nucleotide and amino acid sequences of F. merguiensis FoxL2.

The FoxL2 domain is depicted in yellow. The red line underlines two glycosylation sites, and phosphorylation sites are circled in black.

Fig 6. Diagram of the FoxL2 domain and phylogenetic analysis of FoxL2 proteins in banana shrimp and other species.

(A) The chart of the FoxL2 domain according to species. Ha: H. americanus; Mr: M. rosenbergii; Es: E. sinensis; Sp: S. paramamosain; Pt: P. trituberculatus; Pm: P. monodon; Fm: F. merguiensis; Pc: P. clarkii; Cq: C. quadricarinatus. (B) The neighbor-joining phylogenetic tree of FoxL2 amino acid sequences from different organisms was constructed by MEGA 11.0, and 1,000 bootstrap replicates were used to assess the confidence in each node.

The highest nucleic acid sequence identity of FoxL2 was 95.93%, which was shared with P. monodon, and the lowest was 62.73% with M. rosenbergii (S1 Fig). Phylogenetic analysis revealed that the FoxL2 proteins of F. merguiensis were most closely related to those of P. monodon (Fig 6B). The amino acid sequence of the forkhead domain between F. merguiensis and P. monodon is 100% identical. Comparisons between S. paramamosain and F. merguiensis showed 89.58% similarity in the forkhead domain, where the conserved domain was identical.

Prediction of the FoxL2 protein Vtg promoter binding site

Before experimenting with an animal model, the FoxL2 protein was screened for binding to the Vtg promoter using in silico binding. A three-dimensional protein structure model of the FoxL2 forkhead domain was constructed and contained three alpha helices and two beta sheets (Fig 7A). The appearance of such a three-dimensional structure follows the winged helix, which is characteristic of the forkhead box domain. The model had a QMEANDisCo global score of 0.63 ± 0.09 and a sequence identity of 76.09% to human FoxL2 (PDB ID: 7vou.1.C). The JASPAR program was used to predict binding sites between the FoxL2 domain of F. merguiensis and the Vtg promotor of S. paramamosain. A total of five positions were predicted. The top two binding positions with the highest relative scores (rs) were 5′-AGAAAATAAACAAA-3′ (rs: 0.89) and 5′-ATTTTGTAATCACG-3′ (rs: 0.84) (Fig 7B). The second binding position is consistent with the prediction of FoxL2 binding to the Vtg promoter of S. paramamosain [33].

Fig 7. A model of the three-dimensional structure of F. merguiensis FoxL2 binding to the Vtg promoter.

(A) The three-dimensional structure of the FoxL2 domain of F. merguiensis. (B) Two nucleotide-binding regions bind to the FoxL2 proteins, predicted using the JASPAR database. (C) The first binding site and (D) the second binding site in the FoxL2 domain (pink) connect with the promoter segment of the Vtg gene in S. paramamosain (green).

The model illustrates the interaction site between FoxL2 and the Vtg promoter, with the two positions depicted in Fig 7C and 7D, respectively. Predictions show that FoxL2 at SER184, TYR185, LYS223, SER231, HIS234, and ASN235 binds to two DNA binding sites of the Vtg promoter through hydrogen bonds or other types of electrostatic interactions. The first binding site has the shortest distance between the nitrogen atom of HIS234 and the nitrogen atom at the second position of the phosphate group of adenine-729 of the Vtg promotor, which is 0.9 angstroms. Meanwhile, FoxL2 at ARG233, HIS234, SER237, LYS244, and TRP259 binds to the first antisense DNA binding site with the shortest distance is 2.0 angstroms. The first antisense site is between the nitrogen atom of HIS234 and the nitrogen atom located at the third position in the nucleotide sugar ring of thymine, which is complementary to adenine-731 of the Vtg promotor. In addition, the identical amino acid residues of FoxL2 bound with the second Vtg promoter binding site. The minimum distance recorded at this binding site is 0.9 angstroms, particularly between HIS234’s nitrogen atom and the primary position nitrogen atom of adenine-315 within the Vtg promoter.

Furthermore, FoxL2 at TYR207, ASN230, ARG233, HIS234, SER237, LYS244, and TRP259 are associated with the antisense DNA strand using hydrogen bonds or other electrostatic interactions. This specific location is the shortest at 2.0 angstroms, directly linking the nitrogen atom of HIS234 with the nitrogen atom at three specific positions within the nucleotide sugar ring of thymine. This particular interaction corresponds with adenine-314 of the Vtg promoter.

Effects of FoxL2 knock-down in female banana shrimp

As a positive result of in silico Vtg promoter binding, FoxL2 knocked-down in female banana shrimp was performed. After dsRNA-FoxL2 was injected at 2, 4, and 6 d, the expression level of FoxL2 genes was significantly down-regulated than those of the control group. While the expression levels of Vtg were found to be significantly higher (p-value < 0.05) than those of the control group at 2, 4, and 6 d post-injection with dsRNA-FoxL2 (Fig 8). In addition, the morphology and histology of the ovaries after down-regulated FoxL2 expression were overall developed, corresponding to the increase of the Vtg expression, shown in Fig 9. These results indicated that FoxL2 expression was successfully knocked down by dsRNA-FoxL2. Meanwhile, Vtg expression was activated when FoxL2 was reduced.

Fig 8. Expression levels of FoxL2 and other ovarian genes after FoxL2 knock-down.

Expression levels of FoxL2 and Vtg genes at 2, 4, and 6 d post-injection (dsRNA-FoxL2) were analyzed using three replicates. The symbol * indicates a significant difference between sample groups, p-value < 0.05.

Fig 9. Morphology and Histological investigation of ovarian samples after FoxL2 knock-down.

The ovarian samples after FoxL2 knock-down experiment were sectioned (5μmthick) to determine ovarian cell development after downregulation of FoxL2 expression. Red arrows indicate areas where ovary discoloration to yellow was observed. Og: Oogonia; Oc: oocytes; Fc: follicular cells. The tissue image was taken at 20X magnification under a BX-53 light microscope equipped with a DP-72 digital camera (Olympus, Tokyo, Japan).

Discussion

F. merguiensis is popular for consumer and commercial species in several countries, including Southeast Asian countries, India, and Australia. Aquaculture of this species may economically benefit these countries in terms of food security and income. Understanding the reproductive system of this shrimp is essential for aquaculture applications. Although the RNA-seq approach has been reported for F. merguiensis [4, 6, 10, 11] and other shrimps, more information focusing on sexual and reproductive development is required to increase seed quality and obtain sustainable cultures. This study used transcriptome analysis to investigate the dynamic change in gene expression between juvenile and adult ovaries and testes. This study revealed that the detected unigenes in the juvenile testes and ovaries were about 50% of those in the adult stage. Interestingly, the total number of DEGs (unigenes) in the testis was 36.2% and 5.2% in juvenile and adult testis, respectively. In contrast, the total number of DEGs in the ovary was 9.5% and 31.4% in the juvenile and adult stages, respectively (Table 2). This indicates that the development in the adult ovary required more upregulated genes than in the adult testis. A comparison between the ovary and testis transcriptome data was also reported in P. serratus and demonstrated that genes were more upregulated in the adult ovary when compared to the adult testis [4], which is similar to the data shown in this study.

In this study, histological techniques after H&E staining of ovarian tissue were classified into 4 stages as previous reports: undeveloped stage (oocyte<65 in diameter), early development (stage I, oocyte 75–125 in diameter)), developing (stage II, oocyte 100–200 μm in diameter, fat droplet, cortical rods) and mature (stage III, oocyte 125–250 μm in diameter, fat droplet, cortical rods) [58–61]. In this study, the H&E staining juvenile’s ovarian sample contained oogonia, primary oocyte (<65 μm in diameter), and primary follicle cell was classified as undeveloped ovary stage. The adult’s ovarian sample comprised of oogonia, oocyte (75–125 μm in diameter), and primary follicle cell, defined as early development stage I (Fig 1). Therefore, the expressed genes in the adult’s ovarian sample were more than in the juvenile’s ovarian sample. In the juvenile ovary, meiosis occurs to accumulate arrested oocytes in previtellogenic growth. A secondary oocyte maturation (vitellogenin) takes place at the adult stage [62]. Then the secondary oocyte maturation (vitellogenin) resumes meiosis by the time spawning. It corresponds with the DEGs that were shared both in the juvenile and adult stages. The SOP1, SOP2, Ptgr1, SOP, Vmo1, Tra-2c, EST, Vgr, TR-beta A genes. Meanwhile, the DEGs of Vtg, JHE1, and Pxt genes were significantly found in the adult ovary.

In this study, the SOP of F. merguiensis was the high expressed during the juvenile (early previtellogenic stage, which corresponds with the expression of the SOP of Masupenaeus japonicus. The SOP of M. japonicus in the ovary was higher in the previtellogenic stage than the endogenic stage, early exo-vitellogenic stage, late exo-vitellogenic stage and mature stage [63]. The same pattern of Vtg mRNA in the ovary was reported to slightly increase in the previtellogenic stage and higher expression in the endogenic stage than in the early, late, and mature stages. In contrast, the Vtg expressed in the hepatopancreas at a low level during the previtellogenic to the endogenic stage, increased the level in the exo-vitellogenic stage and the highest level in the mature stage. SOP and Vtg were also produced from extra-ovary (hepatopancreas) in M. japonicus, P. vannamei, P. semisulcatus, P. monodon [63–66]. Vtg of F. merquiensis is initially upregulated expression in the previtellogenic in this study and in the previous study [67] and is the highest expression in the vitellogenin stage [60].

Vmo1 is one of the essential proteins in the ovary and mRNA was detected in the previtellogenic ovary in this study. In comparison, Vmo1 was not found in the ovary of L. vannamei and was reported only produced from the hepatopancreas during vitellogenesis and imported into the ovary [55]. The different stages and species of shrimp provided different sources of protein in the ovary. For example, Vtg of M. rosenbergii was produced only from hepatopancreas [68], while Vtg of M. japonicus, P. vannamei, P. semisulcatus, P. monodon was from ovary and hepatopancreas [63–66].

Testis, where spermatogenesis occurs and comprises spermatocytogenesis and spermiogenesis stages. The spermatocytogenesis produces haploid spermatids from diploid spermatogonia. Spermiogenesis is the process that transforms the mature spermatids into spermatozoa [69]. Due to the juvenile testis sample containing most spermatogonia and primary spermatocyte to transform into spermatid, many expressed genes were found. While the adult testis sample consisted of many spermatids, the primary process is spermiogenesis, and fewer genes were expressed in the adult testis than in the juvenile testis (Figs 1 and 2). Additionally, this study was supported by the report that DNA synthesis in testes at the pre-molt stage is higher than at the post-molt and inter-molt stages in M. rosenbergii [70, 71]. However, the spermatogenesis in L. vannamei is continuous without related to the molt cycle [72].

Since FoxL2 is known to be involved in female development in humans and mice, the suppression of Sox-9 expression in humans and mice suggests that the FoxL2 gene plays a role in regulating cell differentiation and proliferation [20]. FoxL2 is involved in the synthesis of estrogen and aromatase in mammals and fish [22–26]. In F. merguiensis, transcriptome analysis and qPCR showed that the FoxL2 gene was expressed in the juvenile testis and was upregulated as it progressed into the adult stage (Fig 2C and 2D). The reason for the higher expression of FoxL2 in the testis than in the ovary of the shrimp remains unclear. FoxL2 gene expression in M. rosenbergii was reported to be highest in males (zz), followed by females (zw) and super females (ww) [73]. Moreover, the expression pattern of FoxL2 was analyzed using fluorescence in situ hybridization in S. paramamosain, which revealed signals in both the ovary and testis. In the ovary, signals were identified in follicle cells. In contrast, in the testis, signals were detected in the epithelium of the seminiferous tubules, spermatogonia, spermatocytes, and spermatids. This suggests that FoxL2 may be crucial in testis differentiation and development [74].

In crustacean species, E. sinensis [30] and S. paramamosain [33] showed high expression levels of Vtg in the ovary, while low expression of FoxL2 was found. In addition, a FoxL2 mechanism was proposed to bind to Ftz-F1, inhibiting Ftz-F1 from binding to the cytochrome P450 promoter and suppressing Vtg expression [30]. Moreover, it has been reported that FoxL2 inhibits Vtg expression by binding the Vtg promoter in S. paramamosain [33]. In the transcriptome analysis of this study, the Ftz-F1 and FoxL2 transcripts were higher in both the juvenile and adult testes when compared to those of the ovary (Table 3), which may support the FoxL2 mechanism affects Vtg expression in F. merguiensis. In addition, knock-down by the dsRNA injection significantly suppressed FoxL2 when compared to the control 2, 4, and 6 d post-injection (Fig 8). Consequently, Vtg expression was activated when compared between 2, 4, and 6 d post-injection. These phenomena indicate that FoxL2 reduction is essential for Vtg expression. Further investigation is required to confirm mechanism of these findings.

Supporting information

S1 Fig. Multiple alignments of the deduced amino acid sequence of the FoxL2 gene.

Multiple alignments of the deduced amino acid sequence of the FoxL2 gene, GenBank accession no. OQ870552, with H. americanus: accession no. KAG7154875.1, M. rosenbergii: accession no. USJ75257.1, E. sinensis: accession no. AIS92518.1, S. paramamosain: accession no. QQY98966.1, P. trituberculatus: accession no. WAA68168.1, P. monodon: accession no. XP_037795163.1, P. clarkii: accession no. ALD48735.1, and C. quadricarinatus: accession no. UWX37250.1. The red boxes represent the FoxL2 conserved domain.

(TIF)

Data Availability

All relevant data are within the paper. The transcriptome data are held in NCBI database, project numbers: PRJNA961319 for testes and PRJNA997123 for ovaries.

Funding Statement

I confirm that the funding organization did not play a role in the study design, data collection, and analysis, decision to publish, or preparation of the manuscript. The funder provided support in the form of salaries for W. Potiyanadech and C. Choomee but did not have any additional role in the study design, data collection, analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section. 1) A Ph.D. candidate scholarship from The Royal Golden Jubilee Graduate Program of the Thailand Research Fund (TRF) (4.J.PS/58/B.1). 2) This research has received funding support from Fundamental Fund, Prince of Songkla University, contact no. SCI6505113b. 3) This research has received funding support from the NSRF via the Program Management Unit for Human Resources & Institutional Development, Research, and Innovation [grant number B05F630026].