Transcriptomic profiling reveals molecular regulation of seasonal reproduction in Tibetan highland fish, Gymnocypris przewalskii
The Tibetan highland fish, Gymnocypris przewalskii, migrates from Lake Qinghai to its spawning grounds every summer. This seasonal reproduction is critically regulated by intrinsic and extrinsic signals. However, the molecular mechanisms that process environmental oscillations to initiate the seasonal mating are largely unknown.
A transcriptomic analysis was conducted on the brain and gonad of male and female G. przewalskii in reproductive and nonreproductive seasons. We obtained 2034, 760, 1158 and 17,856 differentially expressed genes between the reproductively active and dormant female brain, male brain, ovary and testis. Among these genes, DIO2 was upregulated in the reproductively active brain and gonad of both males and females. Neuroactive ligand-receptor genes were activated in male and female brain. Functional enrichment analysis suggested that retinol metabolism was uniquely stimulated in reproductively active males. Genes involved in GnRH signaling and sex hormone synthesis exhibited higher expression levels in brain and gonad during the reproductive season. A co-expression network classified all the genes into 9 modules. The network pinpointed CDC42 as the hub gene that connected the pathways in responsible for modulating reproduction in G. przewalskii. Meanwhile, the sex pheromone receptor gene prostaglandin receptor was identified to link to multiple endocrine receptors, such as GnRHR2 in the network.
The current study profiled transcriptomic variations between reproductively active and dormant fish, highlighting the potential regulatory mechanisms of seasonal reproduction in G. przewalskii. Our data suggested that the seasonal regulation of reproduction in G. przewalskii was controlled by the external stimulation of photoperiodic variations. The activated transcription of neuroendocrine and sex hormone synthesis genes contributed to seasonal reproduction regulation in G. przewalskii, which was presumably influenced by the increased day-length during the breeding season.
KeywordsSeasonality Reproductive migration Neuroendocrine Day-length RNA-seq WGCNA Tibetan highland fish
Differentially expressed gene
False discovery ratio
Galanin receptor 2
- HPG axis
Acetylcholine receptor 9
Neuropeptide Y receptor
Relaxin receptor 2
Retinol X receptor α
Retinol X receptor γ
Weighted gene coexpression network analysis
Seasonal reproduction is an evolutionary adaptive strategy for animals living in areas with evident seasonality [1, 2, 3]. To maximize the survival of offspring, animals in the mid and high latitudes reproduce at the most beneficial time of the year, which has moderate climate and sufficient food . In vertebrates, the timing of mating behavior is tightly controlled by the coordination of both environmental cues and internal factors. Seasonal changes in temperature, food availability, and day-length are environmental factors required to initiate reproduction in many animals [4, 5]. Based on the relative day-length during the reproductive season, animals are classified into long-day (LD) and short-day (SD) breeders . The light information received by the retina of mammals and extra retina tissues of lower vertebrates influences neuroendocrine production, transmitting environmental information to affect the physiological activities of animals, including reproduction . For LD breeders, increased day length induces gonadotropin-releasing hormone (GnRH) production, which facilitates follicle-stimulating hormone (FSH) and luteinizing hormone (LH) synthesis as well as gonad development [7, 8]. The mechanisms controlling seasonal reproduction have been firmly established in mammals, which involve the stimulation of gonad development via the hypothalamic-pituitary-gonadal (HPG) axis depending on the seasonal fluctuation of melatonin production due to the day-length change .
Seasonal reproduction is widely observed in temperate fish species as well . For example, medaka is characterized as a LD breeder that develops gonads from spring to summer, while salmonid fishes are SD breeders who migrate back to home rivers for spawning in autumn [11, 12]. Although seasonal reproduction is pervasive in fish, its molecular basis remains ambiguous. Saccus vasculosus (SV), a specific organ in fish brain that serves as a sensor responding to photoperiodicity, expresses all the components in the regulation of the photoperiodic signaling pathway [13, 14]. Meanwhile, a variety of neurotransmitters, such as melatonin, dopamine, acetylcholine, GABA, and Kiss1, as well as their receptors, have been reported to translate environmental stimuli into internal signals and synchronize reproduction in different fish species [10, 15, 16, 17]. Therefore, seasonal reproduction is likely to be controlled by multiple neuroendocrine molecules in fishes in response to photoperiodicity, with regulatory mechanisms distinct from those of mammals.
In the current study, we attempted to examine the role of the transcriptional regulation in controlling seasonal reproduction in G. przewalskii. To achieve this goal, brain and gonad samples from both male and female G. przewalskii were collected in the RS and NRS. Gene expression analysis revealed that the number of differentially expressed genes (DEGs) was 2034 in female brain, 760 in male brain, 1158 in ovary and 17,856 in testis genes between fishes in RS and NRS. DEGs were enriched in pathways including steroid hormone synthesis, neuroactive ligand-receptor interaction, and retinol metabolism. Gene co-expression network analysis revealed CDC42 as a hub bridging genes in reproduction-related pathways. Taken together, our results showed that the timing of breeding in G. przewalskii was highly likely to be controlled by the increased day length in the summer, which probably induced the transcription of genes in neuroendocrine regulation of sex hormone synthesis.
De novo assembly and annotation of G. przewalskii transcriptome
Summary of de novo assembly using Trinity
Number of raw reads
Number of clean reads
Number of unigenes
N50 of unigenes (bp)
Average length of unigenes (bp)
Number of unigenes
Annotated in Nr
Annotated in KOG
Annotated in Swiss-Prot
Annotated in KEGG
Annotated in all databases
Annotated in at least one database
Quantification of transcriptional abundance between RS and NRS samples
Number of DEGs between the NRS and RS in female and male fish
Functional enrichment analyses of DEGs
Identification of a seasonal-reproduction-related coexpression network
Reproduction in most fishes is a seasonal phenomenon that occurs at a precise time of the year to ensure the survival of the offspring. The timing of the spawning migration is tightly controlled by the coordination of internal and external signals. Studies of the genetic and physiological bases of reproduction have been carried out in several fish species, and the results suggested both universality and diversity in the modulation of seasonal breeding [11, 22, 23, 24, 25, 26]. Among environmental cues, photoperiod is considered the primary and most reliable determinant of the timing of reproduction via neuroendocrine regulation [27, 28]. G. przewalskii inhabits in a region with evident seasonality. Long-term observations have demonstrated that the water temperature in this region shows significant fluctuation in the spring and early summer due to the sunshine, wind and precipitation. However, the day length peaks from May to Aug (Fig. 1c) [20, 21]. Therefore, we proposed that increased day length played the critical role in regulating seasonal reproduction in Tibetan highland fish. In the current study, we compared reproductively active and dormant G. przewalskii at the transcriptomic level, aiming to uncover the regulatory mechanisms that synchronize environmental cues and reproduction. The high quality of the reference transcriptome laid a foundation for the study of the molecular basis of seasonal breeding in G. przewalskii.
Potential interplay between photoperiod and reproduction in G. przewalskii
Several neuroendocrine molecules were identified in RS brain samples, which was the possible outcome of the increased day-length in RS. DIO2, the key enzyme in the conversion of thyroid hormone to its bioactive form [8, 29], is induced by LD exposure in birds and mammals during the RS [30, 31]. The enhanced transcription of DIO2 found in reproductively active fishes may result from the increased day-length, based on the similar regulatory mechanisms in other vertebrates. This result highlights its potential role in facilitating seasonal reproduction via photoperiodicity in G. przewalskii. Meanwhile, several neuroendocrine receptor genes that have been reported to control sexual maturation and reproduction in other species were stimulated in G. przewalskii during the RS. For example, massive studies have confirmed that Kiss1 and its receptor mediate the photoperiodic control and sexual maturation in male fishes and mice via promoting GnRH production [32, 33]. In reproductively active male G. przewalskii, the activation of Kiss1r during the RS may bridge the gap from LD sensing to gene expression in GnRH signaling. It has been reported that neuroendocrine receptor genes, NPYR, nAChR9, RXFP2, GALR2 and OXTR exhibit robust circadian expression patterns and regulate HPG axis in many species [34, 35, 36, 37, 38]. The activation of these genes in female individuals suggests their gender-specific roles in regulation of genes in GnRH in G. przewalskii through the increased day-length in RS [34, 35, 36, 37, 38].
The function of GnRH in reproduction has been well established in a variety of species. GnRH promotes the production of LH and FSH, which in turn result in gonad development and breeding behaviors [9, 39, 40]. The simultaneous upregulation of GnRH2 and LHβ in brain, as well as that of testosterone 17-β-dehydrogenase 3 and estradiol 17-β-dehydrogenase 1 in gonad, highlights the importance of the GnRH signaling pathway in the coordination of sexual hormone synthesis in G. przewalskii during the RS. Overall, the transcriptomic analysis indicated the involvement of photoperiodic and neuroendocrine regulation in the seasonal reproduction of G. przewalskii.
Melatonin is known to control the timing of reproduction by influencing GnRH production in many species, including fish . However, we did not observe DEGs in melatonin synthesis in the current study, indicating that melatonin probably was not engaged in directing seasonal reproduction in G. przewalskii. In addition to melatonin, dopamine, GABA and NPY were thought to regulate gonadal development in fish , therefore, we considered that other neuroendocrine molecules, such as NPY, acetylcholine, relaxin, galanin and oxytocin, contributed to seasonal breeding in G. przewalskii.
Dual functions of retinol metabolism in male G. przewalskii reproductive regulation
The functional enrichment analysis indicated that retinol metabolism probably exerted an important function in controlling reproduction in male G. przewalskii, due to the activation of the pathway in both brain and testis during RS. Retinol metabolism involves a series of reactions that convert dietary vitamin A to all-trans retinoic acid (RA), the cellular active metabolite . First, vitamin A is processed to all-trans retinol, which is stored predominantly in the liver. In response to the body’s requirements, retinol is secreted from the liver and delivered to target tissues, where RA is generated . RA binds to its receptors, retinol X receptor α (RXRA) and retinol X receptor γ (RXRG), which exert diverse functions depending on the tissue and cell types [39, 43, 44, 45]. Accumulated evidence suggests that RA and its receptors control spermatogenesis. RALDH is the rate-limiting enzyme for RA biosynthesis . Mice deficient in genes encoding RALDH and RA receptors showed phenotypes of low RA levels in testis and impaired spermatogenesis, indicating that RA is crucial for male reproduction [47, 48]. In the current study, we observed that genes responsible for RA synthesis were stimulated in RS testis, probably causing elevated RA concentration. Given this information and the increased transcription of RA receptor genes, it was reasonable to speculate that the upregulation of genes in retinol metabolism and RA signaling may contribute to the seasonal formation of sperm in male G. przewalskii.
On the other hand, retinol metabolism regulated the seasonal reproduction through the deep brain photoreceptors in lower vertebrates, by influencing neurogenesis, locomotion, and synaptic plasticity . It is well accepted that nonmammalian vertebrates detect light via deep brain photoreceptors lying outside the retina . Vertebrate ancient (VA) opsin and 11-cis-retinal form the photopigments that respond to photoperiodism in the hypothalamus of birds . In fish species, all components of the photoperiodic transduction machinery are integrated into a brain region called SV, where opsin family genes are expressed . The activation of genes in retinal synthesis suggests that they probably participate into the photoperiodic response by interacting with opsin in male G. przewalskii brain.
Molecular network controlling seasonal reproduction in G. przewalskii
Co-expression network analysis uncovered interactions among the genes controlling seasonal reproduction in G. przewalskii. For example, a connection between BMP7 and FSHR in G. przewalskii was revealed, in accordance with a previous finding showing a direct increase in FSHR expression by BMP7 activation in human cell lines . This result improved our confidence in the gene interactions in the network and indicated the universality in reproductive regulation between fish and mammals. In addition, the hub gene CDC42 has been reported to modulate reproduction by stimulation of LH synthesis, oocyte meiotic maturation and fertilization in mammals [53, 54]. We also noticed the connections of CDC42 with genes in GnRH signaling, retinol metabolism and oocyte maturation in the network, emphasizing its importance in coordinating the photoperiodicity, the HPG axis and gonad development in G. przewalskii. Moreover, it was worth noting the involvement of prostaglandin receptor genes PTGER4 and PTGFR in the network, which interacted with PGTR, GnRHR2 and FSHR. Prostaglandins are the major water-borne, hormonally derived sex pheromones in fishes. They induce reproductive behavior by interacting with receptors [23, 55]. Our results indicated they were co-regulated in the reproductively active male gonad, highlighting the potential feedback response in controlling the spawning migration in male population. Further works are required to untangle the role of pheromone in seasonal breeding of G. przewalskii.
The current study was the first attempt to investigate the molecular regulation of seasonal reproduction in Tibetan highland fish. The activation of DIO2 by the LD of the summer may influence sex hormone synthesis in G. przewalskii. The upregulation of neuroactive ligand-receptor genes was involved in the response to day-length change, and these receptors probably switched on genes in GnRH signaling and steroid hormone synthesis. Genes in retinol metabolism were specifically stimulated in reproductively active brain and testis to manage phototransduction and spermatogenesis in male G. przewalskii. The network analysis identified CDC42 as a hub due to its interactions with genes in multiple pathways. Meanwhile, the interaction among prostaglandin receptor genes with GnRHR2 and FSHR suggested the potential feedback regulation in courtship behaviors in male G. przewalskii. Moreover, the well-designed condition controlled experiments are required to untangle effects of environmental factors, such as water temperature and salinity on gene expression variations during the reproductive migration in G. przewalskii. In conclusion, our results profiled the transcriptomic alterations between reproductively dormant and active fish, suggesting that the LD condition in the summer was the principal signal to induce genes in neuroactive ligand-receptors and retinol metabolism in female and male brain. The synchronized upregulation of genes in GnRH signaling and sex hormone synthesis may promote gonad development and spawning migration in G. przewalskii.
Sample collection and RNA isolation
Twelve reproductively active male (body weight = 102.6 ± 30.46 g) and female (body weight = 244.5 ± 51.64 g) G. przewalskii were captured in June 2015 in one of their spawning grounds, the Quanji River, which is connected to Lake Qinghai. Reproductively dormant male (body weight = 208.1 ± 60.11 g) and female (body weight = 197.8 ± 43.46 g) naked carp were collected in Lake Qinghai in November 2015. The field study was authorized by the Qinghai Provincal Bureau of Fishery. All the fish were sexually mature and over 3 years old. The age of each fish was estimated based on its anal scales, dorsal fin spines and otolith, according to a previous publication . Fish samples were euthanized with 200 mg/L MS222 (Sigma, USA), and tissues were harvested and transferred immediately to liquid nitrogen for RNA extraction. The animal experiment was conducted following the procedures and guidelines described in the “Guidelines for Animal Care and Use” manual approved by the Animal Care and Use Committee, Northwest Institute of Plateau Biology, Chinese Academy of Sciences.
Total RNA was purified from the brains and gonads of both male and female G. przewalskii in RS and NRS using TRIzol Reagent (Invitrogen, USA), followed by treatment with RNase free DNase I (Thermo Scientific, USA) to remove DNA contamination. The quality and quantity of RNA were verified with 1% agarose gel, NanoPhotometer® spectrophotometer (Implen, USA), Qubit® 2.0 Fluorometer (Life Technologies, USA), and Agilent Bioanalyzer 2100 system (Agilent Technologies, USA).
Library preparation and sequencing
For each sample, we purified RNA from both brain and gonad tissues to construct 2 sequencing libraries. The sequencing included 3 biological replicates of two genders collected in reproductive and nonreproductive seasons. In total, 24 libraries were sequenced. A total of 1.5 μg RNA was used for sequencing library construction. Sequencing libraries were built using the NEBNext® Ultra™ RNA Library Prep Kit for Illumina® (NEB, USA) according to the manufacturer’s instructions. The sequencing tags were ligated to the sequence library of each sample. Briefly, mRNAs were purified from total RNAs using poly-T oligo-attached magnetic beads, which were fragmented using divalent cations in NEBNext First Strand Synthesis Reaction Buffer (5×). The first-strand cDNA was synthesized using random hexamer primer and M-MuLV Reverse Transcriptase (RNase H−), and the second-strand cDNA was synthesized using DNA polymerase I and RNase H−. Blunt ends were generated, and then, A was added at the 3′ ends, followed by the ligation of NEBNext adaptors for hybridization. The library molecules were selected for cDNA fragments of 150–200 bp. For library amplification, the PCR was carried out with Phusion High-Fidelity DNA polymerase (NEB, USA), universal PCR primers and index (X) primer. After purification of the PCR products, the library quality was measured by Qubit® 2.0 Fluorometer (Life Technologies, USA) as well as Agilent Bioanalyzer 2100 system (Agilent Technologies, USA). In total, 24 RNA-seq libraries were constructed and clustered using the cBot Cluster Generation System by TruSeq PE Cluster Kit v3-cBot-HS (Illumina, USA) based on the manufacturer’s recommendations. Finally, the 24 libraries were sequenced with an Illumina HiSeq™ 4000 platform by GeneDenovo Biotechnology Co. (Guangzhou, China).
De novo assembly and annotation
The raw data were separated and processed to remove adapters, poly-N containing reads (N% > 10%) and low-quality reads (quality score ≤ 10) using custom Perl scripts. Trinity software was applied with the default parameters to assemble clean reads . The assembled transcriptome was annotated by 4 public databases, including Nr, KOG/COG (http://www.ncbi.nlm.nih.gov/COG/), Swiss-Prot (http://www.expasy.ch/sprot) and KEGG (http://www.genome.jp/kegg/) . We used the BLASTx program (http://www.ncbi.nlm.nih.gov/BLAST/) with an E-value less than 1e-5 as the threshold.
DEG expression analysis
Gene expression levels were estimated using RSEM, and RPKM was calculated to represent the expression level for each gene . To obtain the gene expression value for each sample, we first removed rRNA reads from the clean reads. The filtered, high-quality clean reads were aligned to the assembled reference transcriptome using Bowtie2 , and read counts for each gene were recorded. The EdgeR package (v3.20.1) was used to analyze the read counts to determine differential expression. To control false discovery rate (FDR), p-values were adjusted based on Benjamini and Hochberg’s approach. Finally, genes with an adjusted p-value less than 0.05 and an absolute value of fold change greater than 2 were defined as DEGs . A clustering analysis was performed using pheatmap package in R, based on the RPKM of the DEGs.
DEG enrichment analysis
Gene Ontology (GO) enrichment analysis for all DEGs was performed separately for down- and up-regulated genes by Blast2GO software . The functional classification of unigenes was performed using WEGO software . The significantly enriched GO terms were considered those with a p-value less than 0.05. The DEG-related KEGG pathways were obtained using KOBAS software , and pathways with a p-value less than 0.05 were defined as significantly enriched pathways.
Gene expression validation by RT-qPCR
The cDNA was synthesized using 1 μg total RNA with the M-Mul First Strand cDNA Synthesis Kit (Sangon Biotech, China). In the reverse transcription (RT) step, PCR water was used to replace RNA samples as the RT control. The PCR mixture included diluted RT products (1:5), SGExcel FastSYBR Mixture (Sangon Biotech, China) with forward and reverse primers in a final volume of 20 μl, which was incubated in ABI ViiA™ 7 platform (Applied Biosystems, USA). For each gene, 3 fish samples of similar weights were selected within each group as biological replicates to perform RT-qPCR validation. The RT-qPCR experiment was repeated 3 times as 3 independent experiments. The RT-qPCR conditions were as follows: initial incubation at 50 °C for 2 min; 95 °C for 5 min; 40 cycles at 95 °C for 5 s and an annealing temperature for 30 s; 95 °C for 10 s; melt curve detection of 65 °C for 5 s to 95 °C with an increment of 0.5 °C. Two types of control were applied, the RT control and a blank of PCR water, and no amplicon was observed in the real-time PCR controls. The gene expression level for each sample was calculated using the ΔΔCt method using as 18S rRNA internal control to normalize the data . We calculated the logFoldChange of 11 unigenes between the RS and NRS across FB, FG, MB and MG for both RNA-seq and RT-qPCR data. The Pearson correlation coefficient was estimated using all 44 pairs of data using the correlation function in the R environment. The primer information is listed in Additional file 10: Table S1.
Weighted gene coexpression network analysis
To identify potential regulatory networks in the seasonal reproduction, we performed the WGCNA package (v1.47) in R [66, 67, 68]. The gene expression data was imported into WGCNA to generate coexpression modules using the blockwiseModule function for automatic network construction with default parameters. Modules were defined as clusters of highly interconnected genes. The eigengene of the module was the first principal component of a cluster of genes within the module, which represented the module’s gene expression profile. The grey module included genes that cannot be classified into any other modules. Genes in each module were subjected to Gene Ontolgy (GO) analysis and KEGG pathway enrichment tests. The intramodular connectivity of each gene was calculated using the softConnectivity function, and genes with high connectivity were considered as the hub genes. The networks were visualized using a free online platform (http://www.omicshare.com/tools).
We thank the editor and anonymous reviewers for their constructive comments. We greatly appreciated our collaborators, Ms. Daoqin Chen and Ms. Hongfang Wu as well as Messrs. Hongchao Wang, Luxian Yu, Weiguo Zhou, Hong Zhang and Yang Wang in the Rescue and Rehabilitation Center of Naked Carp for their assistance in sample collection and packaging. We are grateful to Mr. Qichang Chen as well as Misses Xiangyan Hei, Jing Han and Shuying Zhang in the College of Kunlun, Qinghai University for their help in sample preparation and labeling. We also acknowledge the data support from “Lake-Watershed Science Data Center, National Earth System Science Data Sharing Infrastructure, National Science & Technology Infrastructure of China (http://lake.geodata.cn)” and http://www.statsilk.com/maps.
This work was supported by grants from the National Natural Science Foundation of China (31360632 and 31700325) and the Science Foundation of Qinghai Province (2016-ZJ-941Q). Dr. Fei Tian was supported by the “1000 Talents” program of Qinghai Province.
Availability of data and materials
The raw reads produced in this study were deposited in the NCBI SRA with the accession number SRP136464 under Bioproject PRJNA445618.
BX and KZ designed the experiments. BX conducted the RNA-seq and RT-qPCR analyses. FT analyzed the data and wrote the paper. SJL conducted WGCNA analysis and revised the manuscript. JS and HQ collected fish samples. All authors read and approved the final manuscript.
Ethics approval and consent to participate
The field study was authorized and supervised by Qinghai Provincial Bureau of Fishery. All animal experiments were approved by the Animal Care and Use Committees of the Northwest Institute of Plateau Biology, Chinese Academy of Sciences.
Consent for publication
The authors declare that they have no competing interest.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 3.Danzmann RG, Kocmarek AL, Norman JD, Rexroad CE 3rd, Palti Y. Transcriptome profiling in fast versus slow-growing rainbow trout across seasonal gradients. BMC Genomics. 2016;17:60.Google Scholar
- 5.Kennaway DJ , Rowe SA. Melatonin binding-sites and their role in seasonal reproduction. J Reprod Fertil. 1995;49(1):423–35.Google Scholar
- 13.Nakane Y, Ikegami K, Iigo M, Ono H, Takeda K, Takahashi D, et al. The saccus vasculosus of fish is a sensor of seasonal changes in day length. Nat Commun. 2013;4. https://doi.org/10.1038/ncomms3108.
- 16.Nocillado JN, Levavi-Sivan B, Carrick F, Elizur A. Temporal expression of G-protein-coupled receptor 54 (GPR54), gonadotropin-releasing hormones (GnRH), and dopamine receptor D2 (drd2) in pubertal female grey mullet, Mugil cephalus. Gen Comp Endocrinol. 2007;150(2):278–87.PubMedCrossRefGoogle Scholar
- 17.Cerda-Reverter JM, Sorbera LA, Carrillo M, Zanuy S. Energetic dependence of NPY-induced LH secretion in a teleost fish (Dicentrarchus labrax). Am J Phys. 1999;277(6 Pt 2):R1627–34.Google Scholar
- 18.Chen Y, Cao W. Cypriniformes, Fauna Sinica vol. III. Beijing: Science Press; 2000.Google Scholar
- 19.Wu Y, Wu CZ. Notes on fishes in the Huanghe drainage of Qinghai Province, with a faunal analysis. Acta biologica plateau sinica. 1987;7:141–53.Google Scholar
- 20.The environment and ecology in the Lake Qinghai; 2008.Google Scholar
- 21.Renwu T. The hydrologic conditions of the Lake Qinghai. Trans Oceanol Limnol. 1988;4:18–25.Google Scholar
- 27.Nakane Y, Yoshimura T. Universality and diversity in the signal transduction pathway that regulates seasonal reproduction in vertebrates. Front Neurosci-Switz. 2014;8. https://doi.org/10.3389/fnins.2014.00115.
- 28.Isorna E, Pedro ND, Valenciano AI, et al. Interplay between the endocrine and circadian systems in fishes. J Endod. 2016;232(3):R141. https://doi.org/10.1530/JOE-16-0330.
- 46.Yao C, Minghan T. Role of retinoic acid in the regulation of spermatogenesis. Chin J Cell Biol. 2014;36(10):1335–43.Google Scholar
- 47.Raverdeau M, Gely-Pernot A, Feret B, Dennefeld C, Benoit G, Davidson I, et al. Retinoic acid induces Sertoli cell paracrine signals for spermatogonia differentiation but cell autonomously drives spermatocyte meiosis. Proc Natl Acad Sci U S A. 2012;109(41):16582–7.PubMedPubMedCentralCrossRefGoogle Scholar
- 56.Chen Yi-Feng HD-K, Yi-Yu C. Age discrimination of selincuo schizothoracini (gymnocypris selincuoensis) in selincuo lake, tibeten plateau. Acta Zool Sin. 2002;48(2):667–76.Google Scholar
- 66.Zhang B, Horvath S. A general framework for weighted gene co-expression network analysis. Stat Appl Genet Mo B. 2005;4. https://doi.org/10.2202/1544-6115.1128.
- 67.Langfelder P, Horvath S. Eigengene networks for studying the relationships between co-expression modules. BMC Syst Biol. 2007;1(1):54–0.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.