The late-evolving salmon and trout join the GnRH1 club

Abstract

Although it is known that the whitefish, an ancient salmonid, expresses three distinct gonadotropin-releasing hormone (GnRH) forms in the brain, it has been thought that the later-evolving salmonids (salmon and trout) had only two types of GnRH: GnRH2 and GnRH3. We now provide evidence for the expression of GnRH1 in the gonads of Atlantic salmon by rapid amplification of cDNA ends, real-time quantitative PCR and immunohistochemistry. We examined six different salmonid genomes and found that each assembly has one gene that likely encodes a viable GnRH1 prepropeptide. In contrast to both functional GnRH2 and GnRH3 paralogs, the GnRH1 homeolog can no longer express the hormone. Furthermore, the viable salmonid GnRH1 mRNA is composed of only three exons, rather than the four exons that build the GnRH2 and GnRH3 mRNAs. Transcribed gnrh1 is broadly expressed (in 17/18 tissues examined), with relative abundance highest in the ovaries. Expression of the gnrh2 and gnrh3 mRNAs is more restricted, primarily to the brain, and not in the gonads. The GnRH1 proximal promoter presents composite binding elements that predict interactions with complexes that contain diverse cell fate and differentiation transcription factors. We provide immunological evidence for GnRH1 peptide in the nucleus of 1-year-old type A spermatogonia and cortical alveoli oocytes. GnRH1 peptide was not detected during other germ cell or reproductive stages. GnRH1 activity in the salmonid gonad may occur only during early stages of development and play a key role in a regulatory network that controls mitotic and/or meiotic processes within the germ cell.

Keywords: Gamete biology; Gene expression; GnRH; Immunolabeling; Ovary; Testis.

Fig. 1

The complete GnRH1 cDNA and proximal promoter of the Atlantic salmon gnrh1 gene. a The full-length GnRH1 cDNA and amino acid sequences. The coding regions for the signal peptide (22 aar), GnRH decapeptide (in bold), amidation/proteolytic cleavage site (underlined) and GnRH-associated peptide (57 aar) are shown. The 5′- and 3′-untranslated regions are shown in lower case. The polyadenylation signal is shown in bold. b The proximal promoter and transcription and translation start sites of the gnrh1 gene. Several different composite regulatory elements are shown. Two composite POU-domain elements could engage multiple members of the diverse POU family of transcription factors (Table 1). An element with motifs that could bind WT-I, EGR-1 and SP-1 (shown in orange) with two flanking SMAD half-sites is presented (−410 to −379). The two potential SMAD half-sites and a SMAD palindromic sequence are underlined and in bold. Two composite elements that could recognize SOX-2 and OCT-4 are underlined. There may be two TATA boxes associated with transcription of gnrh1: one that may bind TBP more weakly (1) (−128 to −122) and one with canonical sequences (2) (−29 to −23), respectively. Immediately preceding each TATA box are stretches of DNA that could interact with WT-1, EGR-1 and SP-1 (orange). The motifs AAGCTGC (asterisks below) and AGTGGAG (underlined) in proximity to the TATA boxes may facilitate assembly of the core transcriptional machinery. Translation of preproGnRH1 begins at the start codon (ATG) (green) and the mature hormone processing site (GKR) is underlined. Positions of base pairs (bp) and nucleotides (nt) are based on position relative to the transcription start site

Fig. 2

gnrh1 comparative analysis. a Atlantic salmon homeologous regions surrounding gnrh1 on chromosomes Ssa20 and Ssa24. Only gene symbols are shown. Synteny between the regions is shown with dotted lines. The region on Ssa24 did not have an annotated gnrh1 gene. b Splign alignment of the gnrh1 mRNA sequence to Ssa20 (top) and Ssa24 (bottom). Based on the Splign alignment, the first exon is non-coding and there has been significant accumulation of mutations in the other two exons for the gnrh1 copy on Ssa24. c Comparison of homeologous copies of gnrh1 in Atlantic salmon, a Salvelinus spp., and sockeye salmon. Each genomic sequence was aligned to the whitefish mRNA (AY245104.2). d A dotplot of alignment of homeologous regions Ssa20 and Ssa24 surrounding gnrh1 (genomic position indicated in a). The position of gnrh1 is shown at the bottom. npy8br neuropeptide Y receptor-Y8b, actr1b beta-centractin-like, kctd9 BTB/POZ domain-containing protein KCTD9-like (potassium channel tetramerization domain containing 9a), gnrh1 gonadotropin releasing hormone 1, ankrd39 ankyrin repeat domain 39, fgfr1 fibroblast growth factor receptor 1-A-like, lgi3 leucine-rich repeat LGI family member 3-like

Fig. 3

Constitutive expression of five different gnrhs in various tissues of adult Atlantic salmon. a Constitutive expression of gnrh1. Relative quantities are presented as mean ± standard error (n = 4 per sex) and are placed in descending order. Different letters denote significant (p < 0.05) differences between tissues (1-way ANOVA; data pooled between sexes). The asterisk represents significant (p < 0.01) differences between sexes for a given tissue (t-test). The number in parentheses below the asterisk is the mean fold-change between sexes. Expression of gnrh1 was not detected in liver tissue and in two of the muscle samples (one male and one female). b Constitutive expression of gnrh2a, gnrh2b, gnrh3a and gnrh3b. Relative quantities are presented as mean ± standard error (n = 4 per sex). Different letters denote significant (p < 0.05) differences between tissues (gnrh2a, t-test; gnrh2b and gnrh3a, 1-way ANOVA; data pooled between sexes). The asterisk represents significant (p < 0.05) differences between sexes for a given tissue (t-test). The number in parentheses below the asterisk is the mean fold-change between sexes. Expression was not detected in gill, heart, head kidney, posterior kidney, liver, gonad, stomach, pyloric caecum, midgut, hindgut, skin or fin tissues for all four transcripts. ND not detected

Fig. 4

LM images of Richardson-stained MBM-embedded sections showing the general morphology of the ovaries. a Low magnification section of a 1-year-old ovary containing oocytes during primary growth stages. Scale bar 100 μm. b A higher magnification view of a showing more clearly the morphology of the nucleoli near the membranes of the PGO nuclei. Scale bar 50 μm. c A section of a 2.5-year-old ovary showing two large oocytes together with a few PGOs. PGO denotes an oocyte examined more closely in image d. Scale bar 500 μm. d A higher magnification view of the PGO from c shows the perinuclear nucleoli and follicle cells much more clearly. Scale bar 50 μm. e A section of a large, mature oocyte in comparison to the maturing and smaller oocytes imaged for c. In this section, the yolk proteins have filled the ooplasm. Scale bar 500 μm. f A magnification of e showing the zona radiata surrounding the mature oocyte that is overlaid with follicle cells. Scale bar 100 μm. PGO primary growth oocyte, ST stroma, NO perinuclear nucleoli, FC follicle cells, YP yolk protein granules, ZR zona radiata proteins

Fig. 5

LM images of Richardson-stained MBM-embedded sections showing the general morphology of the testes. a A section from a 1-year-old testis shows developing seminiferous tubules containing type A spermatogonia (black arrows) surrounded by Sertoli cells. Leydig cells, blood vessels and red blood cells are shown between the developing tubules (white arrows). b A section from a 2.5-year-old testis showing different germ cell types from late type B spermatogonia to spermatozoa within different cysts. Scale bar 50 μm applies to each image. LBS late B spermatogonia, PS primary spermatocytes, SS secondary spermatocytes, SD spermatids, SZ spermatozoa

Fig. 6

Representative fluorescent and phase-contrast images showing labeling of MBM-embedded ovarian sections. The left sides of each panel a–d are the fluorescence results while the right sides are their corresponding phase contrast images. a One-year-old ovarian section incubated with no primary. b One-year old ovarian section incubated with GF-6 primary. c Two-and-a-half-year-old ovarian section incubated with GF-6 primary. The scale bar 50 μm in c applies to the panels a–c. d Two-and-a-half-year-old ovarian section incubated with GF-6 primary. Scale bar 100 μm. PGO primary growth oocytes, YP yolk protein granules, ZR zona radiata proteins

Fig. 7

Representative fluorescent and phase-contrast images showing labeling of MBM-embedded testicular sections. The left sides of each panel a–c are the fluorescence results while the right sides are their corresponding phase contrast images. a One-year-old testicular section incubated with no primary b One-year-old testicular section incubated with GF-6 primary c Two-and-a-half-year-old testicular section incubated with GF-6 primary. The scale bar 50 μm in c applies to all panels

Fig. 8

TEM image of stained Epon-embedded sections showing the general morphology of the testes. a A section from a 1-year-old testis shows developing seminiferous tubules containing type A spermatogonia surrounded by Sertoli cells. Leydig cells, blood vessels and red blood cells are shown between the developing tubules. Scale bar 10 μm. The arrow points to the area in a that is shown at a higher magnification in b. Scale bar 4 μm. tA type A spermatogonia, SC Sertoli cells, LC Leydig cells, RBC red blood cells

Fig. 9

Examination of GF-6 labeling of MBM-embedded testis sections by TEM. a Low magnification image showing developing seminiferous tubules. Scale bar 5 μm. The enclosed areas labeled “b” and “c” show the areas highlighted below. b Higher magnification of “b” from a. The left side shows the raw data and the right side shows positive GnRH immunolocalizations. Electron dense colloidal gold markers have been false colored red to enhance their appearance within the image. Scale bar 1.0 μm. c Higher magnification of “c” from a. The left side shows the raw data and the right side shows positive GnRH immunolocalizations. The colloidal gold locations have been colored red for clarity. Scale bar 1.0 μm. tA type A spermatogonia, SC Sertoli cells, LC Leydig cells, RBC red blood cells

Fig. 10

A schematic diagram presenting the organization of the genes encoding the three distinct GnRH forms of the salmonids. The sections that encode the preprohormone are shown by the signal (green box), GnRH (black box) and GnRH-associated peptide (GAP) (purple boxes). Promoter and intronic sequences are represented by horizontal black lines and untranslated regions by white boxes. Some variation exists in the lengths of the sequence that encodes the GnRH1 preprohormone in exon 2 and its terminus in exon 3 among the salmonids (see text for more details). The lengths of each segment of the GnRH2 and GnRH3 genes were averaged and a composite for each paralog is presented