Animals, rearing and sampling of tissue
The use of the experimental animals in this study was performed in strict accordance with the Norwegian Animal Welfare Act of 19th of June 2009, in force from 1st of January 2010. All year classes of fish reared were approved by the Norwegian Animal Research Authority (http://www.fdu.no/fdu/NARA, permit number 5741. A completed ARRIVE guidelines checklist can be viewed in [Additional File 5]. All the fish were reared and sampled at Matre Aquaculture Research Station, Matredal, Norway. Four groups of fish were applied, of which groups 1-3 have been published in previous studies; consequently, only 3 experimental animals were sacrificed solely for the current study (group 4).
Group 1: 7 Atlantic salmon males, 1 year old, reared and sampled as described previously [3]. The following tissues were included for RNA sequencing to screen for genes expressed in gonadal somatic cells: GCF/dnd-knockout (n=4) and WT testis (n=3).
Group 2: 6 Atlantic salmon (3 females, 3 males), 1-2 years old, reared and sampled as described previously [9]. The following tissues were included for a qPCR analysis to screen for inha and gsdf expression in various parts of the body: ovary, testis, brain, eye, muscle, fin, heart, gill, spleen, gut, liver, head kidney, skin (Fig. 4).
Group 3: 83 Atlantic salmon (24 GCF/dnd-knockout, 10 immature and 9 early vitellogenic females; 22 GCF/dnd-knockout, 9 immature and 9 mature males), 2 years old, reared and sampled as described previously [22]. The following tissues were included for qPCR analysis to reveal if inha and gsdf expression in the gonads change through puberty: GCF, immature and early vitellogenic ovary, and GCF, immature and mature testis (Fig. 5). The selected gonadal stages have been described previously based on gonado-somatic index, plasma levels of sex steroids and histology [22].
Group 4: 3 Atlantic salmon (1 immature and 1 vitellogenic female; 1 immature male), reared in indoor tanks under standard rearing conditions and a maturation inducing regime [65]. The fish were fed ad libitum with a standard commercial diet. Prior to sampling, all fish were anesthetized with 2 ml/L finquel vet, and sacrificed by cutting into the medulla oblongata, the connection between the spinal cord and the skull. Gonad tissue was collected and fixed in 4% paraformaldehyde in PBS over-night at 4°C. Subsequently, the samples were washed 2x2 hours in PBS, immersed in 25% sucrose in PBS, and stored over-night at 4°C. Finally, the samples were embedded in Tissue Tek and stored in -80°C until ISH. The following tissues were included for ISH to reveal the cellular localization of gsdf and inha transcripts in the gonads: immature testis and ovary, and vitellogenic ovary (Fig. 6).
In silico analysis and RNA sequencing
Firstly, to identify genes expressed specifically in salmon gonads, we screened available salmon multi-tissue (ovary, testis, brain, eye, gill, gut, head kidney, heart, liver, muscle, nose, ovary, pyloric caecum, skin, and spleen) transcriptome sequencing data (GenBank GBRB00000000.1; [23]). Genes with expression in extra-gonadal tissues were removed from the analysis.
Secondly, to identify which of the gonad specific genes that were expressed in the somatic part of the gonads, and which of them that were expressed within germ cells, we sequenced total RNA from 3 WT and 4 GCF testis (raw sequence reads accession no. PRJNA550414; https://www.ncbi.nlm.nih.gov/bioproject/PRJNA550414) using the HiSeq.2000 sequencing platform (Illumina). RNA-seq paired end sequences were mapped with Bowtie2 against the gene model transcripts of Atlantic salmon genome (ICSASG_v2) with standard Bowtie2 parameters [66]. Raw count table for each gene was extracted using SAMtools idxstats [67]. The read counts were normalized to the total reads in the sample with the smallest number of reads. A summary of the mapping can be viewed in [Additional File 6]. KEGG pathway analysis was performed by mapping the KEGG annotated genes to KEGG pathways as described in the KEGG Mapper tool [68]. Consequently, germ cell specific genes could be identified due to lack of expression in the GCF group, while gonadal somatic genes could be identified due to expression in both the GCF and WT group.
Phylogeny and synteny
Phylogeny and chromosomal synteny were investigated for inha to confirm the identity of this gene. For the phylogenetic analysis, Inha protein sequences from the following species were retrieved from https://www.ncbi.nlm.nih.gov/: Atlantic salmon (GenBank XP_014007683.1), rainbow trout (GenBank XP_021466674.1 (1) and NP_001117672.1 (2)), zebrafish (Danio rerio) (GenBank NP_001038669.1), Japanese medaka (Oryzias latipes) (GenBank XP_020564073.1), Common carp (Cyprinus carpio) (GenBank XP_018966331.1), Ballan wrasse (Labrus bergylta) (GenBank XP_020497817.1), Atlantic herring (Clupea harengus) (GenBank XP_012676518.1), Spotted gar (Lepisosteus oculatus) (GenBank XP_015214647.1), Chicken (Gallus gallus) (GenBank NP_001026428.1) and Xenopus (GenBank NP_001027522.1 (1), OCT63342.1 (2) and AAI70257.1 (3)). The sequences were imported into MEGA version X [69], and sequence alignments were performed using MUSCLE (Multiple Sequence Comparison by Log-expectation). All positions containing gaps and missing data were eliminated. The phylogenetic analysis was performed applying the Neighbor-Joining method [70]. A bootstrap test [71] was used to test the reliability of the inferred trees.
For the chromosomal synteny, BLAST searches for orthologous genes for salmon inha (GenBank 106575603) were performed at https://www.ncbi.nlm.nih.gov/. The following fish species were included in the analysis: rainbow trout (Genbank 110528842 and 100135804), zebrafish (GenBank 570520), Japanese medaka (GenBank 101157430) and Ballan wrasse (GenBank 109990152). The Atlantic salmon sequences and annotation were obtained from the official genome annotation (NCBI Salmo salar Annotation Release 100).
RNA extraction, cDNA synthesis and Real-time, quantitative PCR
For the group 1 samples, RNA extraction was performed using the MiRNeasy Mini kit (Qiagen), according to the manufacturer’s instructions. Up to 50 mg tissue was used for each sample, and the extracted RNA had absorbance ratios 260/280 of 1,9-2,1 (NanoDrop Spectrophotometer/ThermoFisher Scientifics), and RNA intergrity numbers 7,6-9 (Bioanalyzer/Agilent Technologies). For the group 2 and 3 samples, RNA was extracted and DNase-treated as described previously ([9] and [22], respectively). cDNA was synthesized from 125 ng RNA (group 2 samples) and 500 ng RNA (group 3 samples) using the Superscript VILO cDNA synthesis kit (Invitrogen), according to the manufacturer’s instructions. Primers and probe sequences for inha and gsdf were designed online (https://www.genscript.com/ssl-bin/app/primer (Genscript®)), and can be seen in Table 3. Primers and probe sequences for the housekeeping gene elongation factor 1-alpha (ef1α) were published previously [72]. QPCR was performed in duplicates in 384-well optical plates in a QuantStudio 5 Real-Time PCR system (ThermoFisher Scientific) (all group 3 samples, and group 2 samples for gsdf) or SDS 7900HT Fast Real-Time PCR system (Applied Biosystems) (group 2 samples for inha) using default settings. One µl cDNA was used in a 5 (gsdf) or 10 µl (inha) Fast Taqman qPCR reaction (ThermoFisher Scientific). No-template controls for each gene were run in all qPCR plates. The relative gene expression level was calculated using the comparative Ct (or 2−ΔΔCt) method. All values were normalized to ef1a and calibrated to the average ΔCt of the testis tissue (group 2 samples) or the immature WT gonads (group 3 samples).
In situ hybridization
A PCR product was generated using gsdf or inha gene-specific primers (Table 3) for the synthesis of cRNA-probes for ISH. The PCR products were sequenced following a PCR cleanup using illustra ExoProStar 1-Step (GE Healthcare Life Sciences), according to the manufacturer’s instructions. The returned sequences were blasted in the NCBI database (https://www.ncbi.nlm.nih.gov/) against several species, confirming that the primers had amplified the gsdf and inha genes. ISH cRNA antisense and sense probes were synthesized from 1 µg PCR product applying primers containing Sp6 or T7 sequence, respectively (Table 3), together with a digoxigenin-alkaline phosphatase (DIG-AP) RNA Labeling Kit (SP6/T7) (Roche Diagnostics). The probes were precipitated, washed, and resuspended in MilliQ water as described by Weltzien et al. [73]. Probe size and quality were checked with a Bioanalyzer (Agilent Technologies), and the DIG incorporation in the probes was inspected by performing a spot-test. ISH by DIG-AP was performed as described by Weltzien et al. [73].
Table 3 Primer and probe sequences
Gene
|
Forward primer
|
Reverse primer
|
Probe
|
Application
|
gsdf
|
GGCAGCATTTCAGACCACTA
|
GACAAAGCAGTGGCTGTACC
|
TGCTGCAGGACCCTCAGCCTGG
|
qPCR
|
atttaggtgacactatagGGTGAGGGTGCTGAACTCAT
|
taatacgactcactatagggTGCCATGGAGAGTTGTTGAA
|
|
ISH probe synthesis
|
inha
|
TGGTGGCTCTCTCCTCTGAT
|
ATGAGCAAGTCATCCTCTTCC
|
CCAGCTCTGGCTCTACCTGTGATAGCT
|
qPCR
|
atttaggtgacactatagGTAGGTGGTCCAGCCATCAG
|
taatacgactcactatagggTTGGACTGGTTCAAACAGCA
|
|
ISH probe synthesis
|
Primer and probe sequences (5’-3’) designed for gsdf and inha and used in this study. Sp6 and T7 sequences are shown in lower case letters.
Statistics
Statistical tests were performed using GraphPad Prism 7.02 (GraphPad Software Inc.). All qPCR datasets were tested for normal distribution using a D’Agostino & Pearson omnibus normality test. For datasets with no normal distribution, or too few n to test for normal distribution, non-parametric tests were performed to calculate differences between groups. A Kruskal-Wallis with Dunn’s multiple comparisons post test was applied for inha in different stages of ovary tissue (Fig. 5C), and expression of gsdf and inha in multi tissues of adult immature salmon (Fig. 4). An ordinary one-way ANOVA with Tukey’s multiple comparisons post test was applied for gsdf and inha expression in different stages of testis tissue (Fig. 5B, D), and gsdf in ovary (Fig. 5A).