Chinese sturgeons (30–500g) were provided by the Yangtze River Fisheries Research Institute (Taihu Lake, Jingzhou, China). Recombinantly expressed Chinese sturgeon IFN-γ protein was preserved in our laboratory (21).

Chinese sturgeon was euthanized by MS-222, then the body surface of a Chinese sturgeon in a good growth state was swabbed with alcohol, and the head kidney and middle kidney were dissected. The excised kidneys were rinsed three times with 1 × phosphate-buffered saline (PBS). After cleaning, the kidneys were placed in 75% medical alcohol for 45 s. The tissues were ground and filtered into a 15 mL centrifuge tube. M199 complete medium (Invitrogen, Carlsbad, CA, USA) was added while grinding. After grinding, tissues were centrifuged at 500 × g for 15 min, and the supernatant was discarded. The centrifugation step was repeated twice, and the cells were resuspended in M199 complete medium. Three new 15 mL centrifuge tubes were taken, and 4 mL of 52% Percoll (Solarbio, Beijing, China) and 4 mL of 34% Percoll were added successively. Finally, the cell suspension was slowly added to the upper layer of 34% Percoll. The cells were centrifuged at 500 × g at room temperature for 50 min (with slow increases and decreases in the centrifuge speed). After centrifugation, the cells formed six layers, with the macrophages forming layer four. The supernatant was removed, and the fourth layer was slowly extracted into a new 15 mL centrifugation tube. M199 medium (10 mL) was added, the cells were centrifuged at 500 × g for 15 min, the supernatant was discarded, and M199 (containing 10% fetal bovine serum (FBS; Invitrogen)) was added for re-suspension. The cells were successively divided into six-well plates and cultured normally. When the cell coverage reached 80% and growth appeared normal, the experimental groups were treated with 100 ng/mL IFN-γ in PBS (labeled as AM 1, 2, and 3, respectively), and the control groups were treated with an equal volume of PBS (labeled as AMC 1, 2, and 3, respectively); the cells were collected after 24 h of induction.

The cells with a coverage rate of more than 80% of the 25 cm² cell culture flask and in a good growth state were washed three times with 1 × PBS, the residual liquid was discarded, 1 ml of Trizol (Invitrogen) was added for lysis, and the cells were mixed by aspiration into a 1.5 ml RNase-free EP tube (Axygen, SFO, USA), and left at room temperature for 5 min to fully lyse. Then, 200 μl of chloroform (Sinopharm Chemical Reagent Co. LTD, Shenzhen, China) was added, the mixture was shaken vigorously for 30 s to mix the aqueous phase and organic phases left to stand at room temperature for 5 min, after centrifugation for 10 min at 10000 × g at 4°CC, the samples were divided into three layers (upper layer of RNA, middle layer of protein, and lower layer of cell debris). The upper aqueous phase was carefully removed into a new 1.5 ml RNase-Free EP tube, and pre-cooled isopropyl alcohol (Sinopharm Chemical Reagent Co. LTD) was added to the tube. After gentle and full mixing, the tube was stored at -20°CC for 15 min. The supernatant was discarded after centrifugation at 10000 × g at 4°CC for 10 min. Pre-cooled 75% ethanol (Sinopharm Chemical Reagent Co. LTD) was added to wash the precipitate 1-2 times (with centrifugation for 10 min at 7500 × g and 4 °CC; after adding 75% ethanol, it was only necessary to gently reverse the EP tube to precipitate and float the RNA). Put the RNA precipitation into the ultra-clean workbench, turn on the fan and blow for 2-3 min, so that the alcohol can be quickly swept away. DiethyI pyrocarbonate (DEPC; Takara, Dalian, China) water was added to dissolve the precipitate. The RNA concentration and quality in the samples were determined using a spectrophotometer and agarose gel electrophoresis, and the RNA was stored at −80 °CC for later use. High-quality RNAs were used to synthesize cDNA for quantitative real-time PCR.

After DNA digestion using DNase, the mRNA was enriched using Oligo(dT) magnetic beads. The fragmentation buffer was added to break the mRNA into short fragments. Reverse transcription PCR and random six-base primers were used to generate first-strand cDNA using the fragmented mRNA as a template, followed by a second-strand synthesis reaction to produce double-stranded cDNA, which was purified using PCR cleanup and gel extraction kits (Takara). After end repair and the addition of poly(A), the short fragments were linked with sequencing adapters, PCR amplification was performed after the fragment size was selected. After the constructed libraries were quantified with Agilent 2100 Bioanalyzer (Agilent, Santa Clara, CA, USA), they were sequenced using Illumina HiSeq 2500 sequencer (Illumina, San Diego, CA, USA) to produce 125 bp or 150 bp paired-end reads.

Raw reads in the FASTA format sequences were subjected to quality filtration to obtain high-quality reads for subsequent analysis. Firstly, Trimmomatic software (22) was used for quality control and removal of adapters. Low-quality bases and N bases were filtered out to obtain high-quality clean reads. Trinity (version: 2.4) (23) was used to splice the clean reads to produce transcript sequences using the paired-end method. The longest transcript was selected as a unigene according to their sequence similarity and length. Then, CD-HIT software (24) clustering was used to remove redundancies to obtain a final set of unigenes, which was used as the reference set for subsequent analysis.

Using BLASTX (http://www.ncbi.nlm.nih.gov/BLAST/) with an E-value threshold of 1e-5 performed the unigene functional annotation. The databases being utilized include non-redundant (NR) (http://www.ncbi.nlm.nih.gov), Clusters of Orthologous Groups (COG)/eukaryotic Orthologous Groups (KOG) (http://ftp://ftp.ncbi.nih.gov/pub/COG/KOG/), and Swiss-Prot protein (http://www.expasy.ch/sprot) databases. Based on the result of Swiss-Prot, we mapped Swiss-Prot IDs to gene ontology (GO) terms to obtain GO annotation of the unigenes. Finally, the unigenes were subjected to Kyoto Encyclopedia of Genes and Genomes (KEGG) (25) using diamond software (26), and HMMER (27) was used to compare the Pfam database (28) for functional analysis of unigene.

The FPKM (fragments per kilobase of transcript per million mapped reads) and count values of the unigene were analyzed using bowtie2 (29) and eXpress (30). The DEGs were identified using the DESeq functions estimate Size Factors and nbinom Test (31). DEGs (P-value < 0.05 and foldchange > 2) were picked out, and DEGs GO enrichment and KEGG pathway enrichment analyses were performed. Unsupervised hierarchical clustering of the DEGs was carried out, and their expression patterns among different samples were displayed in the form of heat maps.

The transcript levels of eight immunity-related genes (AsIFN-e1-3, AsIFN-γ, IRF1, 2, 3, and 7) were screened using quantitative real-time RT-PCR (qPCR) for verification. The specific primers are shown in Supplementary Table 1 .

Raw reads were obtained by high-throughput sequencing, and 268,120,424 and 280,376,796 reads were detected in the experimental group (AM 1, 2, and 3) and the control group (AMC 1, 2, and 3), respectively. After further quality control and filtering of raw reads, high-quality clean reads were obtained by Trimmomatic. There were 272,115,614 reads in the AM group and 273,895,990 reads in the AMC group. The average sequences numbers in the AM and AMC group were 90,705,204 and 210,075,147, respectively. The Q30 of the original data for each sample was 92.99-93.28%, the effective data volume was 12.00-13.40 G, and the average GC content was 46.29% ( Supplementary Table 2 ). The filtered clean reads were spliced into 88,879 unigenes, with a total length of 93,916,393 bp, a longest length of 26,838 bp, an average length of 1,057 bp, and an N 50 value of 1,727 bp ( Supplementary Table 3 ). The main length distribution was 23,506 (301~400 bp), 13,380 (401~ 500 bp) and 12,525 (>2,000 bp) ( Figure 1 ). The above data indicated that the library of primary macrophages induced by Chinese sturgeon IFN-γ is reliable and can be further studied and analyzed.

The database annotation results of the unigene were as follows: 27679 (31.14%) genes were annotated to the NR database; 23190 (26.09%) genes were annotated to the Swiss-Prot database; 16011 (18.01%) genes were annotated to the KEGG database; 17081 (19.22%) genes were annotated to the KOG database; 25138 (28.28%) genes were annotated to the eggNOG database; 20900 (23.52%) genes were annotated to the GO database; and 16700 (18.79%) genes were annotated to the Pfam database ( Figure 2A and Table 4 ). The annotation distribution of NR species showed that the similarity between Chinese sturgeon and Lepisosteus oculatu in terms of genomic homology and phylogenetic analysis was the highest, which was consistent with the homology rate of 38.42% of Lepisosteus oculatus in this experiment. Latimeria chalumnae, Scleropages formosus, and Chinese sturgeon all belong to the Osteichthyes, which occupy an extremely important position in the evolutionary history from fish to vertebrates, the homology rates between the two species and the Chinese sturgeon were 4.12% and 5.46%, respectively. Chinese sturgeon, Cyprinus carpio, Salmo salar, and zebrafish belong to the Actinopterygii, among which, Cyprinus carpio, Salmo salar and Chinese sturgeon are similar to each other and have a migratory life habit. As a model organism, zebrafish has a higher genetic homology with human, up to 87% ( Figure 2B ). Therefore, the homology of Chinese sturgeon with Cyprinus carpio, Salmo salar and Oncorhynchus mykiss was 1.57%, 1.23%, and 1.61%, respectively. In addition, 40.07% of unigene were homologous to other vertebrate genes.

A total of 8490 DEGs were screened out. The distribution of DEGs was represented by a volcano map ( Figure 3A ). Among the DEGs between the experimental group (AM) compared with the control group (AMC) 4599 were upregulated, 3891 were downregulated, and with false discovery rate value of P < 0.05 & | log₂FC | > 1 ( Figure 3B ). There may be differences between different groups under the same experimental conditions. Therefore, we analyzed the correlations among the DEGs in the three replicates in each group, which showed a strong correlation among the AM1, 2, and 3, and most of the genes with high expression were clustered in the same branch. AMC1 and AMC3, but not AMC2, showed very strong correlation in the control group. In addition, the correlation between genes in the AM group and those in the AMC group was low and the difference was large ( Figure 3C ), indicating that the screened DEGs were reliable and convincing.

After the DEGs were selected, GO enrichment analysis was performed to determine which biological processes were mainly affected by the DEGs. 4004 DEGs were detected in the GO enrichment, 1300 GO annotations were obtained, 830 were upregulated, and 470 were downregulated. The unigene annotated by the GO database can be divided into biological processes (biological processes, BP), Cellular Component (CC), molecular function (MF). The GO level 2 results were further divided into 64 subcategories. BP was divided into 23 subcategories, among which the three categories with the largest number of DEGs were: cellular process (77.54%), regulation of biological process (53.61%), metabolic process (53.08%); CC was divided into 20 subgroups, and the components with the largest proportion of DEGs were: cell (85.69%), cell part (85.38%), organelle (62.69%), membrane (42.23%); MF was divided into 21 subtypes, and the three most important ones were: binding (71.92%), catalytic activity (41.69%), molecular sensing activity (41.69%) ( Figure 4A ). The differences between the enriched DEGs and all unigenes in BP, CC, and MF were compared and analyzed to maximize the visualization of the regulatory effect of Chinese sturgeon IFN-γ on the transcription of Chinese sturgeon in GO enrichment function annotation. In BP, the expression rate of DEGs enriched in biological functions such as cell process, metabolic process, biological regulatory process, and stimulus-response was greater than 50%; in CC, the differential expression rate of DEGs in cells, cell parts and organelles was the highest; in MF, the expression rate was the highest only in the binding process ( Figure 4B ). The results indicated that Chinese sturgeon IFN-γ is involved in the three biological processes of BP, CC, and MF, and had a certain immune regulation effect in all three categories.

KEGG is a major public database related to pathways. Pathway entries with significantly enriched unigenes were found through Pathway analysis of DEGs, providing clues as to which cell pathways might be related to the differences in unigene expression between the samples. In this study, 16011 unigenes annotated by KEGG were combined, 1300 DEGs were enriched for 278 pathways, 222 pathways were significantly downregulated, and 248 pathways were significantly upregulated. The top 20 pathways with the largest number of DEGs were screened. The -log₁₀ P-value corresponding to each entry was sorted from large to small ( Figure 5A and Supplementary Table 5 ). Among them, six signaling pathways related to interferon regulation were identified: The RIGI-like receptor signaling pathway (35 unigenes) and Cytosolic DNA-Sensing Pathway (28 unigenes), NOD-like receptor signaling pathway (41 unigenes), Toll-like receptor signaling pathway (TLR) (27 unigenes), and the JAK-STAT signaling pathway (16 unigenes). Pathways enriched for DEGs and all unigenes at KEGG Level 2 level were divided into six categories: Organic systems (5297, 57.05%), metabolism (3027, 32.59%), human diseases (7187, 77.4%), genetic information processing (2229, 24.01%), environmental information processing (3277, 35.3%), cellular processes (3021, 32.54%) ( Figure 5B ). In terms of the expression trends of the unigenes, 143 unigenes were upregulated in human infectious diseases, followed by 131 unigene that were upregulated in the immune system of the organic system, and 58 unigenes were downregulated in signal transduction in environmental information processing ( Figure 5C ). These results indicated that the recombinant IFN-γ protein of Chinese sturgeon could regulate the immune pathways of Chinese sturgeon to exert its related biological functions, and these data provide a certain theoretical basis for further in-depth studies of the mechanism of IFN-γ related to immune regulation of the Chinese sturgeon.

To further analyze the function of IFN-γ in Chinese sturgeon and verify the transcriptome data, eight immunity-related genes were screened from the transcriptome, namely, AsIFN-e1, 2, 3, AsIFN-γ, IRF1, 2, 3, and 7, and analyzed using qPCR. The results showed that IFN-γ could induce the expression of IRF1, 2, 3, and 7 at 1, 8, and 24 h with the same significantly upregulated trend as observed in the transcriptome data. From 1 h to 24 h, their expression increased to a maximum of 20.75 times, 2.76 times, 8.12 times, and 20.63 times, respectively ( Figure 6A ). AsIFN- e1, 2, 3, and AsIFN-γ were induced in Chinese sturgeons. Except for AsIFN-e1, the expression of AsIFN-e2, 3, and AsIFN-γ showed the same upregulation trend as observed in the transcriptome data. The expression levels of AsIFN-e1, 2, 3, and AsIFN-γ were all lower at 1 h; however, IFN-e1, 2, 3, and IFN-γ expression levels were all upregulated to the highest degree at 24 h, by 6.25 times, 7.84 times, 1.69 times, and 4.88 times respectively ( Figure 6B ); In addition, qPCR results were similar to the up-regulation multiple of transcriptome differential genes ( Figure 6C ). The above results indicated that AsIFN-γ induced primary macrophages of Chinese sturgeons and had strong immunomodulatory effects on interferon (AsIFN-e1,2,3 and AsIFN-γ) and interferon regulatory factors (IRF1, 2, 3, and 7). The qPCR results were similar to the transcriptome data, within the range of allowable error, demonstrating the transcriptome data reliability and validity.