ATAC-seq and RNA-seq analysis unravel the mechanism of sex differentiation and infertility in sex reversal chicken

Transcriptional profiling of embryonic chicken gonads

To determine systematic differences in the transcriptome landscape between embryonic chicken gonads treated with and without fadrozole, RNA-seq was performed on E10 chicken left gonad samples taken from male chickens (M, n = 7), female chickens (F, n = 8) and sex-reversed chickens (R, n = 23) (Fig. 1a). The number of clean reads and the mapping statistics of each sample were displayed in Additional file 4: Table S1, which showed the high reliability of the data. A Pearson’s correlation heatmap illustrated high correlation within each group (Additional file 1: Fig. S1a). A principal component analysis (PCA) was carried out based on the gene expression landscape, and organized the replicates into three distinct groups corresponding to the M, R and F conditions, with PC1 representing 33.2% of the variance in the data (Fig. 1b). To identify differentially expressed genes (DEGs), we performed pairwise differential expression analysis with the three groups. The DEGs were shown in Fig. 1c and Additional file 1: Fig. S1b, c. The majority of the DEGs were located on autosomes (87%), and nearly 10% of the DEGs were on the Z chromosome. Interestingly, we found that the comparison between M and F led to a larger number of DEGs than that between M and R (Additional file 1: Fig. S1d). This difference was mainly reflected in the DEGs located on autosomes, indicating that fadrozole reduced the gap in the differences between genetic males and females and caused female chicken masculinization.

Fig. 1figure 1

Analysis of RNA-seq data from embryo chicken left gonads. a Schematic of experiments done in embryo and adult chicken. The developmental stages are shown below. b PCA plot of RNA-seq data from embryo chicken left gonads. c Number of DEGs and chromosomal allocation of embryo chicken left gonads, including three pairwise comparisons: male vs. female (M vs. F), male vs. reversal (M vs. R), reversal vs. female (R vs. F). Since male chicken do not have W chromosomes, the expressed genes on W chromosomes are marked as “down-regulation” in M vs. F and M vs. R. d Expression profiles of DEGs in three pairwise comparisons. Left: DEGs expression heatmap. Right: seven clusters of DEGs revealed by fuzzy c-means algorithm. The x axis represents each group, while the y axis represents log2-transformed, normalized intensity ratios in each group. e Top ten significantly enriched Gene Ontology terms for DEGs in each cluster

The DEGs in all the groups were clustered using the fuzzy c-means algorithm. In total, seven different clusters of genes were identified based on their differential expression levels among the M, R and F groups (Fig. 1d; Additional file 4: Table S2). When compared to those in the female group, a total of 1,170 DEGs (454 and 716 in Clusters 1 and 5, respectively) were significantly downregulated or upregulated, and these DEGs were expressed at similar levels between the sex-reversed and male groups. The female development-related genes DDX4 and TNFAIP6 and the male development-related genes DMRT3 and AKR1B10 were among the DEGs expressed at similar levels (Fig. 1d). A functional enrichment analysis indicated that the DEGs in Clusters 1 and 5 were mainly involved in cell morphogenesis, extracellular matrix, organic hydroxy compound metabolic process, and organic hydroxy compound catabolic process pathways (Fig. 1e). Interestingly, we identified a total of 1,634 DEGs (687 and 947 in Clusters 2 and 6, respectively) that were expressed at an intermediate level in the sex-reversed group (Fig. 1d). In Cluster 2, the expression of the DEGs in the sex-reversed group was significantly lower than that in the female group but higher than that in the male group. These genes included FOXL2, CYP19A1, CYP17A1, and FGFR3, which were enriched in the meiotic cell cycle and cellular processes involved in reproduction in multicellular organism pathways and played important roles in female sex differentiation (Fig. 1d, e). However, the DEGs in Cluster 6 presented completely opposite expressional patterns than those in Cluster 2. These DEGs included a set of male sex differentiation-related genes, such as DMRT1, AMH, and TOX3, which were involved in processes, such as tube morphogenesis (Fig. 1d, e). Moreover, in Clusters 3 and 7, the expression of a number of DEGs (491 and 836, respectively) showed no change between the female and sex-reversed groups. This gene set included WNT4, GATA6, NR5A2, SOX9, and DNAI1. (Fig. 1d). A functional analysis indicated that the DEGs in Clusters 3 and 7 were mainly enriched in gamete generation, reproductive structure development, response to hormone, and glomerulus development pathways (Fig. 1e). These results suggested that fadrozole regulated functional gene expression at different levels, affected the phenotype acquisition, and induced physical changes in the sex-reversed group.

Analysis of chromatin accessibility in embryonic chicken gonads

To explore changes in accessible regions in the genome after treatment with fadrozole, we performed an ATAC-seq analysis with the three groups. The mapping rate of the sequencing data is shown in Additional file 4: Table S3. A Pearson’s correlation heatmap illustrated that replicates were highly consistent within each group (Additional file 2: Fig. S2a). A PCA plot showed that genetic males and females could be divided into two clusters on the basis of PC1 (81.3%), and the three groups could be clearly separated on the basis of PC2 (7.4%) (Fig. 2a).

Fig. 2figure 2

Analysis of ATAC-seq data from embryo chicken left gonads. a PCA plot of ATAC-seq data from embryo chicken left gonads. b Correlation between DEGs and DARs in three pairwise comparisons. c Example of DEGs (DMRT1) showing gradually increasing expression pattern during female to male transition and their associated DARs, as well as DEGs (IGSF9B) showing gradually decreasing expression pattern during female to male transition and their associated DARs. ATAC-seq tracks are shown in the RPKM scale. The y axis of the RNA-seq boxplot shows their group name, and the x axis shows the mean rlog-normalized counts. d Venn diagrams show the shared and unique genes obtained from RNA-seq and ATAC-seq in three pairwise comparisons. e Top five enriched motifs in different biased peaks in three pairwise comparisons

Subsequently, we performed pairwise comparative analysis based on the ATAC-seq data obtained from the three groups to evaluate the differentially accessible regions (DARs). The results were shown in Additional file 2: Fig. S2b, c. The DARs between the female and sex-reversed chickens were mainly located on autosomes, and nearly 4% of the reversal-biased DARs were located on the Z chromosome. Then, we annotated the DARs onto the genome, which was divided into promoter, exon, intron, and distal intergenic regions, to identify their genomic distribution. We found that gene body regions (promoter, exon, and intron regions) harbored nearly 65% of the DARs in the R vs. F comparison. Approximately 50% of the male-biased DARs were located in promoter regions compared with those of the female or sex-reversed chickens (Additional file 2: Fig. S2c).

To establish the relationship between changes in chromatin accessibility and gene expression, we combined ATAC-seq with RNA-seq. As expected, we found a strong correlation between the DARs and DEGs in all groups by calculating fold changes after assigning open chromatin regions to the nearest DEGs (Fig. 2b). The expression of DMRT1 was gradually upregulated during the female-to-male transition and was accompanied by increased chromatin accessibility (Fig. 2c). In contrast, IGSF9B was highly expressed in the female chickens but was gradually reduced in the sex-reversed and male chickens, and was in parallel with the marked decrease in chromatin accessibility (Fig. 2c). These results suggested a strong link between chromatin accessibility and the expression of genes in embryonic chicken gonads.

To better understand how fadrozole influences gene expression, we integrated DEGs with DARs into different groups (Fig. 2d; Additional file 4: Table S4). As the Venn diagrams show, we found 1,157 upregulated DEGs carrying DARs, including genes related to male sex development, such as DRMT1, SOX9, TOX3, AMH, and NR5A2, in the M vs. F comparison. In addition, 616 downregulated DEGs carried DARs functioning as female sex-related genes, including FOXL2, CYP19A1, FGFR3, WNT4, and GATA6. Focusing on the impact of fadrozole on chromatin accessibility, we identified 166 upregulated DEGs containing DARs in the R vs. F comparison. Interestingly, DMRT1, TOX3, and SOX9 were also included in this group, which further proved that treatment with fadrozole resulted in dynamic changes in the chromatin accessibility of male sex-related genes in sex-reversed chickens. However, only 51 downregulated DEGs carrying DARs were identified in the R vs. F comparison, and they included FGFR3, WNT10A and WNT8A. Moreover, 554 upregulated DEGs carrying DARs were found between male and sex-reversed chickens, including SOX9, AMH, NR5A2 and DNAI1. In this comparison, FOXL2, CYP19A1, CYP17A1, FGFR3, and GATA6, were downregulated DEGs with DARs. These results suggested that fadrozole might indirectly influence genomic chromatin accessibility around or inside sex-related genes to regulate their expression in sex-reversed chickens.

According to previous research, fadrozole could competitively inhibit the activity of aromatase and reduce the production of oestrogen, which could cause changes in chromatin accessibility and the binding of related TFs [50]. Therefore, to further investigate the exact changes in TFs, which might influence gene expression, after treatment with fadrozole, we performed motif analysis for three different pairwise comparisons with HOMER. In the comparison between males and females, we found that the binding sites of TFs related to male gonadal and sex differentiation, such as DMRT1, DMRT6, and NR5A2, were enriched in male-biased DARs (Fig. 2e). As expected, these loci were also enriched in male-biased DARs in the replicates of the M vs. R groups. Interestingly, the CTCF motif, a DNA-binding protein that played a key role in the regulation of chromatin interactions and gene expression via cis-regulatory elements, was also significantly enriched in the male-biased DARs in the M vs. R groups [51, 52]. Compared with those in females, reversal-biased DARs showed significant enrichment of motifs associated with DMRT1, DMRT6, and NR5A2, indicating that these TFs likely contributed to the masculinization of sex-reversed chicken gonads (Fig. 2e). However, the binding loci of the GATA transcription factor family were significantly enriched in female and reversal-biased DARs compared with male birds, and LHX9, LHX2, and Jun-AP1 motifs were enriched only in female-biased DARs compared with the reversal group (Fig. 2e). Taken together, these results suggested that fadrozole significantly affected the genomic chromatin accessibility and binding of TFs related to sex determination and differentiation, which further influenced the expression of sex development-related genes in sex-reversed chickens.

Growth performance analysis of adult chickens

Based on the relationship between gene expression and phenotype, we wanted to determine whether the injection of fadrozole during the embryonic period can influence the formation of phenotype of adult chickens. Therefore, we established a sex-reversed chicken group and two control chicken groups, wild-type male and female chickens. The phenotypic sex of the hatchlings was identified by anal swelling, and the genetic sex was determined by PCR. Interestingly, we found that the phenotypic sex of all hatchlings in the sex-reversed group was male. Therefore, the rate of sex reversal was 100% (Additional file 4: Table S5).

We monitored the growth performance of the chickens, including determining their body weight and shank length, over a 30-week period and examined their reproductive tissue development at the age of 30 weeks (Fig. 3a, b). We found that the weights of the male and female chickens diverged at 9 weeks. Interestingly, the weight increase in the sex-reversed chickens was almost identical to that in the females and was significantly different than that in their male counterparts. Furthermore, as expected, the length of the shank in the sex-reversed chickens was also significantly shorter than that in the males and followed a similar growth pattern to that in the female chickens. These results suggested that the growth performance of the sex-reversed chickens was consistent with that of the female chickens and differed from that of their male counterparts.

Fig. 3figure 3

Growth performance and phenotype of adult chicken. a Body weight of adult chicken. Asterisks indicate a statistically significant difference in body weight between reversal and male chicken. p < 0.001. b Shank length of adult chicken. Asterisks indicate statistically significant differences in shank length. p < 0.001. N.S. indicate not statistically significant differences in shank length. p > 0.05. c Physical and gonadal appearance of adult chicken. Left: secondary sexual characteristics of adult chicken. Hackle feathers are marked by black arrow head and wattles are marked by asterisk. Right: left gonadal appearance and inner structure of adult chicken. Late-stage spermatid can be observed in male seminiferous tubules (black arrow). Bar = 50 μm. d Ultrastructural analysis of spermatic ultrathin sections from reversal and male chicken. Clear acrosome (A), perforatorium (P) and nucleus (N) can be observed in male sperm with slightly waved plasma membrane (black arrow). Bar = 2 μm

At the age of 30 weeks, the sex-reversed chickens were divided into three different groups (Degree 1, Degree 2, and Degree 3) based on their secondary sexual characteristics, where the appearance of masculinized characteristics was increasingly evident from Degree 1 to Degree 3 (Fig. 3c). In male group, chickens grew larger combs and wattles and well-developed leg spurs, possessing typical hackle feathers and sickle feathers. The sex-reversed chickens in the Degree 3 group were almost identical in appearance to the males, with large combs, wattles, and leg spurs. However, the Degree 2 group exhibited an ambiguous appearance, showing neither typical male nor female secondary sexual characteristics, especially regarding feather patterns, and they developed much smaller leg spurs than the males. The Degree 1 group exhibited a classic female-type feather pattern, and these chickens were overtly female in appearance. Analysis of the reproductive systems of the sex-reversed chickens revealed that the gonads of the Degree 3 group were approximately symmetrical but much smaller and less ovoid in appearance than those of the males, and the left gonad was still connected with a profoundly regressed oviduct. In the Degree 2 group, the gonads were asymmetrical, and there were some small follicles on the left side with a larger degenerated oviduct. As expected, the gonads of the Degree 1 group were almost identical to those of the female group, and these chickens could lay eggs from the oviduct.

To further examine the inner structure of the gonads in the different groups, histological sections of the gonads were prepared and stained with haematoxylin and eosin (H&E) (Fig. 3c). The reproductive tissue of the male chickens exhibited clear testicular structures with well-organized seminiferous tubules containing late-stage spermatids. The gonads of the female chickens showed a typical thick cortex with oocyte-containing follicles of different sizes. In sex-reversed chickens from the Degree 1 to the Degree 3 groups, the cortex of the gonads gradually regressed and contained fewer follicles, and the female-type vacuolated medulla formed progressively dense tubule structures. However, we failed to identify any late-stage spermatids in the tubule-like structures regardless of the degree of sex reversal. These results suggested that fadrozole exerted a slight impact on growth performance but significantly influenced the differentiation of the secondary sexual characteristics and reproductive tissues of the sex-reversed chickens.

Fertility analysis of adult chickens

To determine whether the chickens in the Degree 3 group could produce fertile sperm, we attempted to milk semen from the Degree 3 group chickens. Interestingly, we found that six sex-reversed chickens ejected semen-like secretions induced by dorsa‐abdominal massage, and only five ejected semen regularly. However, no fertilized eggs were obtained after artificial insemination (Additional file 4: Table S6). Based on these results, we wanted to determine the reason for the infertility in our sex-reversed chickens. We speculated that the high deformity rate and abnormal structure of the sperm might have influenced the fertility of the sex-reversed birds. Therefore, we compared the structure of the sperm between males and sex-reversed chickens via ultrastructure analysis (Fig. 3d). Sperm from the male chickens presented with intact acrosomes, perforators, and nuclei and showed a slightly waved plasma membrane in the postacrosomal region. However, in the sex-reversed chickens, there were fewer sperm in the semen, and the sperm failed to form acrosome structures. Together, these results suggested that treatment with fadrozole led to the maldevelopment of the gonads and the production of severely deformed sperm in the sex-reversed chickens.

Transcriptional profiling of adult chicken gonads

To explore the underlying mechanisms by which fadrozole affects adult chicken gonadal morphogenesis, and the regulatory basis of sperm maldevelopment in the sex-reversed chickens, RNA-seq was performed on samples taken from the left gonad of the 39-week-old chickens. We intentionally selected the six chickens from the sex-reversed group that could be milked to obtain a semen-like secretion, and we randomly chose six male chickens as controls. The mapping statistics were displayed in Additional file 4: Table S7. Six replicate samples in both groups showed significant correspondence (Fig. 4a). A PCA plot based on highly expressed genes showed a clear separation between the two groups (Fig. 4b).

Fig. 4figure 4

Analysis of RNA-seq data from adult chicken left gonads. a Correlation matrix of RNA-seq samples from adult chicken. b PCA plot of RNA-seq data from adult chicken left gonads. c Top 20 significantly enriched Gene Ontology terms for DEGs from adult chicken left gonads in M vs. R. d Venn diagrams show the shared and unique genes obtained from RNA-seq of embryo and adult chicken left gonads and ATAC-seq of embryo chicken left gonads in M vs. R

By analysis and filtering, we detected a total of 4,327 DEGs, including 2,539 upregulated genes and 1,788 downregulated genes in the M vs. R comparison (Additional file 3: Fig. S3a). However, the majority of the DEGs were located on autosomes, and approximately 12% resided on the Z chromosome (Additional file 3: Fig. S3b). A functional enrichment analysis, as expected, indicated that the upregulated genes were primarily involved in sperm development processes, such as sperm flagellum, spermatogenesis, and the mitotic cell cycle pathways (Fig. 4c). The downregulated genes were enriched in cell activation, lipid binding and response to hormone pathways (Fig. 4c). These results suggested that in the adult sex-reversed chickens, the expression level of spermatogenesis-related genes was significantly lower than that in the male chickens.

To examine the influence of fadrozole on the development of the gonads and sperm during this 39-week period, we integrated RNA-seq data from both embryonic and adult gonads with the ATAC-seq data (Fig. 4d). As the Venn diagrams show, 257 DEGs were upregulated in both groups, and 170 of them harbored DARs that were enriched in 9 + 2 motile cilium, mitochondrial matrix, and regulation of the meiotic cell cycle pathways (Additional file 3: Fig. S3c). These DEGs contained SPEF2, DNAI1, and TACR3, which were related to flagellum development (Additional file 4: Table S8). However, 185 DEGs were downregulated, and 93 of them harbored DARs were involved in uterine development, urogenital system development and response to estradiol pathways (Additional file 3: Fig. S3c). These genes included ESR1 and FOXL2, which were related to ovarian development (Additional file 4: Table S8). These results indicated that fadrozole failed to sufficiently upregulate genes related to testicular development and spermatogenesis and to fully downregulate genes required for ovarian and uterine development, leading to infertility in the sex-reversed chickens.

留言 (0)

沒有登入
gif