Crosstalks between NOD1 and Histone H2A Contribute to Host Defense against Streptococcus agalactiae Infection in Zebrafish

Correlation studies about NOD1 and histones have not been reported. In the present study, we report the functional correlation between NOD1 and the histone H2A variant in response to Streptococcus agalactiae infection. In zebrafish, NOD1 deficiency significantly promoted S. agalactiae proliferation and decreased larval survival. Transcriptome analysis revealed that the significantly enriched pathways in NOD1−/− adult zebrafish were mainly involved in immune and metabolism. Among 719 immunity-associated DEGs at 48 hpi, 74 DEGs regulated by NOD1 deficiency were histone variants. Weighted gene co-expression network analysis identified that H2A, H2B, and H3 had significant associations with NOD1 deficiency. Above all, S. agalactiae infection could induce the expression of intracellular histone H2A, as well as NOD1 colocalized with histone H2A, both in the cytoplasm and cell nucleus in the case of S. agalactiae infection. The overexpression of H2A variants such as zfH2A-6 protected against S. agalactiae infection and could improve cell survival in NOD1-deficient cells. Furthermore, NOD1 could interact with zfH2A-6 and cooperate with zfH2A-6 to inhibit the proliferation of S. agalactiae. NOD1 also showed a synergetic effect in inducing the expression of many antibacterial genes, especially antibacterial pattern recognition receptors PGRP2, PGRP5, and PGRP6. Collectively, these results firstly highlight the roles of NOD1 deficiency in the regulation of immune-related and metabolic pathways, and the correlation between zebrafish NOD1 and histone H2A variant in the defense against S. agalactiae infection.


Introduction
Streptococcus agalactiae, also known as Group B Streptococcus (GBS), is a Gram-positive bacteria, which infects a variety of hosts. Among them, the most commonly studied species are humans, cattle, and fish [1]. In humans, GBS is an important cause of neonatal meningitis, pneumonia, sepsis, and other severe invasive diseases in neonates and other infants, pregnant women, immunocompromised individuals, and older adults [2][3][4][5]. In teleost, streptococcal infections are responsible for global economic losses in farmed tilapia [6]. In addition to tilapia, S. agalactiae is also a major pathogen infecting other saltwater or freshwater fish, such as Javanese medaka Oryzias javanicus [7], giant Queensland grouper Epinephelus lanceolatus [8], cultured golden pomfret Trachinotus blochii [9], and olive flounder Paralichthys olivaceus [10].
In mammals and teleost fish, the innate immune response has been shown to play a vital role in controlling in vivo growth of S. agalactiae. Upon infection, the host innate immune system senses S. agalactiae by distinct classes of pattern recognition receptors (PRRs) such as Toll-like receptors (TLRs) and NOD-like receptors (NLRs) [11][12][13][14]. It has

NOD1 Deficiency Mainly Affects Metabolism and Immune System Processes in Adu Zebrafish in Response to S. agalactiae Infection
Our previous study revealed that the significantly enriched pathways were involved in the metabolism and immune system for zebrafish larvae from WT and 1IS⁻ ⁄ ⁻ collected at 10 dpf [29]. Signaling pathways regulated by NOD1 deficiency a unclear in adult zebrafish in the case of bacterial infection. The systemic infection by the intraperitoneal injection was associated with the expansion of S. agalactiae blood, and S. agalactiae was found in the brain vasculature after 24 h post-injectio As mutations or allele in mammalian NOD1 are known to be associated with susc ity to inflammatory bowel disease (IBD), which is a chronic immune disorder of th tine [36,37], the intestines from WT and NOD1-1IS⁻ ⁄ ⁻ adult zebrafish were used fo scriptome sequencing to make clear the intestinal immune response regulated by deficiency with or without S. agalactiae infection. The results showed that a total genes were considered to be differentially expressed genes (DEGs) for NOD1-1IS⁻ ⁄ ⁻ WT group injected with PBS, 5407 DEGs at 24 hpi for NOD1-1IS⁻ ⁄ ⁻ vs. the WT gro fected with S. agalactiae, and 12,240 DEGs at 48 hpi for NOD1-1IS⁻ ⁄ ⁻ vs. the WT gro fected with S. agalactiae, respectively ( Table 1). All of the DEGs for each group

NOD1 Deficiency Mainly Affects Metabolism and Immune System Processes in Adult Zebrafish in Response to S. agalactiae Infection
Our previous study revealed that the significantly enriched pathways were mainly involved in the metabolism and immune system for zebrafish larvae from WT and NOD1-1IS −⁄− collected at 10 dpf [29]. Signaling pathways regulated by NOD1 deficiency are still unclear in adult zebrafish in the case of bacterial infection. The systemic infection caused by the intraperitoneal injection was associated with the expansion of S. agalactiae in the blood, and S. agalactiae was found in the brain vasculature after 24 h post-injection [34]. As mutations or allele in mammalian NOD1 are known to be associated with susceptibility to inflammatory bowel disease (IBD), which is a chronic immune disorder of the intestine [36,37], the intestines from WT and NOD1-1IS −⁄− adult zebrafish were used for transcriptome sequencing to make clear the intestinal immune response regulated by NOD1 deficiency with or without S. agalactiae infection. The results showed that a total of 820 genes were considered to be differentially expressed genes (DEGs) for NOD1-1IS −⁄− vs. the WT group injected with PBS, 5407 DEGs at 24 hpi for NOD1-1IS −⁄− vs. the WT group infected with S. agalactiae, and 12,240 DEGs at 48 hpi for NOD1-1IS −⁄− vs. the WT group infected with S. agalactiae, respectively ( Table 1). All of the DEGs for each group had an enrichment analysis of the KEGG pathways performed. The most DEGs were found at 48 hpi, and were mainly concentrated in the PI3K-Akt signaling pathway (154 DEGs), cytokine-cytokine receptor interaction (109 DEGs), Th17 cell differentiation (62 DEGs), and other immune-related signaling pathways. The second DEGs were found at 24 hpi, and were mainly concentrated in the complement and coagulation cascades (65 DEGs). The numbers of DEGs were the fewest for the NOD1-1IS −⁄− vs. WT group without bacterial infection, and were mainly concentrated in the complement and coagulation cascades (36 DEGs), NF-kappa B signaling pathway (16 DEGs), and other signaling pathways ( Figure 2 and Supplementary File S1). Furthermore, many signaling pathways related to lipid metabolism, including Glycerophospholipid metabolism (41 DEGs), Glycerolipid metabolism (31 DEGs), Arachidonic acid metabolism (29 DEGs), Steroid hormone biosynthesis (27 DEGs), Sphingolipid metabolism (26 DEGs), Fatty acid degradation (22 DEGs), Ether lipid metabolism (21 DEGs), Primary bile acid biosynthesis (18 DEGs), Linoleic acid metabolism (18 DEGs), alpha-Linolenic acid metabolism (14 DEGs), Biosynthesis of unsaturated fatty acids (13 DEGs), and Steroid biosynthesis (11 DEGs), were significantly enriched at 48 hpi for the NOD1-1IS −⁄− vs. WT group infected with S. agalactiae ( Figure 2 and Supplementary File S1). A previous report characterized the transcription patterns of the intestines from WT adult zebrafish infected with S. agalactiae. A number of genes involved in the PI3K-Akt signaling pathway (125 DEGs), NOD-like receptor signaling pathway (65 DEGs), T cell receptor signaling pathway (50 DEGs), NF-kappa B signaling pathway (48 DEGs), Toll-like receptor signaling pathway (45 genes), B cell receptor signaling pathway (40 DEGs), and so on, were significantly downregulated at 48 hpi, which suggested the suppression of immune responses observed in the intestines and were consistent with the increasing mortality rate from 1 dpi to 4 dpi [38]. However, in the present study, NOD1 deficiency induced the expression of numerous DEGs involved in these immune signaling pathways in the case of S. agalactiae infection ( Figure 3 and Supplementary File S1). Although many of the genes involved in the NOD-like receptor signaling pathway were regulated by NOD1 deficiency, the NOD-like receptor signaling pathway was not one of the significantly enriched pathways. The induced expressions of most immune-related DEGs in the intestines from NOD1-1IS −⁄− adult zebrafish seem very obscure for us at present, based on the fact that the reduced survival rate was observed in NOD1-1IS −⁄− adult zebrafish compared with WT counterparts with the S. agalactiae infection ( Figure 1B). Different from the results in the intestines, many KEGG pathways, including Cytokine-cytokine receptor interaction (28 DEGs), Apoptosis (25 DEGs), IL-17 signaling pathway (19 DEGs), NF-kappa B signaling pathway (17 DEGs), and Hematopoietic cell lineage (17 DEGs), were significantly enriched for down-regulated DEGs in the skins from the same NOD1-1IS −⁄− adult zebrafish (Supplementary File S2). As the gut microbiota and its secreted metabolites contribute to regulating transcription, ROS modulation, and inflammation in the gut, they thereby play substantial roles in regulating the host metabolic and immune functions [39,40], which have been shown to act on both NOD1 and NOD2 receptors [41]. S. agalactiae-induced abnormal expression of most immune-related DEGs in the intestines from NOD1-1IS −⁄− adult zebrafish might be as a result of the outcome of immune imbalances regulated by NOD1 deficiency, the effects of intestinal microbiota and S. agalactiae infection in fatal conditions (100% mortality at 4 dpi). It is interesting to further investigate the effect of piscine NOD1 on the intestinal microbiota composition and activity with or without pathogen infection. A previous report characterized the transcription patterns of the intestines from WT adult zebrafish infected with S. agalactiae. A number of genes involved in the PI3K-Akt signaling pathway (125 DEGs), NOD-like receptor signaling pathway (65 DEGs), T cell receptor signaling pathway (50 DEGs), NF-kappa B signaling pathway (48 DEGs), Tolllike receptor signaling pathway (45 genes), B cell receptor signaling pathway (40 DEGs), and so on, were significantly downregulated at 48 hpi, which suggested the suppression of immune responses observed in the intestines and were consistent with the increasing mortality rate from 1 dpi to 4 dpi [38]. However, in the present study, NOD1 deficiency induced the expression of numerous DEGs involved in these immune signaling pathways in the case of S. agalactiae infection ( Figure 3 and Supplementary File S1). Although many of the genes involved in the NOD-like receptor signaling pathway were regulated by NOD1 deficiency, the NOD-like receptor signaling pathway was not one of the significantly enriched pathways. The induced expressions of most immune-related DEGs in the intestines from NOD1-1IS⁻ ⁄ ⁻ adult zebrafish seem very obscure for us at present, based on the fact that the reduced survival rate was observed in NOD1-1IS⁻ ⁄ ⁻ adult zebrafish compared with WT counterparts with the S. agalactiae infection ( Figure 1B). Different from the

NOD1 Deficiency Regulated the Expression of Histone Variants and those Gene Variants Associated with Bile-Acid Signalling
As RIP2 deficiency impairs the expression of histones [31], and NOD1 is also localized in the nucleus and is associated with chromatin [33], we are quite interested to know

NOD1 Deficiency Regulated the Expression of Histone Variants and those Gene Variants Associated with Bile-Acid Signalling
As RIP2 deficiency impairs the expression of histones [31], and NOD1 is also localized in the nucleus and is associated with chromatin [33], we are quite interested to know whether an interaction exists between NOD1 and histone. We firstly investigated the expression regulations of NOD1 deficiency on the histones through transcriptome analysis. Among the DEGs for NOD1-1IS −⁄− vs. WT group injected with PBS, two histone variants were identified. The numbers of histone variants were increased to 6 at 24 hpi for NOD1-1IS −⁄− vs. WT group infected with S. agalactiae, and 74 at 48 hpi for NOD1-1IS −⁄− vs. WT group infected with S. agalactiae (Table 1 and Figure 4A). Especially at 48 hpi for the NOD1-1IS −⁄− vs. WT group infected with S. agalactiae, 5 H1 variants, 27 H2A variants, 8 H2B variants, 14 H3 variants, and 13 H4 variants were significantly decreased by NOD1 deficiency ( Figure 4A). All these data suggest that NOD1-RIP2 signaling regulated the transcription of histones.
Many studies have shown that bile acids and their receptors are involved in immune regulation of the host, in addition to playing a crucial role in glucose and lipid metabolism. Bile acids significantly regulate host immune function by activating G protein-coupled bile acid receptor 1 (TGR5, also known as GPBAR1) and FXR (namely NR1H4) receptors, causing changes in the host's ability to resist bacteria, virus, and parasites [42][43][44]. Bile acids also activate the PKA-NLRP3 pathway through the TGR5 receptor, thereby improving the pathogenesis of inflammatory diseases including sepsis, celitis, and type 2 diabetes mellitus by inhibiting NLRP3 inflammasome [45]. Although the correlation between the NOD1 and bile acid metabolism has never been reported, research has shown that NOD2 deficiency resulted in an increased renal excretion of bile acids, which was mediated by an increased expression of the bile acid efflux transporters MRP2 and MRP4 in mice. This regulation of NOD2 deficiency on the hepatic bile acid concentration in turn significantly affected hepatocyte death and cholestatic liver disease [46]. Furthermore, it was suggested that bile acid receptors TGR5 and FXR might be therapeutic targets for digestive diseases and inflammatory diseases, including IBD [47][48][49][50][51]. Interestingly, in teleost, we demonstrated the conserved antiviral role of GPBAR1 and its function in regulating glycerophospholipids metabolism [52]. In the present study, we found that 51, 117, and 282 lipid-related DEGs were regulated by NOD1 deficiency in the NOD1-1IS −⁄− vs. WT group injected with PBS, the NOD1-1IS −⁄− vs. WT group infected with S. agalactiae for 24 h, and the NOD1-1IS −⁄− vs. WT group infected with S. agalactiae for 48 h, respectively. Among them, 4, 4, and 30 DEGs involved in bile-acid signaling were regulated by NOD1 deficiency in NOD1-1IS −⁄− vs. WT group injected with PBS, NOD1-1IS −⁄− vs. WT group infected with S. agalactiae for 24 h, and NOD1-1IS −⁄− vs. WT group infected with S. agalactiae for 48 h, respectively (Table 1 and Figure 4B). Especially at 48 hpi for the NOD1-1IS −⁄− vs. WT group infected with S. agalactiae, 10 bile acid synthesis related genes, 8 bile acid transport related genes, and 2 bile acid receptor genes were found to be associated with NOD1 ( Figure 4C). All of these results suggest that NOD1 may regulate bile acid metabolism and influence metabolic diseases via crosstalk with bile acid receptors such as NR1H4 and GPBAR1. In future work, we would like to further study how the mutual regulation and interaction between NOD1 and bile acid receptors affect the occurrence of infectious diseases or metabolic diseases.  The expressions of 6 histones and 32 genes involved in bile acid metabolism were further confirmed by qRT-PCR in the NOD1-1IS⁻ ⁄ ⁻ vs. WT groups infected with S. agalactiae for 48 h. Nucleotide polymorphism of H2A, H2B, H3, and H4 could not be distinguished  The expressions of 6 histones and 32 genes involved in bile acid metabolism were further confirmed by qRT-PCR in the NOD1-1IS −⁄− vs. WT groups infected with S. agalactiae for 48 h. Nucleotide polymorphism of H2A, H2B, H3, and H4 could not be distinguished by qRT-PCR, and the total expressions of H2A, H2B, H3, or H4 were detected by qRT-PCR ( Figure 5A). At 48 hpi, the expression of all of the 38 genes tested was in agreement with the transcriptome data ( Figure 5). Most histone genes were down-regulated by NOD1 deficiency (Figure 5A), however more genes involved in bile acid metabolism were upregulated by NOD1 deficiency (Figure 5B). These results suggest that NOD1 deficiency does regulate the expressions of histones and bile acid metabolism-related genes.
tibiotics 2021, 10, x FOR PEER REVIEW 10 of up-regulated by NOD1 deficiency ( Figure 5B). These results suggest that NOD1 deficienc does regulate the expressions of histones and bile acid metabolism-related genes.

Interaction Network Analysis of DEGs Related to Immunity and Metabolism Regulated by NOD1 Deficiency
WGCNA correlation analysis was performed on those DEGs related to immunity an metabolism. After ranking the edge weight of each connection, the first 500 edges wer The expressions of candidate DEGs for bile acid metabolism-related genes confirmed by qRT-PCR. Data represent the means ± the SEM and are tested for statistical significance using two-tailed student's t-test. *, p < 0.05; **, p < 0.01; ns, not significant.

Interaction Network Analysis of DEGs Related to Immunity and Metabolism Regulated by NOD1 Deficiency
WGCNA correlation analysis was performed on those DEGs related to immunity and metabolism. After ranking the edge weight of each connection, the first 500 edges were selected for visualization (all edges were selected for groups less than 500). The results showed that after screening, the lowest edge weight of the NOD1-1IS −⁄− vs. WT group without infection was 0.8307. The lowest edge weight of the NOD1-1IS −⁄− vs. WT group infected with S. agalactiae for 24 h was 0.9224, and was 0.9449 for the NOD1-1IS −⁄− vs. WT group infected with S. agalactiae for 48 h. These three groups contain 107, 89, and 172 nodes, respectively, and all these nodes belong to the same module.
The results of the WGCNA correlation analysis showed that some lipid metabolism genes are closely related to immune-related genes ( Figure 6). More complement genes including C2, C3, C5, C8A, C8B, and C9 were found in the NOD1-1IS −⁄− vs. WT group without infection ( Figure 6A). Relatively few complement genes, including C3, C7, C8A, and C8G, were found in the NOD1-1IS −⁄− vs. WT group infected with S. agalactiae for 24 h ( Figure 6B). However, only C6 was found in the NOD1-1IS −⁄− vs. WT group infected with S. agalactiae for 48 h ( Figure 6C). Histone H2A and bile acid synthesis gene CYP7A1 play important roles in the NOD1-1IS −⁄− vs. WT group without infection, but not critical in NOD1-1IS −⁄− vs. WT group infected with S. agalactiae for 24 h ( Figure 6A,B). More DEGs related to lipid metabolism, including SLC10A2, SLC27A2, SLCO2B, SCP2, and so on, were found in the NOD1-1IS −⁄− vs. WT group infected with S. agalactiae for 48 h. Bile acid transport associated protein SLC10A2 and histone H2B were critical in the NOD1-1IS −⁄− vs. WT group infected with S. agalactiae for 48 h ( Figure 6C). The WGCNA-derived "hub proteins" for H2A, H2B, H3, and DEGs related to lipid metabolism are beneficial for us to capture data for identifying novel histone binding complexes and reconstructing metabolic networks.

Histone H2A Is Translocated to the Cytoplasm and Colocalizes with NOD1 in the Case of S. agalactiae Infection
Histones are synthesized in the cytoplasm, and rapidly imported into the nucleus by association with specific nuclear import receptors such as karyopherins (Kaps) or importins to promote their own nuclear localization [53][54][55]. However, after treatment with bleomycin, histone H1.2 was translocated from the nucleus to the mitochondria and colocalized with Bak in the mitochondria [56]. In apoptotic microglia, histone H3 did not stay concealed in the nucleus, and was found to leak from the nucleus into the cytoplasm [57]. In the present study, endogenous histone H2A was detected in the nucleus from the uninfected ZF4 cells, whereas no staining was observed in the cytoplasm using the anti-H2A antibody ( Figure 7A). Endogenous histone H2A was distributed in the nucleus and cytoplasm in the ZF4 cells infected with S. agalactiae using anti-H2A antibody ( Figure 7B). Furthermore, the merged images confirmed that NOD1 was colocalized with histone H2A both in the cytoplasm and cell nucleus in the case of S. agalactiae infection ( Figure 7B). Although further studies are necessary to elucidate the regulatory mechanisms of histone H2A from the nucleus to the cytoplasm, our results suggest that NOD1 appears to be strongly linked with histone H2A.   plasm in the ZF4 cells infected with S. agalactiae using anti-H2A antibody ( Figure 7B). Furthermore, the merged images confirmed that NOD1 was colocalized with histone H2A both in the cytoplasm and cell nucleus in the case of S. agalactiae infection ( Figure 7B). Although further studies are necessary to elucidate the regulatory mechanisms of histone H2A from the nucleus to the cytoplasm, our results suggest that NOD1 appears to be strongly linked with histone H2A.

NOD1 Interacts and Cooperates with Histone H2A Variant to Protect against S. agalactiae Infection
Our previous study showed the antibacterial property of the complete histone H2A against Gram-negative bacteria E. piscicida [31]. We also found that nucleotide polymorphisms existed for zebrafish and grass carp H2A, and that nucleotide polymorphisms of piscine H2A significantly affected the antibacterial activities of piscine H2A [58]. In the present study, we selected a zebrafish histone H2A variant (zfH2A-6, GenBank accession number MT726195) with an antibacterial effect against E. piscicida infection for further study. Similar to the effect of zfH2A-6 in E. piscicida infection, the overexpression of zfH2A-6 significantly inhibited the bacteria proliferation of S. agalactiae in zebrafish larvae

NOD1 Interacts and Cooperates with Histone H2A Variant to Protect against S. agalactiae Infection
Our previous study showed the antibacterial property of the complete histone H2A against Gram-negative bacteria E. piscicida [31]. We also found that nucleotide polymorphisms existed for zebrafish and grass carp H2A, and that nucleotide polymorphisms of piscine H2A significantly affected the antibacterial activities of piscine H2A [58]. In the present study, we selected a zebrafish histone H2A variant (zfH2A-6, GenBank accession number MT726195) with an antibacterial effect against E. piscicida infection for further study. Similar to the effect of zfH2A-6 in E. piscicida infection, the overexpression of zfH2A-6 significantly inhibited the bacteria proliferation of S. agalactiae in zebrafish larvae ( Figure 8A). In the case of S. agalactiae infection, NOD1 deficiency enormously impaired cell survival, whereas the overexpression of zfH2A-6 in WT and NOD1-deficient cells promoted cell survival, which suggest that the effect of zfH2A-6 in promoting cell survival is independent of NOD1. Remarkably, the overexpression of zfH2A-6 in NOD1-deficient cells completely rescued the impaired cell survival caused by NOD1 deficiency ( Figure 8B). As both NOD1 and zfH2A-6 significantly inhibited the bacteria proliferation of S. agalactiae and promoted the cell survival of infected cells, and that NOD1 deficiency decreased the expression of histone H2A, we concluded that the effect of NOD1 in the defense against S. agalactiae infection was associated with H2A. Furthermore, compared with the group alone transfected with NOD1 or zfH2A-6, the transient cotransfection of NOD1 and zfH2A-6 showed a cumulative antibacterial effect against S. agalactiae infection ( Figure 8C), which indicated that the antibacterial activities of NOD1 and zfH2A-6 against S. agalactiae infection were not functionally redundant and antagonistic.  The transcription regulation of NOD1 on the antibacterial genes in the case of S. agalactiae infection with or without the existence of zfH2A-6. Data are expressed as mean ± SEM of three independent experiments. * p < 0.05; ** p < 0.01; ns-not significant. The asterisk above the error bars indicate statistical significance using the group transfected with empty plasmid as the control group. The asterisk above the bracket indicate statistical significance between the two groups connected by the bracket. (E) NOD1 interacts with zfH2A-6. The interaction between FLAG and GFP or zfH2A-6 is used as the negative controls.
To check whether NOD1 physically interacts with the histone H2A variant, we performed co-immunoprecipitation in EPC cells. As shown in Figure 8E, no GFP (lane 1) and zfH2A-6 (lane 3) bands were observed, which confirmed that GFP and zfH2A-6 proteins were not pull-downed by FLAG. However, the pull-downed zfH2A-6 proteins by NOD1-FLAG (lane 4) were readily detected by immunoprecipitation analysis, which proved the interaction between NOD1 and zfH2A-6. Furthermore, previous studies have suggested that histones function as an important nuclear DAMP, and some TLRs contributed to histone recognition [59,60]. The interaction between NOD1 and zfH2A-6 in the present study may suggest that NOD1 is another receptor for histone recognition.
In mammals, CIITA and NLRC5 are well-known NLRs with transcriptional functions for inflammasome or MHC genes [61,62]. Our previous study showed that NOD1 was detected in the nucleus and chromatin of piscine and mammalian cells, and may have a regulatory function for gene transcription involved in the immune system [33]. The obvious co-localization and interaction of NOD1 and histone H2A, and the significant synergetic effect of NOD1 and the histone H2A variant in inducing the transcriptions of the upstream antibacterial PRRs may suggest that NOD1 and histone H2A have a strong positive feedback for regulating antibacterial signaling pathways in response to S. agalactiae infection.

Bacterial Infection for Adult Zebrafish and Sample Collection
Zebrafish wild-type AB/TU and NOD-1IS −⁄− mutants were obtained from the China Zebrafish Resource Center (CZRC), and were maintained in an aerated recirculation system at 28 • C with a photoperiod of light/dark (12 h/12 h). For bacterial infection of adult zebrafish, the healthy WT and NOD1-1IS −⁄− zebrafish at the age of 7 months were divided into two groups. Thirty WT or NOD1-1IS −⁄− zebrafish were intraperitoneally injected with 10 µL S. agalactiae (6.25 × 10 5 cfu/µL), and another 30 WT or NOD1-1IS −⁄− zebrafish for 10 µL of PBS.
Nine fish from WT or NOD1-1IS −⁄− zebrafish injected with S. agalactiae or PBS (three fish for each sample and three samples for each group) were sacrificed at 24 and 48 hpi. Samples from the intestines and skins were flash-frozen in liquid nitrogen, and used for transcriptome sequencing. Samples from the intestines were also used for qRT-PCR verification. For the plate count or colony count, the mixtures from the liver and spleen were taken and lysed in 500 µL of PBS. The diluted bacterial suspension was plated onto BHI agar, then the CFUs were counted after 16 h of incubation at 37 • C.
Twenty fish for each group were used for the survival assay. The numbers of surviving zebrafish were counted daily for 5 days. GraphPad Prism 6 was used to generate survival curves, and the log-rank test was used to test differences in survival between the WT and NOD1-1IS -/zebrafish injected with S. agalactiae.

Bacterial Infection for Zebrafish Larvae and In Vivo Antibacterial Analysis of zfH2A-6 and/or NOD1
The p3 × FLAG-CMV-14 empty plasmid, NOD1-FLAG, or zfH2A-6-FLAG, were diluted to the desired concentration of 100 ng/µL. For the in vivo antibacterial assay of histone zfH2A-6, the p3 × FLAG-CMV-14 empty plasmid or zfH2A-6-FLAG were microinjected into fertilized eggs at the one-cell stage. The typical injected volume was 2 nl. At 4 days after fertilization, the hatched larvae microinjected with p3 × FLAG-CMV-14 or zfH2A-6-FLAG were infected with 2 × 10 8 CFU/mL S. agalactiae. Ten larvae per group were collected at 24 and 48 hpi, and were rinsed and lysed in 1 mL of PBS. The diluted homogenates were plated onto BHI agar, and the CFUs were counted after 12 h of incubation at 37 • C.
For the cumulative antibacterial effect of NOD1 and histone zfH2A-6, the plasmids of p3 × FLAG-CMV-14, zfH2A-6-FLAG, and NOD1-FLAG were microinjected with combination constructs of two plasmids (1:1) into one-stage embryos. At 4 days after fertilization, the hatched larvae microinjected with indicated plasmids were infected with 2 × 10 8 CFU/mL S. agalactiae. Ten larvae per group were collected at 48 hpi, and were rinsed and lysed in 1 mL of PBS. The diluted homogenates were plated onto BHI agar, and the CFUs were counted after 12 h of incubation at 37 • C.

Bacterial Infection for Zebrafish WT and NOD1 −⁄− Cell Lines and Cell Counting kit-8 (CCK-8) Assay
Primary cell cultures were developed from caudal fins of WT and NOD1-1IS −⁄− mutants at the age of 1 month, and were successfully subcultured to the stable cell lines designated as WT or NOD1 −⁄− caudal fin-derived cell lines by the tissue block adherent method. The cells beyond 50 passages were used for the CCK-8 assay (Beyotime, Wuhan, China). Approximately 10 4 WT or NOD1 −⁄− caudal fin-derived cells were seeded overnight in 96-well plates, and then infected with S. agalactiae with the MOI of 1 after 24 h. At 6 hpi, 10 µL CCK-8 reagent was added to a 90 µL DMEM/F12 culture medium to generate a working solution, of which 100 µL was added per well and was incubated for 2 h. The absorbance at 450 nm was measured by a PerkinElmer's EnSpire Multilabel Plate Reader.

cDNA Library Construction and Illumina Deep Sequencing
The total RNA was isolated from the intestines or skins of the WT and NOD1-1IS −⁄− zebrafish injected with PBS or S. agalactiae at 24 hpi and 48 hpi using TRIzol ® Reagent (Invitrogen) according to the manufacturer's introduction. RNA integrity was assessed using an Agilent 2100 bioanalyzer (Agilent, Santa Clara, CA, USA). Samples with RNA integrity numbers (RINs) ≥7.5 were subjected to cDNA library construction using the TruseqTM RNA sample prep Kit (Illumina). To identify the DEGs between the WT and NOD1-1IS −⁄− adult zebrafish, the expression levels were measured using numbers of fragments per kilobase of transcript per million fragments sequenced (FPKM), and were identified based on FDR (false discovery rate) <0.05, log 2 FC (fold change (condition 2/condition 1) for a gene) >1 or log 2 FC < −1. The significantly enriched pathways of these DEGs were determined using the KEGG (Kyoto Encyclopedia of Genes and Genomes) database. Bonferroni correction was used to adjust p-values, and KEGG pathways with adjusted p value (Q-Value) < 0.05 were considered significantly enriched. The FPKMs of DEGs were used for the co-expression network analysis using the WGCNA packages in R [63]. The edge weights were sorted from the highest to the lowest, and the network with the top 500 edges was visualized using Cytoscape_v.3.6.1 [64].  Supplementary Table S1. 3.6. qRT-PCR Analysis of Antibacterial Genes Regulated by NOD1 and/or zfH2A-6

qRT-PCR Validation of DEGs
The plasmids of p3×FLAG-CMV-14, zfH2A-6-FLAG, and NOD1-FLAG at a concentration of 100 ng/µL were microinjected with the indicated combination constructs of two plasmids (1:1) into one-stage embryos. At 4 days after fertilization, the hatched larvae were infected with 2 × 10 8 CFU/mL S. agalactiae. Fifty larvae per group were collected at 48 hpi, and were used for RNA extraction and qRT-PCR using primers specific to the antibacterial genes, including PGRP2, PGRP5, PGRP6, defbl2, defbl3, rnasel2, rnasel3, and lyzc. The primer sequences used for qRT-PCR are described by our previous studies [31,65]. The housekeeping gene GAPDH was used for normalizing the cDNA amounts.

Fluorescence Microscopy
ZF4 cells plated overnight onto coverslips in 12-well plates were infected with S. agalactiae with a MOI of 1. At 24 hpi, the cells were washed three times with PBS, fixed for 1h at room temperature by 4% PFA, incubated overnight with the primary antibodies including anti-NOD1 [24] and anti-H2A (catalog no. 12349S, CST), incubated for 1 h with fluorescent labeled secondary antibodies (catalog no. R37121 and A11029; Thermo, Waltham, MA, USA), and finally stained with DAPI (1 µg/mL) for 15 min. The coverslips were observed with a confocal microscope (SP8; Lecia, Wetzlar, Germany).

Co-Immunoprecipitation (Co-IP) and Western Blotting
To test the interaction between NOD1 and zfH2A-6, the EPC cells were co-transfected with the FLAG empty plasmid, GFP empty plasmid, NOD1-FLAG, or zfH2A-6-GFP with the indicated combinations of two plasmids (1:1). At 48 h post-transfection, the cells were washed with ice-cold PBS three times and then lysed in a Pierce™ IP lysis buffer (Thermo Scientific™, Waltham, MA, USA #87787) containing Protease Inhibitor Cocktail (Thermo Scientific™, #78430). Co-IP was performed using a FLAG Tagged Protein Immunoprecipitation Kit (Sigma, St. Louis, MO, USA) according to the manufacturer's manual. The total lysate and eluted proteins were analyzed by Western blotting using monoclonal mouse anti-FLAG antibody (Sigma, St. Louis, MO, USA, F3165) and anti-TurboGFP polyclonal antibody (Evrogen, Moscow, Russia, CAT. # AB513).

Statistical Analysis
Significance testing in the cumulative survival analysis used log-rank test in GraphPad Prism 6. Expression data by qRT-PCR are presented as means and standard error of mean (SEM). Two-tailed Student's t-test or ANOVA were used to compare the means and SEM between groups. All of the data are representative of three biologic replications. The level of significance is shown as follows: * p < 0.05; ** p < 0.01.

Conclusions
S. agalactiae is a kind of Gram-positive bacteria that exists widely in the natural world, and can cause enormous economic losses both for saltwater and farmed freshwater fish. In the present study, we firstly reported that NOD1 and histone H2A variant contribute to host defense against S. agalactiae infection in teleost. The transcriptional regulations of NOD1 deficiency on the histone variants and those gene variants associated with bileacid signaling were characterized, along with immune-related and metabolic pathways regulated by NOD1 in response to S. agalactiae infection. Furthermore, the colocalization, interaction, and functional correlation between NOD1 and the histone H2A variant in the defense against S. agalactiae infection were confirmed. Further studies are required to address how the crosstalks between NOD1 and other histone variants or bile acid receptors affect the occurrence of infectious diseases and/or metabolic diseases.