The Chromatin Remodeler Chd8 Regulates Hematopoietic Stem and Progenitor Cell Survival and Differentiation During Zebrafish Embryogenesis

Abstract

Chromodomain helicase DNA-binding protein 8 (CHD8), a frequently mutated gene in autism spectrum disorder (ASD), is an ATP-dependent chromatin remodeler with emerging roles in hematopoiesis. While CHD8 is known to maintain hematopoietic stem and progenitor cells (HSPCs) in the bone marrow, its function during developmental hematopoiesis remains undefined. Here, using a zebrafish model, we demonstrate that chd8 loss severely depletes the HSPC pool in the caudal hematopoietic tissue through a p53-dependent apoptotic mechanism. Furthermore, chd8^−/− embryos exhibit a p53-independent expansion of myelopoiesis. chd8 deficiency upregulates brd4, which promotes systemic inflammatory cytokine expression. Inhibiting brd4 alleviates cytokine expression, suppresses excessive myelopoiesis, and restores HSPC development. Our findings reveal a dual regulatory mechanism in which chd8 governs HSPC development by repressing p53-mediated apoptosis and constraining brd4-driven immune cell differentiation.

1. Introduction

Hematopoietic stem and progenitor cells (HSPCs) are the foundation of the vertebrate hematopoietic system, responsible for the lifelong production and maintenance of all blood cell lineages [1]. HSPCs first arise during embryogenesis from the dorsal aorta in the aorta–gonad–mesonephros (AGM) region. Following their emergence, HSPCs migrate to the fetal liver in mammals or the caudal hematopoietic tissue (CHT) in zebrafish before ultimately colonizing the mammalian bone marrow or zebrafish kidney, where they sustain adult hematopoiesis [2,3,4].

In the fetal liver or CHT, HSPCs undergo rapid expansion and differentiation. The extensive proliferation induces replicative stress, which can trigger apoptosis to eliminate compromised cells and preserve the integrity of the HSPC pool [5,6,7,8]. Conversely, concurrent survival signals are essential to maintain a balance between cell death and propagation, ensuring proper HSPC expansion and preventing excessive depletion [9]. This critical equilibrium between apoptosis and survival, and between expansion and differentiation, is tightly regulated by elaborate signaling pathways, including transcriptional and epigenetic mechanisms [5].

Chromatin remodeling enzymes are essential regulators of chromatin structure and function, playing critical roles in stem cell biology. Among these, the chromodomain helicase DNA-binding (CHD) family is a major subtype defined by the presence of two chromodomains, a central helicase/ATPase domain, and a DNA-binding domain [10]. These proteins utilize energy from ATP hydrolysis to modulate DNA accessibility for transcription factors and other epigenetic modifiers, making them indispensable for proper developmental processes [11]. Notably, several CHD proteins, including CHD2, CHD4, and CHD7, have been established as key regulators of stem cell maintenance and differentiation [12,13,14].

CHD8 was originally identified as one of the most frequently mutated genes in autism spectrum disorders (ASDs) [15,16]. Prior research has largely focused on its critical role in brain development, including regulating pluripotency and early neurodifferentiation [17,18]. Recent studies have revealed its functions in hematopoiesis. CHD8 governs HSPC survival and stemness by restricting p53 activity in the adult bone marrow [19,20], promotes erythroid differentiation [21] and regulates Treg cells [22]. Although CHD8 is expressed throughout development, its role in HSPC development during embryogenesis remains unknown because of the early embryonic lethality of Chd8-null mice [23].

We previously demonstrated that a chd8 mutation in zebrafish disrupted the intestinal epithelial integrity, leading to a microbiota imbalance that impairs HSPC development [24]. But whether chd8 intrinsically affects HSPC survival remains unknown. In this study, we found that chd8 is dispensable for the initial emergence and migration of HSPCs, but it is essential for their subsequent survival within the CHT by suppressing p53-dependent apoptosis. We also found that chd8 loss leads to p53-independent myeloid differentiation, and that Brd4 inhibition rebalances HSPC development and myeloid differentiation.

2. Results

2.1. Chd8 Is Expressed in Developing HSPCs and chd8^−/− Zebrafish Exhibit Significant Lethality

To study the intrinsic role of chd8 in hematopoiesis, we first analyzed its expression at different developmental stages and within hematopoietic cells using real-time quantitative PCR (RT-qPCR). Chd8 exhibited a dynamic expression profile, with transcript levels peaking during early developmental stages and gradually declining as embryogenesis proceeded (Figure 1A). Furthermore, chd8 expression is highly enriched in fluorescence-activated cell sorting (FACS)-purified drl-GFP⁺ hematopoietic cells [25] and CD41-GFP⁺ HSPCs [8] compared to the whole embryos. This enrichment was also observed in embryonic kidneys, the site of definitive hematopoiesis in zebrafish (Figure 1B). The specific enrichment of chd8 in HSPC-containing tissues and populations supports its potential for a cell-intrinsic function in these cells.

To study the hematopoietic development, we used two zebrafish chd8^−/− mutant lines previously established in our laboratory, chd8^M1−/− and chd8^M2−/− [24]. The zebrafish Chd8 gene and the mammalian CHD8 protein are highly conserved. The amino acid homology between zebrafish Chd8 and human CHD8 is approximately 70%. The full-length zebrafish Chd8 protein comprises 2555 amino acids, and the two lines produce truncated proteins of 800 aa and 50 aa, respectively, due to premature termination codons (Figure 1C,D). RT-qPCR analysis showed a ~50% reduction in chd8 mRNA in heterozygotes (chd8^+/−) and nearly undetectable levels in homozygous mutants (chd8^−/−) (Figure 1E). Both mutant lines exhibited significant mortality rates, reaching 35–45% by 5 days post-fertilization (dpf), with a continued increase thereafter (Figure 1F). A small fraction of mutants survived to adulthood but were notably leaner than their wild-type (WT) counterparts at 6 months post-fertilization (mpf) (Figure 1G). chd8 heterozygotes showed no significant mortality or morphological abnormalities (Figure 1E,G). The viable chd8 mutant embryos give us a useful tool to study embryonic hematopoiesis. The two mutant lines exhibited similar hematological phenotypes. Unless specifically mentioned, the subsequent results are primarily from the chd8^M1−/−, referred to hereafter as chd8^−/− for simplicity.

2.2. Loss of chd8 Does Not Affect Primitive Hematopoiesis

Similar to mammals, zebrafish hematopoiesis occurs in two consecutive waves: primitive and definitive [4]. We first investigated whether chd8 mutants exhibited defects in primitive hematopoiesis. Using whole-mount in situ hybridization (WISH), we found that the expression of myeloid markers Pu.1 (at 22 hpf), mpx and mfap4 (at 24 hpf) showed no difference between chd8^−/− mutants and their WT siblings. Similarly, the expression of erythroid progenitor marker gata 1 (at 24 hpf), and the erythrocyte markers hbbe3 and hbbe1 (at 22–24 hpf) were unaltered in chd8^−/− embryos (Figure 2). These results suggest that primitive myeloid and erythroid development remains intact in the absence of chd8, indicating that chd8 is not required for primitive hematopoiesis.

2.3. Loss of chd8 Impairs HSPC Development in the CHT

To characterize the role of chd8 in definitive hematopoiesis, we analyzed the expression of HSPC markers cmyb and runx1. At 30 and 36 hpf, the expression of c-myb and runx1 in the AGM was properly specified and slightly elevated in chd8^−/− mutants (Figure 3A,B), indicating HSPC emergence is normal or potentially enhanced. At 2 dpf, cmyb and runx1 expression remained intact, suggesting successful HSPC migration to the CHT. However, by 3 dpf, cmyb expression in the CHT was decreased, and by 5 dpf, it was strongly reduced in chd8^−/− mutants (Figure 3C,D). This reduction persisted in the embryonic kidneys at 7 dpf (Figure 3E). Taken together, these results revealed that while HSPC initiation and migration were unaffected in chd8^−/− mutants, their transient expansion within the CHT is severely impaired. We also examined heterozygous mutants and observed no hematopoietic defects, indicating that a single allele loss of chd8 does not impair hematopoiesis in zebrafish (Figure S1).

Since HSPCs originate from arterial vessels and their development is influenced by vasculogenesis and blood flow [26], we assessed vascular integrity. The expression of arterial (ephrinb2) and endothelial (flk1) markers was normal at 24 and 36 hpf (Figure 3F,G). Live imaging of the vascular plexus in Tg(flk1:EGFP) [5] embryos at 3 and 5 dpf further confirmed the normal vascular morphology in chd8^−/− mutants (Figure 3H). We also did not find any circulation defects in the mutants. These findings together support a specific effect of chd8 in definitive hematopoiesis.

2.4. Loss of chd8 Leads to Enhanced Immune Cell Differentiation

To explore the impact of chd8 on HSPC function, we analyzed a series of lineage markers in chd8^−/− embryos. WISH revealed a significant upregulation of the neutrophil markers mpx and L-plastin at 3, 4 and 5 dpf in chd8^−/− embryos (Figure 4A–C). Similarly, expression of the macrophage markers mfap4, apoE, and neutral red-stained macrophages was increased in the chd8^−/− CHT. We also observed elevated expression of the lymphoid marker rag1. These findings indicate that chd8 deficiency enhances both myeloid and lymphoid differentiation. As with HSPC, chd8 haploinsufficiency did not produce detectable abnormal phenotypes in differentiation (Figure S1).

We next asked whether the increased differentiation potential persists into adulthood. May-Grünwald-Giemsa staining of whole kidney marrow (WKM) from 6-month-old fish showed that, consistent with the embryonic phenotype, precursor cells were reduced while mature lymphoid and myeloid populations were expanded in chd8^−/− (Figure 4D,E). Tu et al. reported that Chd8 loss causes anemia in mice [21], and we observed an increase in the erythroid marker hbbe1 and benzidine-stained globin in chd8^−/− (Figure S2). But we found a significant reduction in hbaa1 expression starting from 1-month-old mutants (Figure 4F), suggesting a potential late-onset erythroid defect. These hematopoietic phenotypes were recapitulated in a second mutant allele, chd8^M2−/− (Figure S3). Thus, our results demonstrate that chd8 loss disrupts HSPCs development and lineage output.

2.5. Loss of chd8 Induces p53-Mediated Apoptosis in HSPCs

Previous studies have shown that chd8 safeguards HSPCs by suppressing p53-dependent apoptosis in the bone marrow [22]. To determine if a similar mechanism operates in zebrafish, we assessed apoptosis in chd8-deficient embryos. Quantitative analysis revealed a significant upregulation of key apoptosis-related genes, including p53, caspase-3, and baxa, in whole embryos at 1, 3, and 5 dpf (Figure 5A). This elevation was even more pronounced within sorted drl^-GFP⁺ hematopoietic cells at 5 dpf (Figure 5B), suggesting hematopoietic cell-intrinsic activation of the apoptotic program. Consistent with this, TUNEL assays demonstrated a substantial increase in apoptotic CD41^-GFP⁺ HSPCs in the CHT of chd8^−/− embryos (Figure 5C,D). These findings imply that the definitive hematopoiesis defect in chd8^−/− originates from increased HSPC apoptosis. To directly test whether p53 upregulation mediates this phenotype, we knocked down p53 in chd8^−/− embryos using a morpholino (MO) [27]. WISH showed a complete rescue of cmyb⁺ HSPCs in both the CHT and the embryonic kidney of chd8^−/− embryos (Figure 5E,F), confirming that the HSPC loss is p53-dependent. Interestingly, p53 knockdown failed to suppress the enhanced myelopoiesis in chd8^−/− embryos (Figure 5G,H), and did not inhibit increased erythroid differentiation (Figure S2D,E). In contrast, it partially rescued rag1 expression (Figure 5I,J). These results suggest the enhanced myeloid and erythroid differentiation in chd8^−/− embryos occurs largely via a p53-independent mechanism. This finding aligns with the work of Tu et al., indicating that Chd8 possesses p53-independent functions in murine erythroid development. Collectively, our data demonstrate that chd8 deficiency depletes HSPCs by triggering p53-dependent apoptosis, but inhibiting p53 restores survival without fully rescuing their lineage differentiation defects.

2.6. BET Inhibitor PFI-1 Restores HSPC Production and Differentiation

To identify additional regulators in chd8-mediated hematopoiesis, we performed a rescue screen with the Cayman Chemical Epigenetics Screening Library, based on the known role of chd8 in epigenetic regulation. Embryos from chd8^+/− intercrosses were treated with library compounds (n = 96) from 3 to 5 dpf and assessed for c-myb expression at 5 dpf (Figure 6A). Several compounds restored c-myb expression in chd8^−/−, with the BET (Bromodomain and Extra-Terminal) proteins inhibitor PFI-1 [28] exhibiting the strongest rescue effect (Figure 6B,C). BET proteins are transcriptional regulators that mainly bind to acetylated histones to control gene expression in fundamental processes such as cell proliferation, differentiation, and apoptosis [29]. To further validate their role, we employed additional BET inhibitors, JQ-1 and ABBV-075 [29]. Treatment with these compounds also significantly restored c-myb⁺-HSPC populations in chd8^−/− embryos (Figure 6D–G). We next investigated whether BET inhibition influences HSPC survival. In chd8^−/− embryos, PFI-1 treatment significantly downregulated the expression of p53, caspase-3, and baxa (Figure 6H). Consistently, the TUNEL assay demonstrated a marked reduction in apoptotic CD41-GFP+-HSPCs in the CHT following PFI-1 treatment (Figure 6I,J), indicating that PFI-I antagonizes p53-mediated apoptosis to rescue HSPC survival. Impotently, PFI-1 treatment also significantly reduced the enhanced expression of mpx and rag1 in chd8^−/− mutant embryos at 5 dpf, suggesting it also counteracts the aberrant lineage differentiation driven by chd8 loss. Notably, these drug doses did not affect hematopoiesis in WT siblings (Figure 6B–G,K–N), indicating a specific corrective effect in the mutant context. Thus, BET inhibition restores HSPC development by suppressing p53-mediated apoptosis and normalizing excessive immune cell differentiation in the chd8^−/− embryos.

2.7. PFI-1 Targets Brd4 in chd8^−/− to Restore HSPC Production and Differentiation

We next sought to identify the potential protein target through which PFI-1 exerts its rescue effect in chd8^−/−. PFI-1 is a potent BET inhibitor, particularly blocking the activity of BRD2 and BRD4 [28]. We first characterized the expression of their zebrafish homologs, brd2a, brd2b, and brd4, and confirmed their presence in hematopoietic cells (Figure S4A,B). To determine which of these is functionally required for the phenotype, we performed MO-mediated knockdown of each gene. While knockdown of brd2a, brd2b or a combination of both failed to restore HSPCs in chd8^−/− mutants (Figure S5A–F), knockdown of brd4 completely rescued HSPC production and suppressed the enhanced myeloid differentiation (Figure 7A–D). Furthermore, brd4 inhibition also partially restored lymphoid differentiation (Figure 7E,F). Moreover, simultaneous knockdown of brd2a and brd2b in the brd4 morphant background did not augment this rescue (Figure S5G–L), indicating that brd4 is the primary target of PFI-1 in chd8^−/−.

Next, we examined the relationship between Chd8 and Brd4. brd4 mRNA levels was unchanged in the whole chd8^−/− embryos (Figure 7G). To test if chd8 regulates brd4 in HSPCs, we knocked down chd8 in Tg(CD41:GFP) embryos using a MO [30], which recapitulated the chd8^−/− mutant phenotypes (Figure 7H). This knockdown significantly increased brd4 expression in CD41-GFP⁺ HSPCs (Figure 7I). Furthermore, Brd4 protein was strongly upregulated in the whole chd8^−/− embryos (Figure 7J,K), suggesting potential post-transcriptional dysregulation. The fact that chd8 loss causes brd4 upregulation, while brd4 knockdown rescues the hematopoietic defects in chd8^−/− mutants, positions brd4 as a critical downstream effector of chd8. Therefore, we conclude that Chd8 mediates definitive hematopoiesis through at least two parallel pathways: by repressing p53-dependent apoptosis and by constraining Brd4 activity.

Finally, we previously demonstrated that the microbiota dysbiosis in chd8^−/− drives elevated cytokine expression, which contributes to the host hematopoietic defects [24]. As Brd4 is known to regulate inflammatory gene expression, we asked whether Brd4 participates in the inflammatory response. Both PFI-1 treatment and brd4 knockdown successfully attenuated the systemic cytokine expression in chd8^−/− mutants, whereas p53 knockdown did not produce this effect (Figure 7L,M). Therefore, Brd4 inhibition restores normal hematopoiesis partially through mitigating the pro-inflammatory cytokine expression.

3. Discussion

Recent studies have shown that CHD8 functions as a key regulator of HSPC survival and stemness in the bone marrow via suppressing p53 activity [19,20]. However, its role in HSPC development during embryogenesis remains unknown. Our study reveals that the zebrafish chd8^−/− embryos retained primitive hematopoiesis but showed severe defects in definitive hematopoiesis in the CHT. chd8 is essential for HSPC survival by suppressing p53-dependent apoptosis. Furthermore, chd8 loss leads to enhanced myeloid differentiation through a p53-independent mechanism. Inhibiting brd4 attenuated myeloid differentiation and promoted HSPC development in chd8^−/− partially via decreasing elevated inflammatory cytokine expression. Our results demonstrate that chd8 plays critical and multidimensional regulatory roles in HSPC development.

Definitive HSPCs transiently emerge from the AGM before migrating to the CHT. Chd8 does not regulate early HSPC emergence but is critical for their survival and development in the CHT from 3 dpf. Nascent HSPCs seldom proliferate in the AGM region but enter an active in cell cycle and undergo rapid division in the CHT [7,31]. Such extensive cell division in a short timeframe poses a risk of replicative stress, potentially activating the DNA replication checkpoint and inducing p53-dependent apoptosis [7,32]. We hypothesize that chd8 functions to suppress unwanted apoptosis in proliferating HSPCs, providing a critical window for DNA damage repair and ensuring the successful expansion of the HSPC pool. Previous studies have shown that dysregulation of alternative splicing activates p53-dependent apoptosis in HSPCs [33,34]. Meanwhile, Wdr5-mediated H3K4 methylation is crucial for HSPC survival in the CHT by maintaining genomic stability [7]. Our study identified another essential regulator involved in this process. In chd8^−/−, both primitive hematopoiesis and vascular systems develop normally. And chd8 mRNA is highly enriched in CD41-GFP⁺ HSPCs. These data support that chd8 likely maintains HSPC survival through intrinsic mechanisms. Future studies, including the use of p53 mutants, are needed to formally prove that this function is cell autonomous. Furthermore, in the mouse bone marrow, CHD8 deficiency activates ATM kinase to stabilize P53 protein and CHD8 also directly binds to P53 to regulate its transactivation [20]. How Chd8 suppresses p53 activation during zebrafish development remains unknown. Future omics studies, such as those utilizing activator-targeted chromatin accessibility sequencing (ATAC-seq), will reveal the chromatin accessibility regulated by Chd8, thereby helping elucidate the mechanism by which it suppresses p53 activation.

In addition to HSPC defects, chd8^−/− zebrafish embryos displayed enhanced immune cell differentiation that persisted into adulthood. Chd8^−/− mice die between embryonic days 5.5 and E7.5 (E5.5-E7.5) [23]. Using the Mx-Cre inducible system, deletion of Chd8 in murine hematopoietic cells causes a significant reduction in HSPCs and lineage differentiation [19,20]. This discrepancy may be largely attributed to the distinction between a constitutive systemic knockout in zebrafish and a hematopoietic-specific deletion in mice. The phenotype in chd8^−/− zebrafish likely involves both cell-autonomous and non-autonomous mechanisms. Chd8 may function cell-intrinsically to maintain HSPC survival. We previously showed that chd8 mutants harbor an imbalanced microbiota that drives elevated inflammatory cytokines (e.g., tnfa and il1b) in the host, which in turn non-cell-autonomously impairs HSPC differentiation and production [24]. In this study, we found that BET inhibitors and Brd4 knockdown reduced systemic cytokine production and restored balanced HSPC development in chd8^−/−. In contrast, p53 knockdown did not affect cytokine production nor restore differentiation, but specifically rescued HSPC survival. This supports that enhanced myelopoiesis is related to elevated cytokines. Moreover, enhanced differentiation persisted in germ-free (GF) chd8^−/− embryos [24], indicating that additional, microbiota-independent, genetic mechanisms promote differentiation. Previous studies showed that Chd8 directly regulates pluripotency and early neural differentiation through chromatin structure in mouse embryonic stem cells (ESCs) and neural progenitor cells (NPCs) [17,18]. We downloaded the RNA-seq data from mouse HSPC (WT vs Chd8^−/−) [22], and found that despite no apparent enhancement in differentiation, myeloid and lymphoid differentiation and inflammatory pathways were significantly upregulated in mouse chd8^−/− HSCs (Figure S6). This suggests that Chd8 may intrinsically negatively regulate these pathways in HSPCs, possibly through chromatin compaction, thereby contributing to precise differentiation control. The differentiation defects in chd8^−/− zebrafish are likely the collective effects of HSPC-intrinsic regulation and extrinsic signaling. Future studies using tissue-specific knockdown and rescue experiments will help elucidate the cell-autonomous and non-autonomous functions of Chd8 in HSPC development. However, given that CHD8 is a high-risk gene for ASD—a condition often associated with immune dysregulation [35,36]—and its mutations are constitutive, the viable chd8^−/− zebrafish larvae provide a powerful tool for analyzing hematopoietic defects caused by CHD8 deficiency at the organismal level.

PFI-1 is a highly selective BET inhibitor that particularly targets BRD2 and BRD4 by blocking their ability to bind to acetylated histone tails—a binding crucial for gene transcription [28]. Our study revealed that in chd8^−/− embryos, PFI-1 primarily targets Brd4 to alleviate hematopoietic defects. As a key epigenetic regulator, BRD4 cooperates with various master transcription factors to control gene expression [29] and is known to positively regulate inflammatory genes such as IL1B [37,38]. Knocking down Brd4 may directly repress these inflammatory cytosines, thereby restoring the developmental balance of HSPCs. We also found that Brd4 inhibition partially rescues HSPC apoptosis in chd8^−/−, likely by mitigating the pro-apoptotic effects induced by high levels of inflammatory cytokines such as IL1B and TNFα [39,40]. Indeed, we found that MO-mediated inhibition of il1b in chd8^−/− partially reduced the expression of apoptosis-related gene (Figure S6). In addition, brd4 is known to directly promote granulopoiesis in zebrafish [41]. Thus, upregulation of Brd4 may drive myeloid differentiation directly. The exact function of Brd4 during HSPC development requires further study. Although brd4 expression remained unchanged in whole embryos, it was significantly upregulated in CD41⁺-HSPCs. Thus, chd8 may mediate brd4 expression within HSPCs. Notably, the increase in Brd4 protein levels in chd8 mutants far exceeded that of its mRNA levels, implying that Chd8 may regulate Brd4 protein synthesis or stability through post-transcriptional mechanisms. The cell type-specific regulation of Brd4 by Chd8 and the underlying regulatory mechanisms require further investigation. This study reveals that Chd8 restricts Brd4 activity during embryogenesis.

In conclusion, our findings demonstrate that Chd8 maintains HSPC development by suppressing p53-dependent activity. Furthermore, Chd8 loss drives p53-independent enhanced myeloid differentiation, and Brd4 inhibition restores normal myeloid differentiation and supports HSPC development. Our studies elucidate the critical roles of Chd8 in regulating developmental hematopoiesis.

4. Materials and Methods

4.1. Zebrafish Maintenance and Embryo Handling

Zebrafish (Danio rerio) strains, including wild-type AB and transgenic lines Tg(CD41: GFP), Tg(drl: GFP), and Tg(flk: GFP), were raised and maintained with a 14 h light/10 h dark cycle in a circulating water system at 28.5 °C, conductance 450–550 μS, and pH 7.0 ± 0.5 [42]. WT Siblings and mutant embryos were collected at the desired stage and grown in an E3 medium (5 mM NaCl, 0.17 mM KCl, 0.33 mM CaCl₂, 0.33 mM MgSO₄) with a density of 80–120 embryos per 10 cm Petri dish. Embryos were staged by hours post-fertilization (hpf) and days post-fertilization (dpf). Zebrafish were maintained, handled, and bred according to standard protocols from the Institutional Animal Care and Use Committee of Shanghai Jiao Tong University, China.

4.2. Genotyping of Mutant Lines

The chd8^−/−(M1) and chd8^−/−(M2) mutant lines were generate by CRISPR/Cas9 technology as previously described [24]. For genotyping, DNA fragments of 388 bp and 420 bp (flanking the respective chd8 target sites) were amplified by polymerase chain reaction (PCR) from genomic DNA. Mutants were identified using the T7 endonuclease I assay (NEB, M0302S) and confirmed by Sanger sequencing. The primers used to amplify the chd8 target sequences are listed in

Table 1. PCR primers used for genotyping.

Mutant Lines	Forward Primer	Reverse Primer
chd8−/−(M1)	CAAACTTTTTGACAAGATGG	ACCAAGAGAAGAGCTCAGCA
chd8−/−(2)	TTGCCTCTGTTATAGCCATGA	ACACACACTTTTGCTGGCAA

.

4.3. Whole Mount In Situ Hybridization

Whole-mount in situ hybridization (WISH) was performed on 4% paraformaldehyde (PFA) -fixed zebrafish embryos. The anti-sense probes (c-my, pu.1, runx1, hbbe1 and hbbe3, mpx, mfap4, and rag1) were synthesized by in vitro transcription using SP6 or T7 polymerase (Ambion) with Digoxigenin RNA Labelling Mix (Roche, Mannheim, Germany). The WISH procedure was carried out as previously described [43]. Stained embryos were mounted in 100% glycerol and imaged using an Olympus SZX16 stereomicroscope or a BX53 microscope (Olympus, Tokyo, Japan).

4.4. Neutral Red, Benzidine, and Sudan Black Staining

To minimize pigmentation and enhance staining signal, embryos were treated with 1-phenyl 2-thiourea (PTU) until the desired stages. For macrophage visualization, live 4–6 dpf embryos were incubated in the dark for 6–8 h at 28.5 °C with 2.5 µg/mL Neutral Red (A600652, Sangon Biotech, Shanghai, China) in E3 medium [44]. For hemoglobin detection, live embryos were incubated in the dark for 30 min in a benzidine solution, prepared by combining 2 mL of 5 mg/mL benzidine (B108444, Aladdin, CA, USA) in methanol, 16.7 µL of 3 M sodium acetate, 100 µL of H₂O₂, and 2.483 mL of H₂O [45]. Following incubation, embryos were washed with PBS containing 0.1% Tween 20 (PBT) and fixed overnight in 4% paraformaldehyde (PFA).

For Sudan Black staining, embryos at the desired stages were fixed overnight in 4% methanol-free formaldehyde (Polysciences), rinsed three times with PBS, and incubated in Sudan black B (SB; Sigma-Aldrich, Saint-Quentin Fallavier, France) solution at room temperature for 1 h [46]. Embryos were thoroughly washed with 70% ethanol and rehydrated with PBS and 0.1% Tween 20 (PBT). To better visualize the staining, embryos were incubated in a bleach solution (10% KOH, 30% H₂O₂, 10% Tween 20) for 5 min and then washed with PBT. Images of the stained embryos were taken through an SZX16 stereomicroscope or a BX53 microscope (Olympus, Tokyo, Japan).

4.5. Morpholinos

The antisense morpholinos (MOs) were purchased from Gene Tools and resuspended in nuclease-free water to a 1 mM stock solution. Embryos were injected at the 1–4 cell stage with 1–7 ng/nL of MO into the yolk. Injected embryos were raised to 5 dpf for subsequent staining. The MO sequence is provided in

Table 2. List of Morpholino (MO) used.

Target Genes	MO Sequence	Reference
p53	GCGCCATTGCTTTGCAAGAATTG	[27]
chd8	GAGAATGGAATCATAACTTACTTGA	[30]
brd2a	CCACCTGAGACTAAAACAGAGACAA	[47]
brd2b	AGACTGGTTGATGGCCGCCTCCATC	[47]
brd4	TCATGTCTAATGACACAGAAAGAGA	[47]

.

4.6. Chemical Screen

A screen of epigenetic small molecules (Cayman Chemicals, 11076) was performed at concentrations of 6, 12.5, and 25 µM, using DMSO as a vehicle control. WT and chd8^−/− mutant embryos at 3 dpf were arrayed into 24-well plates (15 embryos per well in 600 µL of E3 medium). Plates were wrapped in aluminum foil to protect light-sensitive compounds and incubated at 28.5 °C until 5 dpf. The chemicals were then washed out 3 times with E3 medium. The toxicity of the chemicals was determined, and healthy embryos at 5 dpf were then fixed in 4% PFA at 4 °C overnight (O/N) for WISH. The additional BET inhibitors were obtained commercially: PFI-1 (Lot# QLDMMEO), (+)-JQ1 (Med Chem Express, Cat# HY-13030/CS-0581), and ABBV-075 (Mivebresib).

4.7. May–Grünwald–Giemsa Staining of Adult Whole Kidney Marrow Cells

Adult zebrafish of the desired genotype were euthanized in 0.4% tricaine. The kidney was then dissected via a midline incision and placed in 0.9× PBS containing 5% FBS. Whole kidney marrow (WKM) cells were isolated by mechanical dissociation through strenuous pipetting in 0.4 mL of the same buffer, followed by filtration through a 40-µm cell strainer (Falcon, New York, NY, USA). The resulting cell suspension was centrifuged onto glass slides using a cytocentrifuge (Sigma-Aldrich) at 400 rpm for 4 min. Slides were stained with May-Grünwald and Giemsa (Sigma-Aldrich) for morphological analysis and differential cell counting.

4.8. Gene Expression by Real-Time qPCR

Total RNA was isolated from whole embryos, kidneys, and sorted cells using Trizol reagent (Thermo Fisher Scientific, 10296028, Carlsbad, CA, USA) according to the manufacturer’s instructions. RNA was reverse-transcribed into cDNA using either the PrimeScript RT reagent Kit with gDNA Eraser (Takara, RR047A, Kyoto, Japan) or the Hifair^® II 1st Strand cDNA Synthesis SuperMix for qPCR (Yeasen Biotech, 11123ES10, Shanghai, China). Quantitative real-time PCR (qPCR) was performed on a StepOnePlus system (Applied Biosystems, San Francisco, CA, USA) with TB Green Premix Ex Taq II (Takara, RR820A, Japan) or Hieff^® qPCR SYBR Green Master Mix (Yeasen Biotech, 11203ES08, Shanghai, China). Gene expression was quantified using the 2^–ΔΔCT method. The RT-qPCR primers are listed in

Table 3. Primers used for RT-qPCR.

Genes	Forward Primer	Reverse Primer
zp53	ACCACTGGGACC AAACGTAG	CAGAGTCGCTTCTTCCTTCG
zbax a	GGCTATTTCAACCAGGGTTCC	TGCGAATCACCAATGCTGT
zcaspase 3	ATGCCAAGCCTCAATCCC	TCACAATGTATCCAAGCTTTCG
zchd8	AAGGAGGACAAAGACTAGCAGTG	AAGATGCAGCTGTAGTGGTGG
zrunx1	TTTGGGACGCCAAATACG	AAACCCTCGCTCATCTTCC
zc-myb	TCGCCAGCTTTCTACCAAA	CAGGGTTGAGGACTTTCTGC
zhbaa1	CAAGGCTGTTGTTAAGGC	ATTCTGGCGAGGGCTTC

.

4.9. Flow Cytometry and Cell Sorting

The CD41 GFP⁺ and drl GFP⁺ cells from WT and chd8^−/− embryos were sorted out using FACS AriaIII (Becton, Dickinson and Company, Franklin Lakes, NJ, USA). Briefly, embryos of the desired stage were anesthetized in 0.4% tricaine, and single cells were collected by chopping the embryos with a blade. Following that, the chopped embryos were incubated for 20 min at 37 °C in 38 µg/mL Liberase (05401119001, Roche, Basel, Switzerland). The liberase reaction was stopped by adding 10% FBS, followed by filtration through a 40 µm filter and centrifugation (1500 rpm, at 4 °C for 10 min). The supernatant was discarded, and cells were resuspended in PBS containing 1% FBS for subsequent studies.

4.10. TUNEL Immunostaining

WT sibling and chd8^−/−/ Tg (CD41: GFP) embryos were fixed at 4 dpf in 4% PFA overnight at 4 °C and dehydrated in methanol at −20 °C for 1 h. After serial rehydration, embryos were permeabilized with proteinase K (20 mg/mL) at room temperature (RT) for 40 min. The apoptotic cells were stained with Bright Red Apoptosis Detection Kit (Vazyme Biotech Co., Ltd., Shanghai, China) for 90 min at 37 °C. After washing with 1XPBS, embryos were blocked for 2 h at RT in blocking buffer (PBS + 1% DMSO + 0.1% Triton X-100+ 1% bovine serum albumin). To amplify the GFP signal, embryos were incubated with an anti-GFP antibody (Dia-an, Cat# 2057), followed by Alexa Fluor 488-conjugated goat anti-mouse secondary antibody (Invitrogen, Waltham, MA, USA) incubation. Images were taken through a confocal microscope (Leica, Wetzlar, Germany).

4.11. Statistical Analysis

GraphPad Prism 8.0.2 (GraphPad Software, San Diego, CA, USA, 2019) was used to analyze the data. The values of all triplicate experiments are presented as mean ± standard deviation (SD), and significance was determined by a 2-tailed Student t-test. For WISH and other staining results, embryos were divided into strong and weak staining groups. For statistical analysis, only the strongly stained groups of wild-type and mutant/morphant were compared.