Stat3 Regulates Developmental Hematopoiesis and Impacts Myeloid Cell Function via Canonical and Non-Canonical Modalities

Abstract

Introduction

Signal transducer and activator of transcription (STAT) 3 is extensively involved in the development, homeostasis, and function of immune cells, with STAT3 disruption associated with human immune-related disorders. The roles ascribed to STAT3 have been assumed to be due to its canonical mode of action as an inducible transcription factor downstream of multiple cytokines, although alternative noncanonical functional modalities have also been identified. The relative involvement of each mode was further explored in relevant zebrafish models.

Methods

Genome editing with CRISPR/Cas9 was used to generate mutants of the conserved zebrafish Stat3 protein: a loss of function knockout (KO) mutant and a mutant lacking C-terminal sequences including the transactivation domain (ΔTAD). Lines harboring these mutations were analyzed with respect to blood and immune cell development and function in comparison to wild-type zebrafish.

Results

The Stat3 KO mutant showed perturbation of hematopoietic lineages throughout primitive and early definitive hematopoiesis. Neutrophil numbers did not increase in response to lipopolysaccharide (LPS) or granulocyte colony-stimulating factor (G-CSF) and their migration was significantly diminished, the latter correlating with abrogation of the Cxcl8b/Cxcr2 pathway, with macrophage responses perturbed. Intriguingly, many of these phenotypes were not shared by the Stat3 ΔTAD mutant. Indeed, only neutrophil and macrophage development were disrupted in these mutants with responsiveness to LPS and G-CSF maintained, and neutrophil migration actually increased.

Conclusion

This study has identified roles for zebrafish Stat3 within hematopoietic stem cells impacting multiple lineages throughout primitive and early definitive hematopoiesis, myeloid cell responses to G-CSF and LPS and neutrophil migration. Many of these roles showed conservation, but notably several involved noncanonical modalities, providing additional insights for relevant diseases.

Introduction

Mammals possess seven signal transducer and activator of transcription (STAT) proteins that are responsible for regulating a myriad of key cellular processes, such as differentiation, proliferation, survival, and functional activation [1]. STAT3 is considered an oncogene, being hyperactivated in a large variety of human cancers acting as a potent transcription factor within malignant cells [2, 3], as well as being involved in the regulation of antitumor immunity [4–6]. Somatic gain-of-function (GOF) STAT3 mutations have also been linked to the development of large granular lymphocytic (LGL) leukemias and multisystem primary immune regulatory syndromes, with most patients also showing signs of autoimmunity [7, 8]. Alternatively, germline loss-of-function (LOF) STAT3 mutations have been implicated in the development of autosomal dominant hyper-IgE syndrome (AD-HIES), which has similar symptoms to those caused by GOF mutations, with recurrent infections and autoimmunity [9, 10]. Various pharmacological inhibitors that inhibit STAT3 activation have proven efficacious in diseases associated with hyperactivation or GOF mutations [11]. However, the mechanisms underlying disease etiology in the context of LOF STAT3 mutations remain poorly understood, meaning current treatments primarily focus on the alleviation of symptoms rather than rectifying the underlying cause [12–14].

Global STAT3 knockout (KO) mice exhibited embryonic lethality, primarily due to disrupted leukemia inhibitory factor (LIF) signaling that impacted embryonic stem cells [15], which has limited the study of STAT3 to particular cell lineages using conditional mouse KOs [16]. Ablation of STAT3 in hematopoietic progenitor cells allowed survival to 4–6 weeks, but these mice ultimately succumbed to prolonged systemic inflammation reminiscent of Crohn’s disease [17]. Ablation of STAT3 in lymphocytes revealed specific roles in the differentiation of CD4⁺ T cells, particularly towards the Th17 subset, largely due to impacts on IL-6 and IL-21 signaling [18, 19]. Ablation of STAT3 directly impacted early B-cell development [20], while B-cell maturation was indirectly affected following STAT3 ablation in T follicular helper cells [21]. Myeloid-specific ablation of STAT3 identified a role in the polarization of macrophages in response to infection [22] and a more prominent role in neutrophils, particularly in response to emergency situations such as bacterial infection or injury [23–25]. These roles have been ascribed to the canonical mode of STAT function as inducible transcription factors downstream of cytokine and other receptors, although several noncanonical modes of STAT function have also been identified that require consideration [26].

Zebrafish STAT3 is highly conserved with its mammalian counterparts [27], and is expressed in similar lineages and activated downstream of conserved pathways suggesting significant conservation of function. However, in contrast to mice, zebrafish harboring Stat3 KO mutations have been shown to survive embryogenesis, allowing functional studies into the juvenile stage [28, 29]. This has been exploited in this study that investigates two different Stat3 mutants: a Stat3 KO mutant and a C-terminal truncation mutant impacting the transactivation domain (ΔTAD) that mimics human STAT3 LOF mutations [9, 30]. These have been thoroughly characterized with respect to primitive and early definitive hematopoiesis, as well as neutrophil and macrophage functionality. This has identified important contributions to immune cell development and function, with both canonical or non-canonical modes of action likely involved, which has implications for human disease.

Materials and Methods

Zebrafish Husbandry

Wild-type (WT) zebrafish and the transgenic lines Tg(mpx:GFP)ⁱ¹¹⁴ [31] and Tg(mpeg1.1:GFP)^gl22 [32] were maintained in accordance with standard husbandry practices in a purpose-built Techniplast aquarium on a 14/10 h light/dark cycle with thrice daily feeding. Embryos were maintained at 28.5°C in E3 water up to 24 h post fertilization (hpf) when this was replaced with 0.003% (w/v) 1-phenyl-2-thio-urea (PTU) in E3 water (E3/PTU) to inhibit pigmentation and promote transparency.

Genetic Manipulation and Line Generation

Single guide RNAs (sgRNAs) were designed against the zebrafish stat3 gene (Acc. No. NM_131479.1) using CHOPCHOP v3 [33] targeting the sequences encoding the N-terminal (sgRNA 1: 5′-GGT​GCT​GCT​TGA​TGC​GCC​GC) or SH2 (sgRNA 2: 5′-GGT​AGA​CCA​GCG​GCG​ACA​CT) domains. Relevant oligonucleotides were cloned into the pDR274 expression vector as previous described [34], and the sgRNAs transcribed using a MegaShortScript™ T7 transcription kit (Thermo-Fischer Scientific) and subsequently purified with a MegaClear™ kit (Thermo-Fischer Scientific). WT embryos at the 1 cell stage were injected with ∼1 nL of 1 × Danieau solution containing 12.5 ng/μL guide RNA (gRNA) and 200 ng/μL TrueCut Cas9 protein v2 (Thermo-Fischer Scientific), raised to adulthood and out-crossed with WT zebrafish. The F1 progeny were screened by high resolution melt analysis [35] and putative mutants confirmed by Sanger sequencing (Australia Genome Research Facility) of PCR products generated with primers flanking the respective sgRNA target sites (Table 1). Zebrafish carrying suitable mutant alleles were further out-crossed with WT zebrafish.

Table 1.

Primers used for PCR applications related to mutant screening and confirmation

Mutation	Application	Primers	Reference
stat3 KO	Genomic PCR/HRM	F: 5′-CCA​ACA​AGG​AGT​CTC​ACG​CC	This paper
		R: 5′-GGG​CTT​CTC​CAG​ATA​CTT​GC
stat3 KO	Genomic PCR/sequencing	F: 5′-CCA​ACA​AGG​AGT​CTC​ACG​CC	This paper
		R: 5′-CAG​TAG​ACG​CTG​CTC​TTC​CCA​C
stat3 KO	RT-PCR/sequencing	F: 5′-GTT​TCC​GTG​GAG​GAC​GAT​TTA	This paper
		R: 5′-AGT​GGG​ATG​GGC​AAC​CTG​G
stat3 ΔTAD	Genomic PCR/HRM	F: 5′-TGA​GAT​CAT​CAT​GGG​CTA​CAA	This paper
		R: 5′-TGT​CAG​GGA​ACT​CAG​TGT​CT
stat3 ΔTAD	Genomic PCR/sequencing	F: 5′-CCA​GAT​TCA​GTC​GGT​TG	This paper
		R: 5′-TGT​CAG​GGA​ACT​CAG​TGT​CT
stat3 ΔTAD	RT-PCR/sequencing	F: 5′-GAG​AGA​GAG​CAA​TCC​TGA​GTC​C	This paper
		R: 5′-CTA​ACT​CTG​GCA​ACA​CAA​GCG

HRM, high-resolution melt.

Experimental Setup

Experiments were performed in a blinded manner with respect to genotype. Embryos derived from heterozygous in-crosses of each mutant were pooled, with samples genotyped post-analysis. Each experiment was repeated three times. For embryos subjected to WISH, genomic DNA was extracted using a genomic lysis buffer (12.5 mm Tris [pH 8.3], 62.5 mm KCl, 1.875 mm MgCl₂, 0.0042% [v/v] Tween 20, 0.0042% [v/v] nonylphenoxy polyethoxy ethanol [NP40]), as previously described [36]. For all other embryos or samples, genotyping was performed using QuickExtract (Bioline), according to the manufacturer’s protocol.

Whole Mount in situ Hybridization

Zebrafish embryos were fixed with 4% (w/v) paraformaldehyde in PBS-T at 4°C prior to whole-mount in situ hybridization using DIG-labeled antisense probes, as previously described [37].

In vivo Analysis

Transgenic embryos at 3 dpf were injected into the venous return with 100 ng/μL lipopolysaccharide (LPS) (#00-4976-93, Invitrogen) in 1 × Danieau containing 1% (w/v) phenol red and incubated at 28.5°C in E3/PTU for 8 h, followed either by imaging or RNA extraction. Alternatively, 100 ng/µL mRNA encoding the zebrafish granulocyte colony-stimulating factor (G-CSF) paralogue, Csf3a, in 1 × Danieau buffer containing 1% (w/v) phenol red was injected into 1 cell stage embryos [38] with imaging or RNA extraction performed at 3 dpf. Neutrophil and macrophage migration was assessed using a wound healing assay [39]. Transgenic embryos at 3 dpf were anaesthetized with 5 μg/mL benzocaine in E3/PTU before the excision of their caudal fin, with the injury site imaged at 4 h intervals using fluorescence microscopy.

Gene Expression Analysis

Total RNA was extracted from juvenile zebrafish with TRIsure (Bioline), according to the manufacturers’ protocols. Single zebrafish embryos were euthanized and homogenized in RLT buffer (Qiagen) using a TissueLyser II (Qiagen) for subsequent RNA extractions with the RNeasy Mini Kit (Qiagen), according to the manufacturer’s protocols for extracting RNA from low sample volumes. For tail wounding assays, whole embryos were used since RNA yield from isolated tails was low and variable. Extracted RNA was subjected to quantitative real-time reverse-transcription PCR (qRT²-PCR) with gene-specific primers (Table 2). Data were normalized relative to actb and fold-change calculated using the ΔCT method [47]. Alternatively, RT-PCR was performed with different primes (Table 1), followed by gel electrophoresis and sequencing of isolated products.

Table 2.

Primers used for quantitative real-time reverse-transcription PCR/RT-PCR

Gene	Accession number	Primers	References
actb	NM131031	F: 5′-TGG​CAT​CAC​ACC​TTC​TAC	[40]
		R: 5′-AGA​CCA​TCA​CCA​GAG​TCC
cd4-1	NM_001135096	F: 5′-CTA​CCC​AGA​GAA​AAG​ATT​GAA​CG	[39]
		R: 5′-AGA​AAT​CTG​CTG​ATA​GAG​AGA​CG
cd8a	NM_001040049	F: 5′-ACT​CTT​CTT​CGG​AGA​GGT​GAC	[41]
		R: 5′-ACA​GGC​TTC​AGT​GTT​GTT​TGA​A
csf1a	NM_001114480.1	F: 5′-ACG​TCT​GTG​GAC​TGG​AAC​TG	[42]
		R: 5′-CTG​TTG​GAC​AAA​TGC​AGG​GG
csf1b	NM_001080076.1	F: 5′-GGA​TTT​GGG​TCG​GTG​AGC​TT	[42]
		R: 5′-TGG​AGA​GGG​GAA​CAC​ACA​GT
csf3a	NM_001145242.1	F: 5′-CTA​CTG​GAG​TCT​GGT​GAT​TGG​C	This paper
		R: 5′-CGT​TGT​GTC​TGC​TCC​TGT​TCC
csf3b	NM_001143754	F: 5′-GTG​TGC​AGC​GGA​TGC​TCA​T	[39]
		R: 5′-CTG​CGA​GGT​CGT​TCA​GTA​GGT​T
csf3r	NM_001113377.1	F: 5′-CAA​CCA​CAC​ACT​AAA​TAT​CAT​GCC​C	[41]
		R: 5′-GTG​ACC​TTC​TTG​CTG​CGA​GGC​GAC
cxcl8a	XM_009306855	F: 5′-CCA​GCT​GAA​CTG​AGC​TCC​TC	[43]
		R: 5′-GGA​GAT​CTG​TCT​GGA​CCC​CT
cxcl8b	NM_001327985.1	F: 5′-GCT​GGA​TCA​CAC​TGC​AGA​AA	[43]
		R: 5′-TCA​AGG​AAA​AGT​TTG​CAG​CA
cxcr1	–	F: 5′-TTC​AGT​TCG​GCT​GCA​CTA​TG	[44]
		R: 5′-GGA​GCA​ACT​GCA​GAA​ACC​TC
cxcr2	XM_017357899.2	F: 5′-TGA​CCT​GCT​TTT​TTC​CCT​CAC​T	[44]
		R: 5′-TGA​CCG​GCG​TGG​AGG​TA
epor	NM_001043334.1	F: 5′-GTC​AGA​TAC​GCT​GTG​GAG​GGA	This paper
		R: 5′-TCA​CAG​GAC​TGG​TCC​AAA​A
fcer1gI	XM_021478398.1	F: 5′-TGG​TGC​TGA​TGA​AGA​TCT​TC	This paper
		R: 5′-GGC​GGA​TTT​CAT​CTT​GAT​GC
hbaa1.1	NM_131257	F: 5′-GGG​AGG​TCT​TGA​GAG​AGC​C	[45]
		R: 5′-GCA​ATC​AGC​GAG​AAG​CCT​GA
ighd	XM_0.21474792.1	F: 5′-CGA​TTG​TGT​TGG​ATT​CAG​CCC	This paper
		R: 5′-GAC​CAA​CAA​CAA​GTC​CGA​G
ighm	AF281479	F: 5′-AAA​GAT​TTG​AGC​GAT​TTT​GTG​C	[39]
		R: 5′-GCT​AAA​CAC​ATG​AAG​GTT​GCT​G
ikzf1	NM_130986	F: 5′-AAG​CGA​AGT​CAC​ACT​GAA​GAA​AG	[39]
		R: 5′-CAG​ATG​TCC​AGT​GAG​AGC​GTC
il4	NM_001170740	F: 5′-CTG​AAT​GGG​AAA​GGG​GAA​A	[39]
		R: 5′-CGA​GAA​ACT​CCT​TCA​TTG​TGC
il6	NM_001261449	F: 5′-TCA​GAG​ACG​AGC​AGT​TTG​AGA​GAG	[39]
		R: 5′-GTT​AGA​CAT​CTT​TCC​GTG​CTG
il13	NM_001199905.1	F: 5′-GCC​TTC​AGA​CCA​CTA​ATC​TTT​TG	This paper
		R: 5′-ACT​TGG​TCT​TGG​GCT​TTT​TTG
il21	NM_001128574	F: 5′-GAA​CTC​AAG​AAG​ATT​CAG​CAG​G	[39]
		R: 5′-GTT​CGG​CTG​TTG​ACC​ATT​G
mmp9	NM_213123.1	F: 5′-ACT​CTC​CTT​TGA​GGA​CCA​CC	This paper
		R: 5′-GCT​TTC​GGT​GGA​GTG​CCC
mpeg1.1	NM_212737	F: 5′-GTT​ACA​GCA​CGG​GTT​CAA	[39]
		R: 5′-TGT​TTT​CAA​TGG​CGT​CAG​C
mpx	NM_001351837	F: 5′-CTG​CGG​GAC​CTT​ACT​AAT​GAT​GG	[46]
		R: 5′-CCT​GGA​TAT​GGT​CCA​AGG​TGT​C
nklc	NM_001311793.1	F: 5′-TGT​GCT​GCT​CAC​TTG​GAG​ATG​C	[39]
		R: 5′-CAT​AGT​CCA​GGC​AGT​TGT​TCC
scl	NM_213237	F: 5′-GCT​TTC​CCT​CTC​CCG​GCA​CG	[39]
		R: 5′-GTT​CGT​GAA​AAT​CCG​TCG​CA
socs3b	NM_213304.1	F: 5′-GGA​TGT​GGA​AGA​AGT​GTC​AGA​GCT	This paper
		R: 5′-GCC​GTT​TTC​ACA​CTG​AGC​GT
stat3	NM_131479.1	F: 5′-TGA​TGC​CTC​TCT​GAT​AGT​G	This paper
		R: 5′-GAG​TGC​GTC​TCA​AGG​TCG
tcra	AF425590	F: 5′-ACT​GAA​GTG​AAG​CCG​AAT	[39]
		R: 5′-CGT​TAG​CTC​ATC​CAC​GCT

Imaging and Image Analysis

Imaging was performed on an Olympus MVX10 microscope using either standard lighting or UV excitation coupled with a GFP filter for transgenic embryos. Images were captured with a DP74 camera using CellSens Dimension 1.6 software (Olympus). The relative area of staining or fluorescence was measured using ImageJ while specific cell populations were quantified by manual counting, including the number of cells in each quadrisect of the yolk relative to the total number of cells during development. Cell numbers for the wounding assay were manually determined using the images from each timepoint, as previously described [39]. Total cell numbers at each timepoint were normalized with respect to the total average cell population of the respective genotype. The rate of cell migration was calculated by dividing the net number of cells migrating to the site of injury by the time taken to reach the peak. Alternatively, the rate of egress was calculated using the decline of cells from the peak to 24 h divided by the time between these points.

Statistics

Statistical analyses were performed using Graph Pad Prism version 8. Normality of data was tested using the D’Agostino-Pearson omnibus normality test and unequal variances using the F test. The difference between more than two groups was tested for significance using ordinary one-way ANOVA coupled with Tukey’s multiple comparison test. In cases where there were multiple variables with more than two groups, a two-way ANOVA with Tukey’s multiple comparison test was performed. If data did not follow a normal distribution, the Mann-Whitney U test was performed. In cases where two groups required comparisons, a Student’s t test was performed with Welch’s correction if variances were unequal. Survival was visualized on a Kaplan-Meier graph, with the Gehan-Breslow-Wilcoxon test used to compare between different groups.

Results

Generation of stat3 Zebrafish Mutants

To produce relevant Stat3 mutants, two separate sgRNAs were designed to target the zebrafish stat3 gene: sgRNA 1 for exon 3, encoding the N-terminal domain, and sgRNA 2 for exon 20, encoding the C-terminal portion of the SH2 domain (Fig. 1a). The sgRNAs were injected separately with Cas9 protein into WT embryos, which were then raised to adulthood and out-crossed with WT fish. Screening of F1 progeny identified one allele from each sgRNA that impacted the corresponding Stat3 reading frame. Zebrafish carrying these alleles were out-crossed again, with heterozygous carrier progeny of each allele in-crossed to produce homozygous F3 mutants. Sequence analysis revealed that the sgRNA 1-generated allele (ko/mdu35) was a 4 bp deletion that resulted in a frameshift and premature stop codon, leading to a severely truncated Stat3 protein representing a loss-of-function KO mutation (Fig. 1b, d). The allele generated from sgRNA 2 (Δtad/mdu36) was an 8 bp deletion that also resulted in a frameshift mutation and early stop codon just after the last binding pocket of the SH2 domain causing the complete truncation of the TAD (ΔTAD) (Fig. 1c, d). Gel electrophoresis and sequence analysis of RT-PCR products confirmed the presence of the respective deletions in the transcribed mRNA, with no apparent change to splicing evident, and also revealed lower stat3 transcript levels in both mutants (online suppl. Fig. 1; for all online suppl. material, see https://doi.org/10.1159/000538364).

Fig. 1.

Generation of zebrafish Stat3 mutants. Schematic of the STAT3 protein outlining the various functional domains: N-terminal, CCD (coiled-coil domain), DBD (DNA binding domain), linker, SH2 (Src homology 2 domain), and TAD (transactivation domain), along with relevant sections of the zebrafish stat3 gene showing introns (black lines) and exons (colored boxes) encoding the indicated protein domains as well its targeting by sgRNAs (a). Sequence traces of representative homozygous wild-type (WT, stat3^wt/wt) and knockout (KO, stat3^ko/ko) (b) and WT and transactivation domain truncation (ΔTAD, stat3^Δtad/Δtad) (c) mutant fish, with nucleotides and corresponding amino acids shown. In each case, nucleotide sequences deleted are boxed in purple with de novo amino acids in red. Schematic of Stat3 proteins expressed in WT, KO, and ΔTAD fish (d).

To further elucidate the roles of Stat3 within the neutrophil and macrophage lineages both mutant lines were crossed onto the Tg(mpx:GFP) [48] and Tg(mpeg1.1:GFP) [32] transgenic backgrounds that mark neutrophils and macrophages, respectively. In subsequent experiments, mutants were compared to those encoding WT Stat3.

Stat3 Contributes to Primitive Hematopoiesis and Migration of Primitive Neutrophils

The effect of the two Stat3 mutations on primitive hematopoiesis was investigated using WISH with lineage-specific markers at relevant timepoints (12 hpf, 16 hpf, and 22 hpf). Compared to the Stat3 WT, KO embryos were found to have a significantly diminished population of early hematopoietic progenitor cells identified using a probe specific for the scl gene [49] at 12 hpf both caudally (Fig. 2a–d) and rostrally (Fig. 2e–h). A similar effect was seen with the early erythroid (Fig. 2i–l) and myeloid (Fig. 2m–p) populations stained with gata1a [50] and spi1b [40], respectively. This extended to mature erythroid cells stained with hbbe1.1 [51] (Fig. 2q–t), neutrophils stained with mpx [52] (Fig. 2u–x) and csf3r [38] (Fig. 2y–b′) and fluorescently marked macrophages in the Tg(mpeg1.1:GFP) line [32] (Fig. 2c′–f′). In contrast, ΔTAD mutants did not show any alteration in early hematopoietic progenitor cells (Fig. 2a–h), early erythroid cells (Fig. 2i–l), early myeloid (Fig. 2m–p) cells or mature erythroid cells (Fig. 2q–t). However, both neutrophil populations (Fig. 2u–x and Fig. 2y–b′) and macrophages (Fig. 2c′–f′) were significantly decreased in ΔTAD mutants compared to WT, although the neutrophil populations remained significantly higher than in KO mutants (Fig. 2u–b′).

Fig. 2.

Effect of Stat3 mutations on primitive hematopoiesis. Representative images of wild-type (WT), knockout (KO), and transactivation domain truncation (ΔTAD) mutant embryos subjected to WISH at 12 hpf with scl (a–c, e–g), at 16 hpf with gata1a (i–k) and spi1b (m–o) and at 22 hpf with hbbe1.1 (q–s), mpx (u–w), and csf3r (y–a′) using light microscopy or those on a Tg(mpeg1.1:GFP) transgenic background at 22 hpf (c′–e′) using fluorescence microscopy. Scale bars represent 200 μm. Quantification of area of staining for scl (d, h), gata1a (l), spi1b (p), and hbbe1.1 (t) and cell numbers for mpx (x), csf3r (b′), and mpeg1.1 (f′) showing values for individual embryos as well as mean and SEM with statistical significance between genotypes indicated (***p < 0.001, **p < 0.01, *p < 0.05, ns: not significant).

As part of normal development, primitive neutrophils and macrophages migrate across the yolk sac [38]. KO embryos exhibited reduced migration of neutrophils marked with mpx (Fig. 2u–w) and csf3r (Fig. 2y–a′) compared to WT embryos. Closer examination of KO embryos revealed the majority of mpx⁺ and csf3r⁺ cells had migrated less than one quarter of the way across the yolk compared to those in WT embryos that had migrated to a greater extent across the yolk (online suppl. Fig. 2A–H). In contrast, ΔTAD embryos resembled the WT (Fig. 2u–w; Fig. 2y–a′; online suppl. Fig. 2A–H). No impact on macrophage migration was evident in either mutant (Fig. 2c′–e′).

Stat3 Is Important in Innate Immune Cell Production during Definitive Hematopoiesis

The effect of the Stat3 mutations on definitive hematopoiesis was also investigated using WISH at 3–5 dpf. Hematopoietic stem cells stained with cmyb [53] were significantly diminished in Stat3 KO mutants both caudally (Fig. 3a–d) and rostrally (Fig. 3e–h). Similarly KO mutants showed a marked decrease in lymphoid precursor cells stained with ikzf1 [54] (Fig. 3i–l) as well as mature lymphoid cells stained with rag1 [55] (Fig. 3m–p) and tcra [56] (Fig. 3q–t) and mature erythrocytes stained with hbbe1.1 (Fig. 3u–x). Neutrophils stained with mpx (Fig. 3y–b′) or csf3r (Fig. 3c′–f′) as well as mpeg1.1⁺ macrophages (Fig. 3g′–j′) were all significantly reduced in KO mutants. In contrast, the ΔTAD mutants did not show any alterations in hematopoietic stem cells (Fig. 3a–h), lymphoid precursor cells (Fig. 3i–l), lymphoid cells (Fig. 3m–t), or mature erythroid cells (Fig. 3u–x). However, both neutrophil populations (Fig. 3y–b′; Fig. 3c′–f′) and macrophages (Fig. 3g′–j′) were all significantly reduced compared to the WT, although the macrophage population remained significantly higher than the KO mutants (Fig. 3g′–j′).

Fig. 3.

Effect of Stat3 mutations on early definitive hematopoiesis. Representative images of wild-type (WT), knockout (KO), and transactivation domain truncation (ΔTAD) mutant embryos subjected to WISH at 3.5 dpf with cmyb (a–c, e–g) and ikzf1 (i–k) and at 5 dpf with rag1 (m–o), tcra (q–s), hbbe1.1 (u–w), mpx (y–a′), and csf3r (c′–e′) using light microscopy or those on a Tg(mpeg1.1:GFP) transgenic background at 5 dpf (g′–i′) using fluorescence microscopy. Scale bars represent 200 μm. Quantification of area of staining for cmyb (d, h), ikzf1 (l), rag1 (P), tcra (t), and hbbe1.1 (x), and cell numbers for mpx (b′), csf3r (f′), and mpeg1.1 (j′) showing values for individual embryos as well as mean and SEM with statistical significance between genotypes indicated (***p < 0.001, **p < 0.01, *p < 0.05, ns: not significant).

Stat3 Plays a Role in the Innate Immune Response to Injury

To further explore the functional impacts of the Stat3 mutations on neutrophils and macrophages, embryos derived from the Stat3 mutants on either the Tg(mpx:GFP) or Tg(mpeg1.1:GFP) transgenic backgrounds were subjected to a caudal fin wounding assay at 3 dpf (Fig. 4a). The Stat3 KO embryos exhibited a marked decrease in the total number of neutrophils migrating to the injury site compared to WT embryos (Fig. 4b–c). However, when these data were normalized to the overall mpx⁺ cell number in the embryos, no significant difference in the proportion of neutrophils migrating was evident (Fig. 4d). Further analysis revealed a significant decrease in the migration rate (online suppl. Fig. 3C), but the cell size (online suppl. Fig. 3A) and egress rate (online suppl. Fig. 3E) were not impacted. No difference was observed in the number of total macrophages migrating (Fig. 4e–f) but normalization revealed that KO embryos had an increased proportion of macrophages migrating compared to WT embryos (Fig. 4g). Interestingly, the size of the macrophages was decreased in KO embryos (online suppl. Fig. 3B), but the rate of migration (online suppl. Fig. 3D) and egress (online suppl. Fig. 3F) were not affected. The ΔTAD mutant did not show any alteration in the number of neutrophils (Fig. 4e, f) or macrophages (Fig. 4b, c) recruited to the injury site. However, when normalized to overall cell counts the ΔTAD mutants showed a higher proportion of neutrophils migrating compared to both WT and KO mutant embryos (Fig. 4d) and of macrophages migrating compared to WT embryos (Fig. 4g). No change was observed in the size (online suppl. Fig. 3A), rate of migration (online suppl. Fig. 3C) or egress (online suppl. Fig. 3E) of neutrophils or the size (online suppl. Fig. 3B), rate of migration (online suppl. Fig. 3D) or egress (online suppl. Fig. 3F) of macrophages in ΔTAD mutants.

Fig. 4.

Effect of Stat3 mutations on the response of neutrophils and macrophages to injury. Wild-type (WT), knockout (KO), and transactivation domain truncation (ΔTAD) mutant embryos on either the Tg(mpx:GFP) or Tg(mpeg1.1:GFP) transgenic background were subjected to a caudal fin wounding assay at 3 dpf and imaged using fluorescence microscopy from 0 to 24 h post wounding (hpw) (a). Representative images of mpx⁺ neutrophil (b) and mpeg1.1⁺ macrophage (e) cells migrating to the vicinity of the injury site (dotted line) at the indicated times, with scale bars of 100 μm. Quantitation of total migrating mpx⁺ cells (c) or normalized relative to overall mpx⁺ cells (d), as well as total migrating mpeg1.1⁺ cells (f) or normalized relative to overall mpeg1.1⁺ cells (g) (n = 5). Gene expression analysis of the indicated genes in WT, KO, and ΔTAD embryos (n = 5–6) at 4 hpw presented as fold change (log₂) relative to WT (h). c, d, f–h show mean and SEM with statistical significance indicated compared to WT (***p < 0.001, **p < 0.01, *p < 0.05), between mutants (^@@@p < 0.001, ^@@p < 0.01, ^@p < 0.05) or between unwounded and wounded (^###p < 0.001, ^##p < 0.01, ^#p < 0.05), identified using a two-way ANOVA with Tukey’s multiple comparison test.

Gene expression analysis was performed at 4 h post wounding, which included genes for chemokines, cxcl8a and cxcl8b [43], and their cognate chemokine receptors, cxcr1 and cxcr2 [57], as well as cytokines, csf3a and csf3b [41], all implicated in neutrophil migration. Expression of cxcl8a, cxcl8b, and cxcr2 were significantly upregulated in WT embryos in response to wounding, while KO embryos demonstrated no such change with cxcl8a and cxcl8b expression significantly reduced compared to WT following wounding (Fig. 4h). Intriguingly, csf3b was significantly upregulated in WT but not in KO embryos while csf3a was upregulated in KO but not in WT embryos in response to wounding. ΔTAD embryos exhibited a significant upregulation of cxcr2, cxcl8b, and csf3a, while cxcr1, csf3b, and cxcl8a remained unchanged in response to wounding, with significantly lower expression of cxcl8a and csf3a compared to WT following wounding.

Stat3 Is Involved in the Response to LPS

To further elucidate the impact of the Stat3 mutations on immune responses, Stat3 mutants on the Tg(mpx:GFP) or Tg(mpeg1.1:GFP) background were subjected to injection with bacterial LPS as an inflammatory mediator. WT embryos showed a significant increase in neutrophils in response to LPS injection (Fig. 5a, d, m), but such an increase was not observed in KO embryos (Fig. 5b, e, m). In contrast, a significant increase in macrophage numbers was observed in both WT (Fig. 5g, j, n) and KO (Fig. 5h, k, n) embryos, although macrophage numbers in injected KO embryos were still significantly lower than their WT counterparts (Fig. 5n). The ΔTAD mutant displayed a similar phenotype to the KO, with lower basal numbers and no change in response to LPS for neutrophils (Fig. 5c, f, m), but a significant increase in macrophage numbers in response to LPS, albeit still lower than the WT (Fig. 5i, l, n).

Fig. 5.

Effect of Stat3 mutations on the response to LPS. Wild-type (WT: a, d, g, j), knockout (KO: b, e, h, k) and transactivation domain truncation (ΔTAD: c, f, i, l) mutant 3 dpf embryos on either the Tg(mpx:GFP) (a–f) or Tg(mpeg1.1:GFP) (g–l) transgenic background were left uninjected (a–c, g–i) or subjected to injection with lipopolysaccharide (LPS) (d–f, j–k) and imaged 8 h post injection showing representative images with scale bars of 100 μm. Quantitation of mpx⁺ cells (m) and mpeg1.1⁺ cells (n) showing values for individual embryos. Gene expression analysis of the indicated genes in LPS-injected WT, KO and ΔTAD embryos (n = 6) presented as fold-change (log₂) relative to uninjected (o). Show mean and SEM with statistical significance indicated between genotypes (m, n) or compared to WT (o) (***p < 0.001, *p < 0.05, ns: not significant), or between uninjected and injected (^##p < 0.01, ^#p < 0.05).

Expression analysis was performed in LPS-injected embryos on genes for an inflammatory cytokine, il6 [58], as well as macrophage-specific csf1a and csf1b [59] and neutrophil specific csf3a and csf3b [41] cytokines, in addition to a negative regulator of cytokine signaling, socs3b [39] (Fig. 5o). LPS-injected WT embryos exhibited an upregulation of il6, csf1a, and csf1b, while KO embryos showed an upregulation of il6 and csf1b only, with expression of socs3b significantly reduced in KO embryos and along with csf3a also compared to LPS-injected WT embryos. In contrast, LPS-injected ΔTAD embryos showed a significant increase in the expression of csf1b and a significant decrease in csf3a expression, which was also significantly decreased compared to LPS-injected WT embryos.

Stat3 Is Required for G-CSF-Mediated Emergency Granulopoiesis

The impact of Stat3 mutations on cytokine-mediated emergency granulopoiesis was assessed by injection of mRNA encoding the mammalian G-CSF paralogue, Csf3a, into embryos on the Tg(mpx:GFP) background. WT embryos showed a significant increase in neutrophil numbers (Fig. 6a, d, g), as described previously [38]. In contrast, KO embryos exhibited no change in neutrophil numbers in response to Csf3a injection (Fig. 6b, e, g). ΔTAD embryos, on the other hand, showed a significant increase in neutrophil numbers following Csf3a injection with numbers comparable to WT and significantly higher than Csf3a-injected KO embryos despite being comparable in uninjected embryos (Fig. 6c, f, g).

Fig. 6.

Effect of Stat3 mutations on the response to G-CSF. Wild-type (WT: a, d), knockout (KO: b, e) and transactivation domain truncation (ΔTAD: c, f) mutant 1 cell embryos on the Tg(mpx:GFP) transgenic background were left uninjected (a–c) or subjected to injection with mRNA encoding the zebrafish G-CSF paralogue, Csf3a, and imaged 3 days post injection showing representative images with scale bars of 100 μm. Quantitation of mpx⁺ cells (g) showing values for individual embryos. Gene expression analysis of the indicated genes in Csf3a-injected WT, KO and ΔTAD embryos (n = 6) presented as fold-change (log₂) relative to WT (h). g, h show mean and SEM with statistical significance indicated between genotypes (g) or compared to WT (h) (***p < 0.001, **p < 0.01, ns: not significant) or between uninjected and injected (^###p < 0.01, ^##p < 0.01, ^#p < 0.05).

Gene expression analysis of Csf3a-injected embryos was performed for the chemokine receptor gene cxcr2 and negative regulator socs3b, both shown to be induced by G-CSF through STAT3 [39, 60]. WT embryos exhibited a significant upregulation of cxcr2 and socs3b in response to Csf3a injection while in KO embryos cxcr2 was downregulated and socs3b remained unchanged, although both were significantly less than in Csf3a-injected WT embryos (Fig. 6h). ΔTAD embryos showed a similar response to WT embryos.

Stat3 Impacts Survival and Juvenile Immunity

Both Stat3 mutants experienced poor survival, with the KO mutants surviving less than 30 dpf, and the ΔTAD mutants less than 40 dpf, both significantly less than the robust survival of WT zebrafish (Fig. 7a), consistent with alternative Stat3 KO zebrafish [28, 29]. While this precluded the study of adults, analysis of 28 dpf juvenile was possible, which enabled analysis when the full immune repertoire, including B and NK cells, was present [56, 61]. Gene expression analysis of juvenile zebrafish included a hematopoietic stem cell marker, scl [49], an early lymphoid marker, ikzf1 [54], an early myeloid marker runx1 [62], neutrophil markers, mpx [48], and csf3r [38], a macrophage marker, mpeg1.1 [32], T-cell markers tcra, cd4-1, and cd8a [63], B-cell markers, ighd, ighm [64], and fcer1gl [65], erythrocyte markers, hbba1 and epor [66], and cytokines, il4 [67], il13 [68], il21, and il6 [69]. Compared to WT zebrafish, KO zebrafish demonstrated a significant increase in expression of the neutrophil marker mpx, the B-cell marker ighm, and the cytokine il13, while expression of il6 was significantly decreased (Fig. 7b). The ΔTAD mutants also showed a significant increase in mpx and il13, along with a decreased expression of il6. In contrast, the T cell marker cd4-1 was also increased in ΔTAD mutants compared to wildtype, while hbba1 was significantly increased compared to both WT and KO mutants (Fig. 7b), indicating a wider disruption of hematopoiesis.

Fig. 7.

Effect of Stat3 mutations on juvenile survival and immune cell development. Relative survival of wild-type (WT), knockout (KO), and transactivation domain truncation (ΔTAD) mutant zebrafish displayed as a Kaplan-Meier curve (n = 15–30) (a). Gene expression analysis of the indicated genes in 28 dpf WT, KO and ΔTAD juveniles (n = 6–12) showing mean and SEM (b). Statistical significance is indicated compared to WT (**p < 0.01, *p < 0.05) or between mutants (^@@p < 0.01, ^@p < 0.05).

Discussion

STAT3 represents a critical transcription factor activated downstream of a myriad of important cytokine and other receptors that impact a wide range of essential cell functions [10]. This is reflected in the various diseases associated with STAT3, including enhanced activation and somatic GOF mutations observed in a variety of cancers, as well as inflammatory disorders such as rheumatoid arthritis and inflammatory bowel disease [70, 71]. In addition, germline STAT3 mutations – presumed to be LOF – have been linked to the development of autoimmunity and impaired innate immunity leading to enhanced susceptibility to infection [10, 72]. Insight from mouse models has identified key roles for STAT3 in hematopoiesis and immunity, including neutrophil and macrophage functionality [16]. However, these studies have been typically limited to specific cell lineages, resulting in a need for additional models to study STAT3 function, including mutants that contribute to disease.

The zebrafish Stat3 protein is highly conserved with mammalian STAT3 proteins, with strong sequence homology across an identical domain structure, with the zebrafish stat3 gene located syntenically with its mammalian counterparts [73]. Zebrafish also shows distinct primitive and definitive waves of hematopoiesis like mammals [74], and possesses neutrophils and macrophages with equivalent functions [75]. Therefore, to further study the contribution of STAT3 to development and function, particularly of neutrophils and macrophages, CRISPR/Cas9 genome editing technology was utilized to create a Stat3 KO mutant that lacked all functional domains, and a Stat3 ΔTAD mutant in which only the transactivation domain and adjacent tyrosine that undergoes phosphorylation to mediate dimerization were deleted [73], similar to human LOF STAT3 [9, 10, 30]. Consistent with studies of alternative Stat3 KO zebrafish [28, 29], both the KO and ΔTAD mutants exhibit reduced survival and decreased expression of stat3 [76], the latter presumably a result of nonsense mediated decay. The survival defect precluded analysis of adults but did allow a thorough comparison of both Stat3 mutants with WT zebrafish during the primitive and early definitive waves of hematopoiesis, including the functional analysis of larval neutrophils and macrophages.

Analysis of the Stat3 KO mutants revealed for the first time that STAT3 ablation impacts primitive hematopoiesis, with a severe reduction across early hematopoietic progenitors to myeloid and erythroid progenitors and their mature counterparts. Early definitive hematopoiesis was similarly affected, with HSCs and each of the lymphoid, myeloid, and erythroid lineages decreased. This is consistent with Stat3 activity observed within early hematopoietic tissues in zebrafish [77] and suggests a role in the early progenitor/stem cell compartment that consequently impacts the development of downstream lineages. While primitive hematopoiesis has not been studied in Stat3 KO mice, hematopoietic-specific deletion of Stat3 in mice resulted in a defect in HSC maintenance during early definitive hematopoiesis [78–81], suggesting a conserved function in this respect.

Stat3 KO mutants also failed to elicit an increase in neutrophil numbers in response to the zebrafish G-CSF paralogue Csf3a. This was consistent with STAT3 being strongly activated downstream of G-CSFR in response to G-CSF [82, 83] and mouse models in which hematopoietic specific ablation of Stat3 blocked G-CSF-mediated granulopoiesis [24, 25], while loss of the STAT3 binding site on the G-CSFR decreased basal neutrophil levels [84]. Zebrafish Stat3 KO mutants also failed to increase neutrophils in response to immune stimulation with LPS, which mirrored the results observed during bacterial infection of mice in which Stat3 was ablated in hematopoietic cells [24], with LPS shown to drive granulopoiesis at least partially through induction of G-CSF [85]. Collectively, this indicates a conserved role for zebrafish Stat3 downstream of its G-CSFR homolog in mediating granulopoiesis.

Neutrophils from Stat3 KO embryos were additionally impaired in their ability to be recruited in response to injury, with the rate of migration significantly diminished, but with no change in other parameters. Impaired neutrophil migration from circulation has also been observed in mice with hematopoietic-specific Stat3 ablation [60], as well as in mice expressing a G-CSFR lacking the STAT3 binding site [84]. Such migration has been shown to be primarily driven by Cxcl8 (Il-8) that acts as the main neutrophil chemoattractant via Cxcr2, with Stat3 shown to be required for both G-CSF-mediated expression of Cxcr2 [25, 60] and NF-κb-mediated expression of Cxcl8 (Il-8) [86, 87]. Long range neutrophil migration in zebrafish has been similarly demonstrated to be driven through Cxcl8 paralogues, Cxcl8a and Cxcl8b, both of which signal through Cxcr2 [44, 88], with Cxcl8b/Cxcr2 mediating neutrophil migration out of the bloodstream [44] and Cxcl8a/Cxcr2 mediating neutrophil migration towards the site of injury [43]. In addition, both Csf3b and Csf3r have also been shown to facilitate migration of zebrafish neutrophils to injury sites [41, 89]. In this study, WT zebrafish embryos showed increased expression of cxcr2, cxcl8a and cxcl8b in response to wounding, but this was not evident in Stat3 KO embryos despite elevated csf3a expression. Furthermore, Csf3a injection was unable to induce expression of cxcr2 in Stat3 KO embryos in contrast to WT embryos. Together this suggests remarkable conservation of a G-CSF/G-CSFR/STAT3/CXCL8/CXCR2 pathway facilitating neutrophil migration. Stat3 KO mutants additionally exhibited defective neutrophil migration over the yolk during primitive hematopoiesis similar to that observed following ablation of zebrafish Csf3a [38, 41] or its receptor (Csf3r) [38, 89]. Whether this utilizes Cxcl8/Cxcr2 or another pathway remains to be determined.

In contrast, relative macrophage recruitment to sites of injury was higher in Stat3 KO compared to WT embryos with macrophages also found to be significantly smaller in size in the Stat3 KO embryos suggesting an alteration in macrophage biology consistent with mouse studies that have identified roles for STAT3 in macrophage polarization and homeostasis [22]. Stat3 KO embryos also exhibited increased macrophage numbers in response to LPS. This is likely driven through Csf1a/b [59] with signaling via the CSF1 receptor not involving strong STAT3 activation [90]. Mouse models have shown that STAT3 plays an anti-inflammatory role in macrophages in response to LPS, inhibiting the activities of IL-1β, TNFα, and IFNγ; as well as stimulating the expression of SOCS3 to provide negative feedback [91, 92]. Interestingly, in this study expression of the SOCS3 paralogue socs3b was found to be significantly diminished in response to LPS in Stat3 KO embryos, and we have recently shown that ablation of socs3b induced an inflammatory macrophage phenotype [39], consistent with a role for a STAT3/SOCS3 module in controlling macrophage activation.

Despite the significantly diminished neutrophil and macrophage populations observed in their embryonic stage, during their juvenile stage the Stat3 KO zebrafish displayed elevated expression of neutrophil and other markers, indicating a rebound in cell numbers, along with elevation of the il-13 gene, encoding a pro-inflammatory cytokine. The alternative Stat3 KO mutant produced in a previous study also demonstrated a similar upregulation of a suite of inflammatory cytokines and chemokines [29]. We have also recently shown that both Stat3 KO and Stat3 ΔTAD mutants exhibited comparable skeletal deformities to the alternative Stat3 KO mutants, further establishing them as loss of function models [76], with no signs of compensation. Due to the similarities between these mutants and the alternative Stat3 KO mutants, it can be speculated that in addition to their skeletal abnormalities these zebrafish suffer from an immune disorder that exacerbates their poor phenotype resulting in the early lethality [29, 76]. While the Stat3 KO and Stat3 ΔTAD mutants appear to possess distinct phenotypes during their early stages, during later development their phenotypes become very similar. These phenotypes mirror those observed in global STAT3 KO mice models as they possess decreased myeloid precursor cells and an increase in neutrophils resulting in excessive inflammation [15, 84, 93].

This study utilized two Stat3 models theorizing that the Stat3 KO mutation would remove both canonical and noncanonical modes of action, but the Stat3 ΔTAD would block canonical but retain at least some noncanonical actions. Interestingly, Stat3 ΔTAD mutants displayed a range of phenotypes that both overlapped with and contrasted to those of Stat3 KO mutants. Thus, unlike Stat3 KO mutants, hematopoietic progenitor/stem cells, erythroid, and lymphoid lineages were unaffected in Stat3 ΔTAD mutants during either primitive or early definitive hematopoiesis. Instead, Stat3 ΔTAD mutants exhibited a selective reduction in neutrophil and macrophage numbers during both primitive and definitive waves, but even this was not to the same extent as Stat3 KO mutants, while the migration of neutrophils during development was also unaffected in Stat3 ΔTAD mutants. In contrast, just like Stat3 KO mutants, Stat3 ΔTAD mutants exhibited no LPS-induced increase in neutrophil numbers. However, in response to Csf3a-injection Stat3 ΔTAD mutants showed an increase in neutrophil numbers compared to Stat3 KO embryos, consistent with elevated induction of cxcr2 and socs3b, as well as elevated relative neutrophil migration in response to wounding, correlating with significant upregulation of cxcl8b expression. Collectively, these results suggest that canonical STAT3 action is required for primitive and definitive myelopoiesis since these are impacted in both mutants. For granulopoiesis, the canonical Stat3 action is presumably mediated via Csf3r, as similar impacts were observed in zebrafish Csf3r KO mutants [89]. In addition, these activities may be negatively regulated by Socs3b (Fig. 8) since zebrafish Socs3b KO mutants displayed enhanced granulopoiesis [39], but this requires confirmation. This canonical pathway also appears to be utilized for LPS-induced granulopoiesis. The prominence of the G-CSFR/STAT3 pathway in granulopoiesis has also been demonstrated in various mice models, with similar phenotypes shown in the absence of G-CSFR and/or STAT3 [25, 84]. However, maintenance of hematopoietic progenitor/stem cells, G-CSF-mediated emergency granulopoiesis and neutrophil migration can instead be attributed to noncanonical modes of STAT3 action. Indeed, others have shown that the impacts of STAT3 on stem cell proliferation are mediated by noncanonical actions in the mitochondria [77, 94]. However, for emergency granulopoiesis and neutrophil migration the noncanonical actions appear to still rely on Csf3a/Csf3r but are independent of STAT3 tyrosine phosphorylation or TAD-mediated transcriptional activation (Fig. 8). With regard to macrophages, both canonical and noncanonical Stat3 actions appear to be involved in their regulation, although exactly which pathways they act via remains to be determined.

Fig. 8.

Model for the role of Stat3 in the development and function of neutrophils. Schematic representation of signaling downstream of Csf3r and Cxcr2. Activation of Csf3r by either Csf3a or Csf3b induces canonical activation of Stat3 to mediate primitive and definitive granulopoiesis and is negatively regulated by Socs3b. Non-canonical Stat3 stimulated by Csf3a induces transcription of cxcr2 to facilitate emergency granulopoiesis. Stimulation of Cxcr2 by Cxcl8b drives a positive feedback loop which increases transcription of cxcl8b to mediate neutrophil migration.

Disease causing STAT3 mutations have been characterized as either GOF or LOF, although it remains unclear which specific STAT3 modes of action are impacted in each scenario [10]. The most widely studied STAT3 associated disease is AD-HIES, which arises as a result of de novo germline heterozygous mutations [9, 10, 95]. While designated as LOF, many mutations fail to significantly affect STAT3 phosphorylation and/or nuclear translocation [9]. Moreover, the phenotypes of AD-HIES patients are much milder than those observed with mouse Stat3 KOs, further suggesting that not all functionality is lost [96]. The majority of AD-HIES patients suffer from eczema and recurrent Staphylococcus infections often resulting in frequent bouts of pneumonia [12], which is presumably due to impaired innate immunity, since neutrophils from AD-HIES patients have been shown to be less responsive to infections [97], and unresponsive to IL-10, resulting in increased inflammatory cytokine production [98]. More than half of AD-HIES patients also share similar skeletal and connective tissue deficiencies resulting in the development of distinct facial features [12]. A mouse transgenic model expressing an AD-HIES patient-derived mutation affecting the DNA binding domain resulted in increased susceptibility to infections [99]. The zebrafish Stat3 ΔTAD mutant has provided an opportunity to more deeply explore the innate immune deficits underpinning AD-HIES, revealing multiple deficits in neutrophil and macrophage responses. Heterozygous Stat3 ΔTAD mutants also displayed similar defects, albeit less pronounced (online suppl. Fig. 4), along with skeletal abnormalities [76], further highlighting their appropriateness as an animal model of AD-HIES. This provides a powerful avenue to test potential therapeutics able to alleviate the various phenotypes observed in AD-HIES patients.

Acknowledgments

The authors thank staff at the Deakin University Zebrafish Facility for their excellent work maintaining the animals and Nagendra Awasthi for experimental assistance.

Statement of Ethics

This study was approved by the Deakin University Animal Ethics Committee, under project numbers: G14/2022, G15/2022, G16/2022, G23/2019, G24/2019, G25/2019.

Conflict of Interest Statement

The authors declare no conflict of interest.

Funding Sources

The research was solely funded by Deakin University, including a postgraduate stipend to MLS.

Author Contributions

Mohamed Luban Sobah designed the research, performed the research, analyzed the data, and wrote the manuscript. Clifford Liongue designed the research and analyzed the data. Alister C. Ward designed the research, analyzed the data, and wrote the manuscript. All authors contributed to editorial changes in the manuscript and read and approved the final manuscript.

Funding Statement

The research was solely funded by Deakin University, including a postgraduate stipend to MLS.

Data Availability Statement

All data generated or analyzed during this study are included in this article or available upon request. Further inquiries can be directed to the corresponding author.