Runx1 regulates zebrafish neutrophil maturation via synergistic interaction with c-Myb

Neutrophils play an essential role in the innate immune defense system in vertebrates. During hematopoiesis, the full function of neutrophils involves maturation of granules and related enzymes. Yet, transcription regulators that promote neutrophil maturation remain largely undefined. Here, two hematopoiesis-defective zebrafish mutants, runx1w84x and c-mybhkz3, were used to investigate the in vivo roles of Runx1 in cooperation with c-Myb in regulating neutrophil maturation. Loss of runx1 impairs primitive neutrophil development. Additional regulation of c-myb+/− and c-myb−/− induces a more severe phenotypes suggesting a synergistic genetic interaction between c-myb and runx1 in neutrophil maturation. Further studies revealed that the two transcription factors act cooperatively to control neutrophil maturation processes via transactivating a series of neutrophil maturation-related genes. These data reveal the in vivo roles of Runx1 in regulating primitive neutrophil maturation while also indicating a novel genetic and molecular orchestration of Runx1 and c-Myb in myeloid cell development. The study will provide new evidence on the regulation of neutrophil maturation during hematopoiesis.

Neutrophils play an essential role in the innate immune defense system in vertebrates. During hematopoiesis, the full function of neutrophils involves maturation of granules and related enzymes. Yet, transcription regulators that promote neutrophil maturation remain largely undefined. Here, two hematopoiesis-defective zebrafish mutants, runx1 w84x and cmyb hkz3 , were used to investigate the in vivo roles of Runx1 in cooperation with c-Myb in regulating neutrophil maturation. Loss of runx1 impairs primitive neutrophil development. Additional regulation of c-myb +/− and c-myb −/− induces a more severe phenotypes suggesting a synergistic genetic interaction between c-myb and runx1 in neutrophil maturation. Further studies revealed that the two transcription factors act cooperatively to control neutrophil maturation processes via transactivating a series of neutrophil maturation-related genes. These data reveal the in vivo roles of Runx1 in regulating primitive neutrophil maturation while also indicating a novel genetic and molecular orchestration of Runx1 and c-Myb in myeloid cell development. The study will provide new evidence on the regulation of neutrophil maturation during hematopoiesis.
Neutrophils are the most abundant phagocytes essential for the first line of defense in the innate immune system. Mature neutrophils play an important role in pathogen clearance, response to tissue injury, and in mediating the inflammatory response (1). Dysregulation of granulopoiesis can cause neutropenia, myeloid leukemia, or neutrophil function disorders (2)(3)(4)(5)(6)(7). Thus, understanding how neutrophils differentiate into functional mature cells might facilitate the development of new therapeutic strategies for the treatment of neutrophilrelated disorders.
The function of mature neutrophils requires the development of characteristic neutrophil granules along with stored enzymes. These neutrophil-specific elements are formed at different stages during neutrophil maturation.
Mammalian neutrophils contain four types of granules: azurophil granules, specific granules, gelatinase granules, and secretory granules. Each subtype of granule contains highly specific storage proteins that carry out different immune functions (8). Neutrophil granule subtypes are released in order to lyse and eradicate microbes when neutrophils are activated during infections (2). Digestive enzymes are key components in neutrophil granules. For example, lysozyme C (Lyz), is a key bactericidal enzyme found in all types of neutrophil granules, and myeloperoxidase (Mpx) is an abundant peroxidase stored in neutrophil azurophilic granules (2). Sorting and packing neutrophil granule proteins by proteoglycans, such as serglycin (Srgn), are essential for neutrophil differentiation (9,10). Neutrophil maturation requires that these granule-related proteins are properly produced, yet the molecular basis controlling the process remains largely unknown.
Several hematopoietic-specific transcription factors are reported to control neutrophil granule-related protein expression. RUNX1 has been described as a pivotal transcription factor during definitive hematopoiesis (11)(12)(13). In neutrophil development, Runx1 is reported to promote granulocytic over monocytic lineage fate choice in zebrafish (14). Yet, the function of RUNX1 in neutrophil differentiation and maturation is still debatable. It has been reported that RUNX1 could regulate Mpx and Elane transcription in myeloid cell lines (15)(16)(17). Similarly, recent mouse data demonstrated that RUNX1-haploinsufficient hematopoietic progenitors impaired in vitro differentiation in neutrophils by repressing Cebpe expression (18). However, another study found that conditional ablation of the Runx1 gene in adult mice paradoxically expands myeloid pools to an extent without incurring any discernible differentiation blockage (19). Therefore, whether RUNX1 plays roles in granulocyte differentiation and maturation in vivo is still unclear, especially in early developmental stages. We previously showed that lyz is a direct target of c-Myb in regulating neutrophil maturation (20). Interestingly, lyz is also a transcriptional target of Runx1 (21); however, whether RUNX1 participates in neutrophil differentiation This article contains supporting information. ‡ These authors contribute equally to this work. * For correspondence: Yiyue Zhang, mczhangyy@scut.edu.cn. and maturation lacks sufficient in vivo evidence. Furthermore, whether the neutrophil maturation process is achieved by the orchestration of these two transcription factors requires genetic verification.
Here, we used two hematopoietic-defective zebrafish mutants, runx1 w84x (22) and c-myb hkz3 (23) to determine the role of Runx1 during neutrophil maturation. These mutants were used to elucidate the genetic interaction of the two transcription factors through genetic epistasis and biochemical analysis. It was found that Runx1 cooperates with c-Myb to control neutrophil maturation in zebrafish embryonic myelopoiesis. This study elucidates the genetic networks that orchestrate primitive myeloid cell development, improving our understanding of the pathogenesis of neutrophil-related diseases.

Runx1 regulates primitive neutrophil maturation
Mature neutrophils are characterized by abundant granules in the cytoplasm, which can be specifically stained by Sudan Black B (SB) (14,24). Similar to the phenotype of c-myb −/− mutants in primitive myelopoiesis (20), runx1 −/− mutants had reduced the number of SB + neutrophils at 36 h postfertilization (hpf) (14) (Fig. 1, A-C). More importantly, the signal intensity for SB staining of the neutrophils was lower in mutants compared with that in siblings ( Fig. 1, A, B and D), suggesting that Runx1 is involved in neutrophil maturation. Since mature neutrophils have abundant granules in their cytosol, videoenhanced differential interference contrast (VE DIC) (25) microscopy was used to observe neutrophil granule status and abundance in live embryos at 2 days postfertilization (dpf) in Tg(mpx:GFP) background. GFP was expressed mainly in neutrophils under the control of the neutrophil-specific promoter (26). Although mpx:GFP + cell numbers were decreased and the expression was weakened in the mutants, we could still catch the remaining mpx:GFP + neutrophils under the microscopy. Indeed, runx1 −/− mutants contained fewer mature neutrophils than observed in siblings ( Fig. 1, E, F, E', F' and G-I). Mature neutrophils were also characterized by banded and segmented nuclei (20,27). Therefore, mature and immature neutrophils were quantified by staining nuclei with May-Grünwald-Giemsa stain. These studies showed a significant decrease in mature neutrophils in 2-dpf runx1 −/− mutants compared with siblings ( Fig. 1, J and K). The data demonstrate that primitive neutrophils were affected by the runx1 mutation in zebrafish.

Runx1 regulates neutrophil maturation accompanying c-Myb
To determine the role of runx1 in relation with c-myb in neutrophil maturation, a genetic approach was used to compare the neutrophil phenotypes of single runx1 or c-myb mutants with composite mutants derived from crossing cmyb +/− ;runx1 +/− double heterozygotes. Since c-myb +/− heterozygous embryos were indistinguishable from wild-type embryos in the number of SB + cells at 36 hpf (20), we focused on the phenotype comparisons of c-myb +/− ;runx1 −/− double mutants and c-myb +/− and runx1 −/− single mutants to determine whether additional c-myb +/− heterozygotes would yield more severe phenotypes. When one allele of c-myb +/− was introduced into runx1 −/− , the number of SB + cells was further reduced (Fig. 2, A-E). Consistently, the c-myb −/ − ;runx1 −/− double homozygotes had nearly no SB + cells left (data not shown). Furthermore, the DIC microscopy of live embryos (Fig. 2, F-I, F'-I' and J-K) showed that the residual neutrophils in c-myb +/− ;runx1 −/− (Fig. 2, I and I') elicited more immature characteristics, with scarce granules almost lacking mpx:GFP expression when compared with c-myb +/− ;runx1 +/+ (Fig. 2, G and G') or c-myb +/+ ;runx1 −/− mutants (Fig. 2, H and H'). To determine whether the synergistic regulation by c-Myb and Runx1 has biological consequences on neutrophil function, we examined the rate of bacterial killing by neutrophils in each group. As expected, additive c-myb +/− further attenuated the runx1 −/− bacterial killing activity, as c-myb +/− ;runx1 −/− mutants prolonged the bacterial clearance time compared with runx1 −/− single mutants (Fig. 2L). The above data suggest that runx1 could genetically cooperate with c-myb to potentiate neutrophil maturation.
These results indicate that one c-myb allele deficiency (c-myb +/− heterozygous) does not affect lyz + signals when the runx1 gene is not mutated. In c-myb +/+ ;runx1 −/− embryos (Fig. 3C), the runx1 −/− homozygous mutation decreased lyz + cells to onethird of those in c-myb +/+ ;runx1 +/+ embryos (Fig. 3A). Importantly, the inadequate lyz + cells in runx1 −/− background mutants were further reduced when one allele of c-myb +/− was introduced, as the average lyz + cell number per embryo was decreased from 30 in c-myb +/+ ;runx1 −/− (Fig. 3C) to 10 in c-myb +/− ;runx1 −/− (Fig. 3, D and E). The above data suggest a synergistic regulation of the two transcription factors on the expression of lyz. Since neutrophil maturation requires proper sequential expression of digestive enzymes, it was hypothesized that the synergistic effects of c-myb and runx1 are likely for lyz, as well as for other related genes. Zebrafish Mpx and Npsn are both neutrophil-specific granzymes used for host clearance of microbe infections (28,29). Similarly, both mpx + and npsn + cells were further reduced when one or two alleles of the c-myb mutation were introduced in runx1 −/− mutants (Fig. 3, F-O, and Fig. S1, B, C, B' and C'). Moreover, the granule-formationrelated glycoprotein-srgn was greatly downregulated in c-myb zebrafish mutants (20), suggesting srgn is also regulated by c-Myb. Expectedly, in c-myb +/− ;runx1 −/− double mutants, srgn + cells showed a greater inhibition compared with c-myb +/− or The above data suggest that there is a synergistic regulation of c-myb and runx1 on downstream neutrophil maturation-related genes. Taken together, these results indicate that Runx1, cooperatively with c-Myb, is essential for neutrophil maturation by positively regulating neutrophil maturation-related genes. Regulation of neutrophil maturation c-Myb and Runx1 physically interact to directly promote neutrophil-specific genes transcription To determine whether the four genes (lyz, mpx, npsn, and srgn) are directly targeted by c-Myb and Runx1, the regulatory regions of these genes were analyzed. It was found that these genes have putative c-Myb and Runx1 binding sites (Fig. 4A). Chromatin immunoprecipitation polymerase chain reaction (ChIP-PCR) assays were performed to ascertain whether the two transcription factors can directly bind to the putative sites in vivo. To address this question, embryos were injected with Myc-tagged c-myb or runx1 plasmids to overexpress c-Myb and Runx1 for ChIP-PCR analysis in zebrafish. The results showed that all putative binding sites of the target genes were coprecipitated using an anti-MYC antibody (Fig. 4B). This  result demonstrates that these neutrophil-specific genes are all directly regulated by c-Myb and Runx1 respectively.
To further support the hypothesis, the in vivo GFP reporter assay was performed by expressing neutrophil-specific reporter constructs together with c-myb or runx1 overexpression plasmids in zebrafish embryos. As reported, by using the zebrafish reporter assay c-Myb overexpression increased lyz promoter transcription activity (20). In the present report, it was also found that c-Myb overexpression increased mpx, npsn, and srgn promoter transcription activity (Fig. 4C). Compared with overexpressing c-Myb only, transcription activation of the four target genes were all enhanced by Runx1 coexpression (Fig. 4C). This suggests a cooperation of c-Myb and Runx1 on targeting downstream genes.
Next, it was asked whether c-Myb and Runx1 physically interact in these molecular functions. Coimmunoprecipitation (Co-IP) experiments were performed in 293T cells cotransfected with runx1-Flag and Myc-c-myb constructs. Cell lysates were immunoprecipitated using an anti-FLAG antibody, followed by western blot analysis using anti-MYC and anti-FLAG antibodies. Immunoprecipitating Runx1-FLAG resulted in coprecipitation of MYC-c-Myb (Fig. 4D). This suggests protein-level cooperation between zebrafish Runx1 and c-Myb. In addition, the co-IP assay was performed in zebrafish embryos. The result was consistent with that observed in the 293T cells (Fig. S2). To identify the regions of c-Myb and Runx1 that are necessary for their interaction, zebrafish c-Myb and Runx1 proteins were further truncated for the co-IP assay. The results showed that the DNA binding domain (DBD) of c-Myb and Runt homologs domain (RHD) of Runx1 are predominantly responsible for the observed interaction (Fig. 4, E and F). However, the constructs for overexpressing truncated c-Myb or Runx1 proteins were injected into c-myb hkz3 and runx1 w84x mutants respectively, but the overexpression could not rescue the neutrophil defects in these two mutants (data not shown). These data suggest that the truncated proteins may also affect the essential functions of these proteins, such as DNA binding activity.
The final experiment was designed to explore whether there exists a mutual dependency on c-myb and runx1 expression in early neutrophil progenitors. Expression levels were checked at 18 hpf, a crucial stage in embryogenesis when zebrafish myelopoiesis arises directly from the rostral blood island (RBI) (24,30). In the RBI region, c-myb expression was found to be unchanged in runx1 −/− embryos (Fig. S3, A and B), and runx1 expression was also unaltered in c-myb −/− embryos (Fig. S3, C and D). These expression data suggest that c-myb and runx1 are transcriptionally independent of each other during early neutrophil development.
From the genetic epistasis and biochemical analysis, it can be concluded that Runx1 functions as a positive regulator for neutrophil maturation in early development. In addition, it interacts and cooperates with c-Myb to transactivate a panel of neutrophil maturation-related genes (Fig. 5).

Discussion
In this study, by utilizing the zebrafish model, the in vivo roles of Runx1 were elucidated in neutrophil maturation. The results demonstrated a cooperative genetic and molecular interaction between Runx1 and c-Myb in regulating neutrophil development. These data give crucial insights into the genetic and molecular orchestration involved in neutrophil maturation.
In hematopoiesis, RUNX1 is known to function in the formation of hematopoietic stem cells (12,31), the fate of macrophages/neutrophils (14), and the maturation of megakaryocytes and lymphocytes (19,32,33). Moreover, RUNX1 is one of the most frequently mutated genes in a variety of hematological malignancies, such as acute myeloid leukemia and familial platelet disorder (34,35). The high incidence of RUNX1 mutations in multiple types of hematologic malignancies provides strong evidence for its essential function in blood lineage development. At present, more than 50 chromosome translocations that affect RUNX1 function have been identified, most of which cause maturation arrest of myeloid cells and even leukemia (36). Recent studies have shown that RUNX1 is part of a transcriptional complex that regulates important target genes in myelopoiesis. The key questions that remain to be answered are precisely what mechanisms and target genes underlie myelopoiesis and leukemogenesis and how they can be used for developing potential targeted therapies for these patients. From previous mouse studies, the precise role of RUNX1 in primitive myelopoiesis is still unknown. The present study provides direct evidence of a novel function of Runx1 and its regulatory mechanism in the primitive neutrophil maturation process. The clarification of in vivo roles of Runx1 in early myeloid cell development will shed new light on a better understanding of RUNX1-related hematological diseases.
Neutrophil maturation requires a set of neutrophil granulerelated products that are gradually produced while granules are properly assembled. The lack of granule contents will lead to immature or dysfunctional neutrophils and ultimately immunodeficiency. In c-myb and runx1 mutants, neutrophil granule formation was severely blocked. In addition, granzyme genes were inhibited as well as an observed downregulation of the granule sorting and packing-related proteoglycan gene, srgn. These results underscored the interplay of c-Myb and Runx1 cooperatively controlling neutrophil development by transactivating a cluster of neutrophil maturation-related genes. It was also reported that the ETS family, C/ebp family, and Runx family of transcription factors all function in forming comprehensive genetic networks in myelopoiesis (37)(38)(39)(40). Thus, how myeloid-lineage-specific genes are regulated by different transcription factors is of great interest for future studies. In addition, precisely why these transcription factor pairs function in neutrophils remains to be determined. It is possible that during neutrophil development, transcription factors might be sequentially expressed in different subpopulations of developing neutrophils and/or they might be activated by the same factor. It will be important to derive a comprehensive target gene list of c-Myb and Runx1. In addition, it is crucial to find upstream factors and interacting proteins to elucidate the comprehensive regulation of the neutrophil maturation process.

Regulation of neutrophil maturation
Taken together, this study has demonstrated the in vivo role of Runx1 in neutrophil maturation during early myelopoiesis. Furthermore, a genetic interaction between the two transcription factors, Runx1 and c-Myb, was shown to regulate neutrophil maturation through a molecular interaction that functions to regulate genes expression in a cooperative manner. This study improved our understanding of the genetic networks that orchestrate primitive myeloid cell development and revealed the molecular basis of neutrophil-related disease pathogenesis.

Zebrafish (Danio rerio) strains
Zebrafish runx1 (Gene ID: 58126) and c-myb (Gene ID: 30519) were used in the study. The following strains were utilized: AB, Tg(mpx:GFP)i114 (26), runx1 w84x (22), and c-myb hkz3 (23). runx1 w84x mutant harbors a G to A nucleotide substitution resulting a premature truncation in the Runt domain of the Runx1 protein. This truncation removes most of the residues important in Runx1 activity, such as CBFb and DNA binding, and nuclear localization signal. c-myb hkz3 mutant harbors a splice mutation that results in the synthesis of a truncated c-Myb protein lacking its transactivation domain. Zebrafish were maintained in accordance with the Guidelines from the Animal Care and Use Committee of South China University of Technology. All experimental protocols were approved by the Division of Cell, Developmental and Integrative Biology, School of Medicine, South China University of Technology.

May-Grünwald-Giemsa staining of embryonic blood cells
Fish embryos were anesthetized in 90% PBS + 10% FBS containing 0.02% tricaine. After tail clipping using syringe needle, blood cells were collected by pipetting and cytospun onto slides by centrifugation at 450 rpm for 3 min using a Cytospin 4 (Thermo Scientific). The slides were then airdried and subjected to May-Grünwald-Giemsa (Merk) staining according to the standard protocol. Each group was collected from 150 embryos.
Histology and heat-shock-inducible experiment WISH and SB staining were performed at 36 hpf onward according to previous report (24,41), as primitive neutrophils were evident for detection. Relative ratio of SB signal

Regulation of neutrophil maturation
intensity was calculated by Image J software. Heat-shockinducible constructs (hsp70:Myc-c-myb and hsp70: Myc-runx1) were generated by inserting Myc-c-myb and Myc-runx1 into the pTol vectors under the control of the heat shock protein 70 (hsp70) promoter. Embryos injected with plasmids were incubated at 39.5 C for 1 h heat shock treatment at 24 hpf.

Cell transfection, immunoprecipitation, and western blot
293T cells were grown in DMEM supplemented with 10% bovine calf serum. Cell transfection, cell extracts preparation, immunoprecipitation, and western Blot have been described (42). Constructs with Myc-tagged zebrafish c-myb and Flagtagged zebrafish runx1 were transfected into 293T cells. Cell lysates were immunoprecipitated with anti-FLAG antibody, and the immunoprecipitants were examined by western blot using anti-MYC and anti-FLAG antibodies. Input represents 10% of total cell lysates used for immunoprecipitation. Anti-MYC and anti-FLAG antibodies were obtained from Santa Cruz (sc-40) and Sigma (ab6658), respectively.

Microscopy and imaging
VE DIC microscopy was done with 60×/1.00 NA waterimmersion objective mounted on Nikon 90i microscope according to previously described (25). In 2-dpf live embryos, the neutrophil granule status is easy to be detected with DIC assay since the neutrophil granules are abundant and neutrophils are accessible to be observed by the lens.

ChIP assay
Putative transcriptional binding sites were identified by JASPAR online software. Embryos were injected with Mycc-myb or Myc-runx1 plasmids at one-cell stage. Six hundred injected embryos were harvested at 2 dpf for brief fixation. Cross-linked chromatin was immunoprecipitated with anti-Myc antibody or anti-IgG antibody (negative control) according to the procedure described by Hart et al. (43). The immunoprecipitants were subjected to semiquantitative PCR. ChIP primers could be found in Table S1.

Bacterial killing experiment
The Tg(mpx:GFP) background zebrafish were anaesthetized at 2 dpf with tricaine (MS-222; Sigma-Aldrich) and injected in the tail muscle with approximately 200 c. f. u. dsRed-labled E. coli. Embryos were anaesthetized with tricaine and mounted in 1% (w/v) low-melting-point agarose for time-lapse microscopy. Bacterial clearance time in mpx:GFP + neutrophils was counted.

Statistical analysis
Data were analyzed by GraphPad Prism6 software using the Student t-test for comparisons between two groups and one-way analysis of variance (ANOVA; with Tukey posttest adjustment) among multiple groups. Significance was accepted when p < 0.05. Data are expressed as mean ±standard error of the mean (SEM). An asterisk indicates a statistical difference (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001).

Availability of data and materials
All data generated or analyzed during this study are included in this published article [and its supplementary information files].
Acknowledgments-We thank Koichi Kawakami for providing pTol vector, Tingxi Liu for sharing Tg(mpx:GFP)i114 constructs, and Paul P. Liu for sharing runx1 w84x mutants.  Conflict of interest-The authors declare no competing or financial interests.