Dysregulation of NIPBL leads to impaired RUNX1 expression and haematopoietic defects

Abstract The transcription factor RUNX1, a pivotal regulator of HSCs and haematopoiesis, is a frequent target of chromosomal translocations, point mutations or altered gene/protein dosage. These modifications lead or contribute to the development of myelodysplasia, leukaemia or platelet disorders. A better understanding of how regulatory elements contribute to fine‐tune the RUNX1 expression in haematopoietic tissues could improve our knowledge of the mechanisms responsible for normal haematopoiesis and malignancy insurgence. The cohesin RAD21 was reported to be a regulator of RUNX1 expression in the human myeloid HL60 cell line and during primitive haematopoiesis in zebrafish. In our study, we demonstrate that another cohesin, NIPBL, exerts positive regulation of RUNX1 in three different contexts in which RUNX1 displays important functions: in megakaryocytes derived from healthy donors, in bone marrow samples obtained from adult patients with acute myeloid leukaemia and during zebrafish haematopoiesis. In this model, we demonstrate that alterations in the zebrafish orthologue nipblb reduce runx1 expression with consequent defects in its erythroid and myeloid targets such as gata1a and spi1b in an opposite way to rad21. Thus, also in the absence of RUNX1 translocation or mutations, additional factors such as defects in the expression of NIPBL might induce haematological diseases.


| INTRODUC TI ON
In all vertebrates, the RUNX family of transcriptional regulators containing the runt domain (RD) comprises three isoforms: RUNX1, RUNX2 and RUNX3 that, together with the non-DNAbinding CBFβ subunit, regulate many developmental processes. 1,2 The RUNX members specify their functions depending on their cellular and tissue expression: RUNX1 plays a key role in blood development, primarily in the haematopoietic stem cells (HSCs), RUNX2 is manly involved in bone morphogenesis, and RUNX3 in cell growth of neurons, epithelial cells and T cells. However, the three RUNX proteins could exert biological activities also in other organs [3][4][5] ; for example, RUNX2 and RUNX3 are known to play a role during haematopoiesis together with RUNX1. In addition, all the RUNX genes are transcribed by a distal and a proximal promoter (P1 and P2, respectively) in two main isoforms that differ in the 5′UTR and in the coding sequence of the first exon. 6,7 The as alternative splicing, acetylation, methylation, phosphorylation and ubiquitination. 9 As transcription factor, RUNX1 targets multiple genes, many of which are also pivotal transcriptional regulators involved in the formation of all haematopoietic lineages including the haematopoietic-specific member of E-twenty-six (ETS) family, PU.1. 10,11 Furthermore, the activity of RUNX1 is carried out by its interaction with different proteins fundamental during haematopoiesis such as GATA1, PU.1, CEBPA, PAX5 and ETS1. 10,[12][13][14] Given the high complexity in RUNX1 expression and function, its deregulation is commonly associated with haematopoietic diseases. Depletion of Runx1 in mice and zebrafish models leads to severe defects or complete absence of definitive haematopoiesis. [15][16][17][18] RUNX1 is frequently involved in chromosomal translocations observed in acute leukaemias, such as ETV6-RUNX1 in t(12;21) and RUNX1-EVI1 in t (3;21), 19 while the formation of the chimeric protein RUNX1-CBF2T1 (AML1-ETO) is associated with the M2 subtype of acute myeloid leukaemia (AML). 20,21 RUNX1 mutations determine the familial platelet disorder with a propensity for AML (AML/FPD) and the minimally differentiated acute myeloid leukaemia (AML/M0). 22 Importantly, regulation of RUNX1 dosage is essential for the maintenance of normal haematopoiesis 23 and several haematopoietic transcription factors are deputed to regulate RUNX1 expression such as Gata2, Ets factors (Fli-1, Elf-1 and Pu.1) and the SCL/Lmo2/Ldb1 complex. 24 In zebrafish, the subunit Rad21 of the cohesin complex has been identified as a regulator of runx1 through a forward genetic screen, 25 and multiple predicted and in vivo validated binding sites of Rad21 have been shown to be involved in the regulation of the zebrafish runx1. 26 In this work, we demonstrate that NIPBL, another member of the cohesin complex, positively regulates RUNX1 expression in two different contexts in which it exerts important functions: normal cord blood megakaryocytes derived from healthy donors and bone marrow samples derived from adult AML patients. In addition, we generate a zebrafish model in which the nipblb-mediated dysregulation of runx1 expression leads to haematopoietic defects resulting in decreased expression of the erythroid marker gata1a and reduction of mature circulating erythrocytes, and increased expression of myeloid precursors positive for the spi1b marker. Our data confirm the regulatory loop between RUNX1-GATA1 and PU.1 during haematopoiesis and highlight a new role of NIPBL on top of this route.

| Patients
Diagnostic bone marrow samples from 34 adult patients affected by AML were collected and characterized for specific molecular aberrancies, including translocations t(9;22), t(8;21) and inv (16)  Human material and derived data were used in accordance with the Declaration of Helsinki.

| Animals
Zebrafish embryos were raised and maintained according to in- Sigma-Aldrich).

| Reverse transcription and real-time quantitative polymerase chain reaction assays (RT-qPCR)
RNA was extracted from human and zebrafish embryos using TRIzol reagents (Life Technologies), following the manufacturer's protocol.   Table 3. Expression levels in the Y-axis were relative to the control.

| In situ hybridization, o-dianisidine and immunofluorescence analyses
Whole-mount in situ hybridization (WISH) experiments were car-

| Injections
Injections were carried out on one-to two-cell stage embryos.
Details of concentration and sequence of nipblb morpholino (nipblb-MO, Gene Tools, Oregon, US) and rad21-MO (Gene Tools) are described in Ref. 34

TA B L E 3 Zebrafish primer sequences used in qPCR experiments
standard deviation (SD) values refer to data from triplicate samples.
In zebrafish, at least three different experiments were done for each analysis.
The degree of linear relationship between RAD21, NIPBL, RUNX1, MPL and SPI1 expression levels was calculated using Spearman's correlation coefficient (r value). software, which allows import and effective integration of data obtained by different experimenters, experimental platforms and data sources. 36 In megakaryocytes (MK) derived from healthy donors, RAD21 expression did not correlate with the expression levels of RUNX1 ( Figure 1A). Conversely, we found a positive correlation between the expression of RUNX1 and that of NIPBL, another member of the cohesin complex ( Figure 1B). To explore the myeloid compartment under pathological condition, we used bone marrow (BM) cells derived from adult AML patients. Similar to TRAM analyses, when RAD21 and RUNX1 expressions were investigated in a cohort of 34 AML adult patients without anomalies in chromosome 21 that contains the RUNX1 locus, no significant correlation was reported ( Figure 1C). Conversely, we observed the positive NIPBL/RUNX1 correlation already detected in megakaryocytes ( Figure 1D).

| TRAM analysis
We previously showed that NIPBL transcript abundance is decreased in AML patients carrying the mutated NUCLEOPHOSMIN1 (NPM1), which transfers NPM1 in the cytoplasm (NPMc+), compared to the NPM1 wild-type (NPM1wt). 34

| Knock-down of nipblb specifically reduces runx1 expression in zebrafish
To confirm the positive correlation between NIPBL and RUNX1 observed in human, we took advantage of a zebrafish model with down-regulation of nipblb, the orthologue of the human NIPBL, previously generated in our laboratory. 34 The expression of runx1 was analysed in embryos at 30 and 48 hpf as definitive HSCs arise from the vascular endothelium from these developmental stages.
Moreover, we verified that both P1-P2runx1 isoforms were highly

| NIPBL-mediated RUNX1 down-regulation impairs the expression of RUNX1 target genes
We further verified whether the NIPBL-mediated RUNX1 reduction affects the expression of RUNX1 haematopoietic downstream targets. In MK cells derived from healthy donors, we observed a positive correlation between the expression of RUNX1 and that of MPL gene, the marker of megakaryocyte/platelet differentiation ( Figure 3A). 39 In BM cells derived from AML human patients, we showed a positive correlation between the expression of RUNX1 and its targets SPI1, the marker of myeloid precursors ( Figure 3B). 40 The expression of confirming the RT-qPCR data and our previous findings. 34 In agreement with the positive regulation exerted by runx1 on spi1b, the injection of the runx1/mRNA further enhanced this phenotype ( Figure 3J). 40 As it has been previously demonstrated that rad21, another member of the cohesin complex, regulates runx1 in zebrafish embryos during primitive haematopoiesis, 25    Also the SPI expression is positively regulated by RUNX1, facilitating the interaction between the SPI enhancer and its proximal promoter. 47 Indeed, we observed a positive correlation between RUNX1 and SPI1 in human samples and in zebrafish when we forced runx1 expression. However, following nipblb down-regulation, we also observed an increase in spi1b expression according to our previous data. 34 This result does not correlate with the runx1 reduction and its positive activity on spi1b expression and raises three possibilities: first that the increased number of myeloid precursors, previously reported in zebrafish following nipblb-MO injection, 34 leads to an augmented number of cells expressing spi1b with a consequent total increase of spi1b transcript. Second, it has been reported that the chromatin structure at the spi1b/PU.1 locus could be differentially regulated during the different stages of haematopoiesis, 11 suggesting the possibility that other mechanisms than RUNX1 might control spi1b expression. For example, we demonstrated that the canonical Wnt pathway, modulated by nipblb, has a pivotal role in regulating spi1b myeloid expression during definitive haematopoiesis in zebrafish. 34 Moreover, in vitro and in vivo studies demonstrated that forced expression of gata1 down-regulates spi1b, while forced expression of spi1b down-regulates gata1. [48][49][50] In this scenario, the nipblb-mediated runx1 down-regulation might lead to spi1b enforced expression that, in turn, reduces gata1a expression.
Alternatively, the two P1 and P2runx1 isoforms might exert different functions on spi1b regulation. Indeed, as for the case of Rad21 zebrafish mutants, 26 we demonstrated that the down-regulation of nipblb differently affects the two isoforms by significantly reducing only the P2runx1. Third, it has been demonstrated that NIPBL might regulate SPI1 by itself, encompassing the Runx1 regulation. 44 Although in this work we did not address the mechanism through which NIPBL regulates RUNX1 expression, we demonstrated that NIPBL positively regulates RUNX1 transcription and that the link between NIPBL dysregulation and RUNX1-driven haematopoietic defects might explain haematological malignancy occurrence. Thus, also in the absence of RUNX1 translocation or mutations, additional factors such as defects in the expression of NIPBL observed in AML patients might contribute to haematological diseases.

ACK N OWLED G M ENTS
The authors thank Carol Burns for the runx1 mRNA. They also thank Dorela Meta and Giulia Salmoiraghi for technical help in data preparation.

CO N FLI C T O F I NTE R E S T
The authors declare no competing financial interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.