BCAS2 promotes primitive hematopoiesis by sequestering β-catenin within the nucleus

Reviewer #1 (Public Review):

Summary:

In this manuscript, Ning et al. reported that Bcas2 played an indispensable role in zebrafish primitive hematopoiesis via sequestering β-catenin in the nucleus. The authors showed that loss of Bcas2 caused primitive hematopoietic defects in zebrafish. They unraveled that Bcas2 deficiency promoted β-catenin nuclear export via a CRM1-dependent manner in vivo and in vitro. They further validated that BCAS2 directly interacted with β-catenin in the nucleus and enhanced β-catenin accumulation through its CC domains. They unveil a novel insight into Bcas2, which is critical for zebrafish primitive hematopoiesis via regulating nuclear β-catenin stabilization rather than its canonical pre-mRNA splicing functions. Overall, the study is impressive and well-performed, although there are also some issues to address.

Strengths:

The study unveils a novel function of Bcas2, which is critical for zebrafish primitive hematopoiesis by sequestering β-catenin. The authors validated the results in vivo and in vitro. Most of the figures are clear and convincing. This study nicely complements the function of Bcas2 in primitive hematopoiesis.

Weaknesses:

A portion of the figures were over-exposed.

Reviewer #2 (Public Review):

Summary:

Ning and colleagues present studies supporting a role for breast carcinoma amplified sequence 2 (Bcas2) in positively regulating primitive wave hematopoiesis through amplification of beta-catenin-dependent (canonical) Wnt signaling. The authors present compelling evidence that zebrafish bcas2 is expressed at the right time and place to be involved in primitive hematopoiesis, that there are primitive hematopoietic defects in hetero- and homozygous mutant and knockdown embryos, that Bcas2 mechanistically positively regulates canonical Wnt signaling, and that Bcas2 is required for nuclear retention of B-cat through physical interaction involving armadillo repeats 9-12 of B-cat and the coiled-coil domains of Bcas2. Overall, the data and writing are clean, clear, and compelling. This study is a first-rate analysis of a strong phenotype with highly supportive mechanistic data. The findings shed light on the controversial question of whether, when, and how canonical Wnt signaling may be involved in hematopoietic development. We detail some minor concerns and questions below, which if answered, we believe would strengthen the overall story and resolve some puzzling features of the phenotype. Notwithstanding these minor concerns, we believe this is an exceptionally well-executed and interesting manuscript.

Strengths:

(1) The study features clear and compelling phenotypes and results.

(2) The manuscript narrative exposition and writing are clear and compelling.

(3) The authors have attended to important technical nuances sometimes overlooked, for example, focusing on different pools of cytosolic or nuclear b-catenin.

(4) The study sheds light on a controversial subject: regulation of hematopoietic development by canonical Wnt signaling and presents clear evidence of a role.

(5) The authors present evidence of phylogenetic conservation of the pathway.

Weaknesses:

(1) The authors present compelling data that Bcas2 regulates nuclear retention of B-cat through physical association involving binding between the Bcas2 CC domains and B-cat arm repeats 9-12. Transcriptional activation of Wnt target genes by B-cat requires physical association between B-cat and Tcf/Lef family DNA binding factors involving key interactions in Arm repeats 2-9 (Graham et al., Cell 2000). Mutually exclusive binding by B-cat regulatory factors, such as ICAT that prevent Tcf-binding is a documented mechanism (e.g. Graham et al., Mol Cell 2002). It would appear - based on the arm repeat usage by Bcas2 (repeats 9-12)-that Bcas2 and Tcf binding might not be mutually exclusive, which would support their model that Bcas2 physical association with B-cat to retain it in the nucleus would be compatible with co-activation of genes by allowing association with Tcf. It might be nice to attempt a three-way co-IP of these factors showing that B-cat can still bind Tcf in the presence of Bcas2, or at least speculate on the plausibility of the three-way interaction.

(2) A major way that canonical Wnt signaling regulates hematopoietic development is through regulation of the LPM hematopoietic competence territories by activating expression of cdx1a, cdx4, and their downstream targets hoxb5a and hoxa9a (Davidson et al., Nature 2003; Davidson et al., Dev Biol 2006; Pilon et al., Dev Biol 2006; Wang et al., PNAS 2008). Could the authors assess (in situ) the expression of cdx1a, cdx4, hoxb5a, and hoxa9a in the bcas2 mutants?

(3) The authors show compellingly that even heterozygous loss of bcas2 has strong Wnt-inhibitory effects. If Bcas2 is required for canonical Wnt signaling and bcas2 is expressed ubiquitously from the 1-cell stage through at least the beginning of gastrulation, why do bcas2 KO embryos not have morphological axis specification defects consistent with loss of early Wnt signaling, like loss of head (early), or brain anteriorization (later)? Could the authors provide some comments on this puzzle? Or if they do see any canonical Wnt signaling patterning defects in het- or homozygous embryos, could they describe and/or present them?

Reviewer #3 (Public Review):

Summary:

This manuscript utilized zebrafish bcas2 mutants to study the role of bcas2 in primitive hematopoiesis and further confirms that it has a similar function in mice. Moreover, they showed that bcas2 regulates the transition of hematopoietic differentiation from angioblasts via activating Wnt signaling. By performing a series of biochemical experiments, they also showed that bcas2 accomplishes this by sequestering b-catenin within the nucleus, rather than through its known function in pre-mRNA splicing.

Strengths:

The work is well-performed, and the manuscript is well-written.

Weaknesses:

Several issues need to be clarified.

(1) Is wnt signaling also required during hematopoietic differentiation from angioblasts? Can the authors test angioblast and endothelial markers in embryos with wnt inhibition? Also, can the authors add export inhibitor LMB to the mouse mutants to test if sequestering of b-catenin by bcas2 is conserved during primitive hematopoiesis in mice?

(2) Bcas2 is required for primitive myelopoiesis in ALM. Does bcas2 play a similar function in primitive myelopoiesis, or is bcas2/b-catenin interaction more important for hematopoietic differentiation in PLM?

(3) Is it possible that CC1-2 fragment sequester b-catenin? The different phenotypes between this manuscript and the previous article (Yu, 2019) may be due to different mutations in bcas2. Is it possible that the bcas2 mutation in Yu's article produces a complete CC1-2 fragment, which might sequester b-catenin?

(4) Can the author clarify what embryos the arrows point to in SI Figure 2D? In SI Figure 6B and B', can the author clarify how the nucleus and cytoplasm are bleached? In B, the nucleus also appears to be bleached.

Author Response:

Thank you very much for your consideration and assessment. We really appreciate the generous comments from the reviewers on our manuscript entitled “BCAS2 promotes primitive hematopoiesis by sequestering β-catenin within the nucleus”. The comments are very helpful for the improvement of our work. We would like to provide the following provisional revision plan to address the public reviews:

1. To clarify if Bcas2 also promotes primitive myelopoiesis by enhancing nuclear accumulation of β-catenin, bcas2 morpholino will be injected into the Tg(coro1a:EGFP) zebrafish embryos at 1-cell stage, and subsequently the β-catenin distribution in the myeloid cells will be examined. Tg(coro1a:EGFP) is commonly used to track both macrophages and neutrophils.

2. According to the reviewers’ comments, we will quantify the fluorescence intensity in the cell nucleus and cytoplasm in Figure 3H. Meanwhile, we will adjust the exposure of Figure 5C and Figure 7E, or replaced the figures with high-resolution ones.

3. Previous studies have reported that β-catenin can bind directly to CRM1 through its central armadillo (ARM) repeats region. β-catenin region containing ARM repeats 10 and the C terminus are essential for its nuclear export (Koike M, et al., The Journal of Biological Chemistry, 2004). In our research, BCAS2 has been demonstrated to bind to the 9-12 ARM repeats of β-catenin. Therefore, it is highly likely that Bcas2 may compete with CRM1 for binding with the nuclear export signal peptide on β-catenin. To further test this possibility, we will transfect HEK293T cells with constructs expressing full-length or truncated forms of β-catenin, and then examine their nuclear distribution.

4. To validate if BCAS2 affects CRM1-dependent nuclear export of other classical factors, we plan to knock down or overexpress BCAS2 in HeLa cells, and detect the distribution of ATG1 and CDC37L, which have been identified as CRM1 cargoes.

5. Considering that the ARM repeats bound by Bcas2 (repeats 9-12) and Tcf (repeats 2-9) might not be mutually exclusive, it is indeed appealing to investigate whether β-catenin can simultaneously interact with Tcf and Bcas2. We will follow review’s suggestion to perform a three-way co-immunoprecipitation assay. Plasmids encoding these three proteins will be co-transfected into cells. Cell lysates will be immunoprecipitated using antibodyspecific to the bait protein (e.g., β-catenin) and eluted proteins will be analyzed using antibodies specific to the other two proteins.

6. To elucidate that canonical Wnt signaling regulates hematopoietic development by activating expression of cdx1a, cdx4, and their downstream targets hoxb5a and hoxa9a as previously reported, we intend to examine the expression of cdx4 and hoxa9a in bcas2+/- embryos at 10 ss by performing in situ hybridization.

7. To further validate whether Wnt signaling is required during endothelial differentiation from angioblasts, wild-type embryos will be subjected to treatment with Wnt inhibitor CCT036477 and the expression of hemangioblast markers npas4l, scl, and gata2 and endothelial markers fli1 will be analyzed using in situ hybridization.

8. In order to clarify whether coiled-coil (CC) domain 1-2 of Bcas2 is sufficient to interact with β-catenin and restore the primitive hematopoietic defect, we will overexpress CC1-2 in Tg(gata1:GFP) embryos injected with bcas2 morpholino, and then investigate the distribution of β-catenin, as well as gata1 expression at 10 ss in these embryos.