Ajay Chitnis, MBBS, PhD, Head, Section on Neural Developmental Dynamics
Motoyuki Itoh, PhD, Postdoctoral Fellow 1
Michael Keller, PhD, Postdoctoral Fellow
Kinneret Rand, PhD, Postdoctoral Fellow
Sang-Yeob Yeo, PhD, Postdoctoral Fellow 2
Gregory Palardy, BS, Research Technician
Chongmin Wang, MS, Research Technician
The earliest differentiating neurons in the zebrafish neural plate are generated in well-defined neurogenic domains that are separated by non-neurogenic domains. Analysis of the mechanisms by which these discrete domains are defined in zebrafish provides important clues about how neurogenesis is regulated in the vertebrate nervous system. Previous studies in the frog had suggested a critical role for zic2 in defining non-neurogenic domains; however, analysis in zebrafish has suggested a more complex role in defining both neurogenic and non-neurogenic domains. Our recent studies show how in both domains the Zic factors regulate neurogenesis by limiting expression of certain Hairy-related transcription factors. The neural plate eventually undergoes cell movements and rearrangements to form the neural tube. In the hindbrain, further patterning and morphogenetic events define segmental compartments called rhombomeres; their boundaries become organizing centers for neurogenesis. Analysis of Mosaic Eyes (Moe), a FERM domain protein, has provided insight into the role of Notch signaling in the morphogenetic events leading to the establishment of rhombomere boundaries. Moe was originally identified in a yeast two-hybrid screen with Mind bomb, a critical component of the Notch signaling pathway. Its analysis provides insight into the poorly understood role of Notch in morphogenesis.
Zic genes limit her9 expression to mediate opposing effects on neurogenesis
Keller, Chitnis
Early neurogenesis occurs in three longitudinal columns within the caudal neural plate in zebrafish and Xenopus embryos. The patterning mechanisms that delineate neurogenic and non-neurogenic domains are poorly understood, but analyses in Xenopus had suggested a role for the zinc-finger transcription factor Zic2 in defining non-neurogenic domains within the neural plate. To determine if Zic2-related factors play a similar role in zebrafish, we used a loss-of-function approach with morpholino-mediated protein knock-down to assess the role of zic genes expressed during early neurogenesis. Rather than defining non-neurogenic domains, we found that zebrafish zic2a, zic2b, and zic3 functioned together to promote neuronal differentiation in the neural plate. However, consistent with their suggested function in Xenopus, the zic genes also contributed to the suppression of neurogenesis in the adjacent trigeminal placode. The opposing roles of zic2a, zic2b, and zic3 genes in the trigeminal placode and neural plate presented a puzzle, and we tried to identify potential transcriptional targets to understand how the Zic factors might influence neurogenesis in different ways in these two adjacent neural tissues.
While our studies were in progress, it was reported that her3, her5, her9, and her11, a subset of Hairy-Enhancer of split-Related (HER) genes, play critical roles in defining non-neurogenic territories in the zebrafish neural plate. To determine if some effects of Zic factors on neurogenesis are directly or indirectly related to the function of some of these HER genes, we examined changes in the expression of her9, a HER gene whose expression was spatiotemporally most closely related to neurons whose differentiation was altered with zic gene knock-down. Consistent with the fact that the combinatorial function of zic2a, zic2b, and zic3 in neurogenesis is related, at least in part, to regulation of her9, the expression of her9 was expanded in both the neural plate and adjacent placodal ectoderm in zic2a+zic2b+zic3 morphant embryos. Furthermore, the concurrent knock-down of her9 in the zic2a+zic2b+zic3 morphant embryos "rescued" neuronal differentiation in both tissues despite their qualitatively opposite phenotypes: neuronal differentiation was restored in the neural plate while the overproduction of neurons in the trigeminal placode was suppressed. These observations led to the conclusion that the function of the zic genes in both the neural plate and adjacent trigeminal placode was dependent on the genes' ability to repress her9. Thus, Zic proteins function by limiting her9 expression, thereby either promoting or suppressing neuronal differentiation in a context-dependent manner. Elucidating the mechanism by which HER genes such as her9 have opposing effects on neurogenesis in the trigeminal placode and neural placode is the next subject of inquiry.
Mosaic eyes helps stabilize Delta and restrict Notch activity to rhombomere boundaries
Rand, Itoh, Palardy, Yeo, Chitnis
Delta-Notch signaling limits the differentiation of neural progenitors as early neurons in the zebrafish neural plate. In this context, a broad failure of effective Notch signaling in mind bomb (mib) mutants results in the production of an excess of early neurons. The gene mib encodes an E3 ligase that promotes Delta ubiquitylation and endocytosis, a step that is essential for Delta's function as a Notch ligand.
In recent studies, we identified a FERM domain protein Mosaic Eyes (Moe) as a potential Mib-interacting protein. In cell culture, Moe is typically associated with the cell surface and, in co-transfection experiments, it alters Mib's cellular distribution, reducing Mib in the cytoplasm and stabilizing it at the cell surface. In contrast, co-expression of Mib with a truncated Moe that lacks the N-terminal fragment responsible for its surface localization fails to stabilize Mib at the cell surface and instead retains it in the cytoplasm.
Mib ubiquitylates DeltaD and promotes its internalization, dramatically reducing surface DeltaD when the two are co-expressed in COS cells. Co-expression of Moe with Mib and DeltaD, however, increases surface DeltaD, suggesting that Moe reduces overall internalization of DeltaD by Mib and/or facilitates recycling to the cell surface. Co-transfection of Mib and DeltaD with N-terminal-deleted Moe also increases surface Delta, though not as effectively as full-length Moe. Remarkably, however, co-transfection of DeltaD with the N-terminal-deleted Moe reduces DeltaD protein, suggesting that retention of Mib in the cytoplasm by N-terminal-deleted Moe may facilitate degradation of DeltaD. Destabilization of Delta by the truncated Moe raised the possibility that stabilization of Mib at the cell surface by full-length Moe may normally contribute to Delta stability.
To test our prediction, we examined endogenous DeltaD expression in zebrafish embryos injected with antisense moe morpholinos. Although we observed no change in DeltaD expression during early development, a dramatic loss of DeltaD protein occurred in the hindbrain at 24 hours after fertilization. Knock-down of moe also interfered with morphogenesis of rhombomere boundaries: genes whose expression is typically restricted to rhombomere boundaries were deregulated and spread throughout the hindbrain in moe morphants. Rhombomere-specific gene expression, however, was not affected, suggesting a specific problem with boundary formation rather than with patterning events that define discrete rhombomere compartments.
Restriction of Notch activity is thought to be essential for morphogenesis of rhombomere boundaries, and cells with relatively high Notch activity segregate to these boundaries. In moe morphants, Notch activity is no longer restricted to boundaries, and the Notch target gene her4 is expressed throughout some rhombomeres. Interestingly, knock-down of DeltaD with morpholinos also deregulated her4 expression, supporting the idea that (1) DeltaD protein has a critical role in restricting Notch activity to rhombomere boundaries and (2) loss of DeltaD protein, resulting from either destabilization in moe morphants or direct knock-down, leads to deregulated Notch activity.
Though it remains unclear how Moe determines DeltaD stability and how Delta limits Notch activity to rhombomere boundaries, our observation identify Moe as a new component of a previously identified patterning mechanism whereby complementary Notch and Wnt signaling collaborate to restrict Notch signaling and expression of various genes to rhombomere boundaries.
Publications related to other work
1 Motoyuki Itoh, PhD, former Postdoctoral Fellow, currently at the Graduate School of Science, Nagoya University, Nagoya, Japan
2 Sang-Yeob Yeo, PhD, former Postdoctoral Fellow, currently at Kyungpook National University, Daegu, Republic of Korea
For further information, contact chitnisa@mail.nih.gov.