In this study we examined the biological role of Huwe1, a HECT-domain E3 ubiquitin ligase whose properties and function in vivo were unclear. We demonstrated the existence of a Huwe1–N-Myc pathway, in which Huwe1 operates on a linear pathway epistatic to N-Myc to allow neuronal differentiation and cell-cycle arrest of stem cells and cortical progenitors. We discovered that the phenotypic and molecular changes resulting from Huwe1 knockdown are fully reversed when expression of the N-myc gene is also silenced. Through experiments conducted in ES cells and neural progenitors in vivo, we showed that the mechanistic explanation for the above findings is the ability of Huwe1 to bind to N-Myc and prime it for Lys 48-linked polyubiquitination and proteasomal-mediated degradation.
The Lys 48-mediated mechanism of ubiquitin linkage on N-Myc by Huwe1 and its recognition by the 26S proteasome for destruction is consistent with in vitro
and in vivo
observations in neuroblastoma cells, differentiating ES cells and developing brain where silencing of Huwe1 expression invariably results in N-Myc accumulation and an extended half-life. Although a previous study found that c-Myc is also a substrate of Huwe1-mediated ubiquitination, it suggested that the polyubiquitin chains are mostly linked through Lys 63 and do not target degradation of c-Myc by the proteasome16
. However, in this study, Huwe1 seemed to assemble different types of polyubiquitin chains (Lys 11, Lys 48, Lys 63). Here, we have conducted a detailed functional comparison of N-Myc and c-Myc as Huwe1 substrates. This analysis led to three conclusions. First, N-Myc and c-Myc are both ubiquitinated in vivo
by Huwe1 through Lys 48-mediated linkages (). Second, ubiquitination of N-Myc by Huwe1 is markedly more efficient than ubiquitination of c-Myc under identical experimental conditions (). This effect is probably the consequence of a more efficient interaction of Huwe1 with N-Myc than with c-Myc (). Third, constitutive and conditional genetic knockout of Huwe1
in ES cells results in the stabilization of endogenous N-Myc and c-Myc, albeit with markedly different rates (N-Myc stabilization is at least 3-fold more pronounced than c-Myc, , ).
The rather poor efficiency by which c-Myc binds, is ubiquitinated and degraded by Huwe1 may be one of the reasons for the failure to observe stabilization of c-Myc in cells where residual Huwe1 expression may still be present after RNA-interference-mediated silencing16
. Furthermore, in the previous study, the ubiquitination assays were performed using a truncated version of Huwe1 that lacks the first 2472 amino acids16
, whereas we have used full-length Huwe1 for all our experiments. Finally, the observation that Huwe1 assembles only Lys 48-mediated linkages on its substrates is consistent with that of a recent report in which a combination of ubiquitin mutagenesis and mass spectrometry experiments revealed that HECT-domain E3 ligases can only form homogeneous ubiquitin chains (Lys 48 or Lys 63; ref. 24
The most striking effect that emerged from inactivation of Huwe1 in ES cells and developing brain is the impaired differentiation along the neuronal lineage. We suggest that the activity of Huwe1 as negative regulator of N-Myc provides a mechanistic explanation for this observation. Support for this idea comes from the sharp gradient of Huwe1 expression in the developing central nervous system, whereby Huwe1 is markedly upregulated as neuronal differentiation proceeds. Accordingly, the cortical plate, which contains only post-mitotic and differentiated neurons that are invariably N-Myc-negative, is the area in the mouse embryo with the highest expression of Huwe1 at mid-gestation. Furthermore, the conclusion that Huwe1 operates on a linear pathway epistatic to N-Myc is firmly supported by the notion that concurrent silencing of Huwe1
not only rescues the differentiation defect of Huwe1
-null neurons but actually converts it into a premature cell-cycle exit/differentiation phenotype that is indistinguishable from that produced by single knockdown of N-myc
. As well as protecting post-mitotic neurons from re-expression of N-Myc, thus preventing cell-cycle re-entry and loss of the differentiated state, the high levels of Huwe1 expression in the cortical plate may also be necessary to target other substrates possibly implicated in specialized neuronal activities. Consistent with this hypothesis is a recent screening in Caenorhabditis elegans
, from which Huwe1
emerged as an essential gene for synaptogenesis46
. However, Huwe1 substrates regulating synaptogenesis have not been found. Regardless of the full scope of Huwe1 substrates in differentiated neurons, our findings indicate that N-Myc is the crucial substrate in differentiating neural progenitors and identify the Huwe1–N-Myc pathway as a key modulator of differentiation and quiescence in the neural tissue.