The mechanisms by which myc regulates both ESC biology and the reprogramming required for iPSC formation are important open questions with critical implications for tumorigenesis as well. Our study suggests a model () in which myc contributes to these pluripotency and self-renewal related functions through inducing expression of pluripotency-related genes including lif and those encoding master stem cell factors KLF2, KLF4, and LIN28B. Our findings indicate that a very similar N-Myc regulated program is at work in neuroblastoma and could play a role in its genesis through promoting an aberrant pluripotent state.
A model of Myc stem-related function in neuroblastoma cells.
expression and expression of klf2
, and lin28b
are likely two independent mechanisms by which N-Myc contributes to pluripotency. The regulation of lif
expression by N-Myc is a mechanism by which it may contribute to neuroblastoma genesis but also ESC and iPSC biology. If N-Myc stimulates the production of lif
during the early stages of neuroblastoma genesis (), the presence of this potent stem cell related ligand could contribute to tumorigenesis through both autocrine and paracrine signaling that could drive the formation or maintenance of neuroblastoma stem cells (). However, lif
expression could also be important later in tumorigenesis, perhaps even in tumor maintenance, as a mechanism for preventing differentiation of neuroblastoma. Importantly, our studies were conducted in human neuroblastoma and mouse NSC. While LIF protein has distinct functions in human and mouse ESC, its role in NSC generally is less well understood and there is not currently any evidence of a distinct role for LIF in NSC or neural tumors of different species. However, there is clear evidence that LIF regulates self-renewal and pluripotency of both mouse and human NSC 
Of interest is our finding that N-Myc does not appear to regulate lif in NSC, suggesting the regulation of lif in the neuronal context could be tumor specific. Our discovery of a link between N-Myc and lif in neuroblastoma also suggests a possible new treatment for neuroblastoma in the form of LIF antagonists that would be predicted to induce regression through stimulating differentiation (). Besides lif expression, we found other differences in stem-related genes regulated by N-Myc in NSC and neuroblastoma. For example, while N-Myc did not appear to be required for nanog expression in neuroblastoma, disruption of N-myc in NSC caused a pronounced decrease in nanog expression. These findings suggest that unique stem cell-related targets exist for N-Myc both in NSC and in neuroblastoma. NSC may be more fully pluripotent, whereas neuroblastoma may express stem-related genes but have at least partially defective pluripotency.
While we have evidence that N-Myc directly regulates lif
through canonical CACGTG E-boxes and triMeK4 in their promoters, it is also possible that Myc's induction of lif
is an indirect mechanism by which it acts to also maintain expression of other important pluripotency associated genes that are dependent on the action of LIF as a growth factor. Our findings of N-Myc regulating lif
in neuroblastoma also fits with previous work indicating overexpression of myc
confers ectopic LIF-independence on ESC 
. Together these data suggest that myc
overexpression in mESC may in part achieve this end through stimulating expression of endogenous LIF. It is important to note that one previous study found that N-myc
overexpression was correlated with reduced LIF protein levels 
in some neuroblastoma suggesting that N-myc
induction of lif
may occur only in a subset of neuroblastoma. Currently it remains unknown what genes mediate Myc function in mESC to specifically maintain self-renewal and pluripotency, but the targets we have identified here in neuroblastoma are candidates as effectors in ESC as well. Also fitting with our data are the recent observations that c-Myc regulates lin28 
and lin28b 
expression. In the case of lin28b
, Myc directly binds a canonical CACGTG E-box, suggesting that our findings of N-Myc regulating lin28b
expression in neuroblastoma and in NSC may be mediated through N-Myc direct binding of this E-box as well. Since LIN28B functions through regulation of miRNA processing including that of let-7, N-Myc activation of LIN28B in neuroblastoma may contribute to maintenance of an miRNA program that enforces an aberrant pluripotent state ().
The regulation of lif
expression by N-Myc through the CACGTG E-box also correlates with regulation of triMeK4, a key euchromatic histone mark associated with active transcription, within the promoter. Decreased N-Myc expression causes a sharp decrease in triMeK4 accompanied by a pronounced decrease in N-Myc binding. There is also some indication that decreased N-Myc reduces triMeK4 in the promoters of lin28b
. Together these findings suggest N-Myc maintains a transcriptionally active chromatin state at pluripotency genes in neuroblastoma. In contrast, at most of the stem cell-related genes tested, decreased N-Myc surprisingly resulted in increased AcK9 in their promoters. Given the recruitment of histone acetyltransferases by Myc proteins, particularly GCN5 
and the dependence on N-Myc for widespread maintenance of AcK9, another key euchromatic mark, in neuroblastoma 
, it is somewhat surprising that decreased N-Myc levels would be accompanied by increased acetylation at these specific genes. At this point it is unclear what mechanism could be responsible for this change and why it would occur specifically at stem cell related genes, but it is possible that pluripotency related genes remain in a poised state rather than be silenced. An increase in AcK9 modification accompanying loss of triMeK4 may prevent a fully silenced state.
Our findings also have implications for iPSC and it is striking that we found N-Myc regulating 3 other known iPSC-related genes, lin28b, klf2, and klf4. These data suggest a model in which overexpressed myc enhances iPSC formation in fibroblasts at least in part by turning on klf family and lin28b gene expression, and through inducing expression of lif. Notably our data also provide the first model for why overexpressed myc, although a potent enhancer of iPSC formation, may not be formally required for the process: if expression levels of klf and lin28 as well as other pluripotency-related genes are high enough, myc may become more dispensable since it is no longer required to turn on their expression. Endogenous lif expression may also be dispensable since ectopic LIF is often added to iPSC media. However, alternatively, the ability to generate iPSC without added myc may very well be a result of high levels of endogenous myc expression. The presence of an iPSC-related gene expression program in neuroblastoma also raises the concern of the tumorigenicity of iPSC.
In our previous model, we proposed that myc genes were most likely contributing to tumorigenesis and perhaps iPSC formation through both gene specific and global chromatin events. Our new findings confirm an important role for Myc's gene specific, classical transcription factor function in neuroblastoma. The potential contributions of a more global Myc chromatin function to neuroblastoma genesis and delineating the mechanisms by which Myc contributes to iPSC biology await future study. Particularly important will be functional genomics assays addressing Myc chromatin function, not just binding, in iPSC and in additional types of tumors.