Snail family zinc finger transcriptional repressors are essential for the formation of the neural crest stem cells, and are also required later in neural crest development for the onset of invasive and migratory behaviour. The proliferative, anti-apoptotic, migratory, and invasive properties that Snail factors confer on neural crest cells are recapitulated when Snai1 and Snai2 and other core regulators of EMT, including Zeb1 and Zeb2, are expressed in epithelial tumor cells. The Snail protein family is highly studied for these reasons, and inhibitors of Snail function are therefore of high potential impact and importance for both basic science and clinical applications.
We have shown Snail factors can be potently inhibited by Co(III)-Ebox in vitro and in vivo. Inhibiting DNA binding by Snail proteins alleviates Snail-mediated transcriptional repression. The effects of Co(III)-Ebox are highly selective, as sequence specific controls with two base substitutions have greatly diminished effects on Snail function, strong evidence that off-target effects are limited. In vitro data further confirms that Co(III)-Ebox does not inhibit other Ebox-binding proteins that do not contain zinc finger domains, including MitF. Together these data strongly substantiate the specificity of Co(III)-Ebox for Ebox-binding zinc finger transcription factors, and indicate that there are at most nominal off-target effects on other zinc finger proteins, or other Ebox-binding proteins such as bHLH transcription factors.
In contrast to RNAi approaches, Co(III)-Ebox-mediated inhibition of Snai1 blocks Snai1 activity without affecting Snai1 protein levels. Co(III)-DNA conjugates therefore represent a novel approach for specifically blocking the function of targets in vitro and in vivo, without the requirement for decreasing the expression level of the protein itself. While in the present report we utilize this approach to block Snail protein function in tumor derived cells and early embryos, this methodology should prove applicable to targeted regulation of transcriptional events by a broad range of potential protein targets.
To demonstrate the utility of the Co(III)-Ebox conjugate, we have used it to probe the temporal requirements for Snail family function during neural crest development in Xenopus. We have used two approaches to controlling the timing of Co(III)-Ebox-mediated Snail inhibition in early embryos. Control over the time of application is a major advantage that chemical inhibitors hold over genetic means of inhibiting protein activity. Reagents such as RNAi affect the translation of new proteins but do nothing to inhibit protein already expressed within a cell, which must turn over before functional effects will be observed. Moreover, the temperature dependence of Co(III)-Ebox function affords an additional level of control that will prove of great value in model organism-based studies. As a consequence the Co(III)-DNA conjugate can be introduced into cells or tissues at one time but kept at a temperature where it is largely inactive. Because the reagent remains stable in the cytoplasm for long periods of time, Co(III)-Ebox-mediated inhibition can subsequently be activated by shifting to a higher temperature ().
Proposed model for the temperature-dependent inactivation of proteins targeted by Co(III)-Ebox.
In this study we demonstrate the power that this level of control affords. We first demonstrate the specificity and temporal control of Co(III)-Ebox by showing that this means of inhibiting Snail function phenocopies the effects of molecular genetic approaches for inhibiting Snail function, while limiting off-target effects 
. Co(III)-Ebox effectively blocks the formation of the neural crest precursor cells and the subsequent migration of these cells, while not impairing the formation of the CNS or mesoderm. Indeed, the current experiments shed further light on the temporal requirements for Snail function at premigratory neural crest stages. These experiments indicate that once neural crest precursors have formed, Snail function is not required for the maintenance of these precursors, as indicated by the continued expression of neural crest markers even in the presence of Co(III)-Ebox. The next functional consequence of Co(III)-Ebox-mediated Snail inhibition was the failure of proper neural crest migration. To examine if there were further requirements for Snail function, such as during neural crest cell fate diversification, we activated Co(III)-Ebox only after neural crest migration was well underway. In this way we found that Snail function was required for the normal formation of at least one neural crest derivative, melanocytes, whereas neural crest derived cartilage and cranial neurons and glia developed normally despite inhibition of Snail function.
Snail repressors are known to directly or indirectly regulate the expression of other transcription factors in the complex gene regulatory network that controls neural crest formation and differentiation. In Xenopus
, Snai1 is known to positively affect Sox10
expression during the formation of neural crest precursor stem cells, presumably indirectly 
. This is significant given that at later stages of development Sox10
has been shown to play an essential role in the formation of the pigment cell lineage including melanocytes 
. Although a role for Snail-dependent regulation of Sox10
at these stages has not been previously reported, it is possible that inhibition of melanocyte formation by Co(III)-Ebox could result from the down-regulation of Sox10
following inhibition of Snail function. FoxD3
also plays roles in pigment cell and glia formation 
and is positively regulated by Snail, however, we found that FoxD3
expression in glial cells was unperturbed at stage 28 in treated embryos. Moreover, given that Sox10
has been shown to promote both melanocyte and glial cell formation while inhibiting neuron formation 
, the normal development of glia and neurons in treated embryos suggests that Sox10
function is not lost in all lineages. Further examination of the temporal requirements of Snail function, and its impact on the expression of other neural crest regulatory factors, should lead to a fuller understanding of the role of Snail in melanocyte formation as well as other neural crest cell derivatives.
Our findings on the effects of Co(III)-Ebox-mediated Snail inhibition in neural crest cells have implications beyond the development of this important cell type. Previous studies have reported the utility of Co(III) Schiff base complexes as targeted protein inhibitors in vitro 
. Moreover, non-targeted Co(III) Schiff base complexes have demonstrated some efficacy as viral inhibitors following topical application. 
. Our findings on the inhibition of transcription factors in tumor-derived cells, and in neural crest stem cells, are the first report of a Co(III) Schiff base complex functioning as an intracellular and even nuclear inhibitor in vivo, and the first used in developmental studies, and as such represent a significant advancement in the field.
Our results further demonstrate the power and potential of Co(III)-DNA conjugates for the study of developmental events regulated by zinc finger transcription factors, and highlight the exceptional temporal control afforded by these agents. Generation of Co(III) Schiff base complexes targeted to other zinc finger transcription factors should prove a broadly applicable strategy for targeted inhibition of factors that play key roles in other developmental and disease processes. The flexible nature of such DNA-conjugated Co(III) Schiff base complexes permits more facile development of inhibitors compared to conventional small molecule drugs, since novel inhibitors can be rationally designed simply by changing the targeting sequence. In the case of Co(III)-Ebox, our findings suggest that this conjugate holds significant therapeutic promise as an inhibitor of Snail-dependent tumor progression and recurrence. We further show that this agent can be used in model organisms to shed important light on the function of this important family of transcriptional regulatory proteins.