All tested fragments. In addition to confirming the intron 5 site as an enhancer, these data suggested that Zfp423 binding might be a negative regulator of its own transcription.shRNA alone. Lecirelin expression of the GFP-marked shRNA construct dramatically reduced Zfp423 immunofluorescence in P19 nuclei using either the E20 (Figure 5D) or custom made (Figure 5E) antibody. Western blots confirmed effective reduction of protein expression after shRNA transfections (Figure 5F,G) and overexpression after transfection with human ZNF423 expression plasmid (Figure 5H).DiscussionZfp423 acts coordinately with several lineage and signalingdependent transcription complexes that are important for precursor cell differentiation, including direct interactions with Ebf, BMP-dependent Smad, retinoic acid receptor, and Notch intracellular domain proteins [5,8,10,12,13]. While dynamically expressed in embryos, postnatal Zfp423 expression attenuates with maturation in many cell lineages and artificially sustained expression in at least the olfactory lineage inhibits normal maturation [4,10,19]. Early observations in the olfactory system by Reed and co-workers indicated that Zfp423 is co-expressed with Ebf1, which Zfp423 inhibits, in immature cells. As these precursors mature, Ebf1 stays on while Zfp423 turns off, promoting Ebf1 transactivation of lineage-specific target genes during differentiation [10]. Conversely, re-expression of Zfp423 reverts the expression of several key genes to an immature state [4]. How this stereotyped progression is regulated is not known, but the sequence of expression states would be consistent with a self-regulating transcriptional network. Our results demonstrate that endogenous Zfp423 directly binds two conserved non-coding sites in its own gene, providing a potential autoregulatory mechanism, in both mouse and human cell culture models; that the intron 5 site acts as an enhancer by classical reporter assays in a P19 cell culture model; and that while this enhancer does not require Zfp423 or its predicted binding site for activity, the enhancer is substantially suppressed by high level expression of ZNF423. These results provide a plausible mechanism for signalinduced progression of precursor cells dependent on Zfp423, using autoregulatory sites to scaffold feed-forward or Licochalcone A biological activity feedback loops, depending upon the cell state, signaling environment, and the state of co-regulatory factors. The intron 5 enhancer element is functionally blocked by overexpression of ZNF423. This may suggest an autoregulatory negative feedback loop under conditions of high Zfp423 expression, even though knockdown of endogenous Zfp423 RNA (to ,24 of control levels) does not appear to alter enhancer activity. It is possible that the repression caused by exogenous expression is indirect and a consequence of ZNF423 activity on other genes. However, as the enhancer is directly bound by Zfp423 in a conserved fashion and mutation of the presumed binding sites eliminates the effect of overexpression (Figure 5), this seems to us less likely. Alternatively, it is possible that in this experimental system and under these conditions, the level of enhancer occupancy (or its context for other binding partners) is not sensitive to reduced levels, while addition of exogenous ZNF423 alters either the percent occupancy or composition of the binding complex sufficiently to repress activation. For example, while knockdown of Ebf1 reduced reporter expression, simultaneousZfp423 O.All tested fragments. In addition to confirming the intron 5 site as an enhancer, these data suggested that Zfp423 binding might be a negative regulator of its own transcription.shRNA alone. Expression of the GFP-marked shRNA construct dramatically reduced Zfp423 immunofluorescence in P19 nuclei using either the E20 (Figure 5D) or custom made (Figure 5E) antibody. Western blots confirmed effective reduction of protein expression after shRNA transfections (Figure 5F,G) and overexpression after transfection with human ZNF423 expression plasmid (Figure 5H).DiscussionZfp423 acts coordinately with several lineage and signalingdependent transcription complexes that are important for precursor cell differentiation, including direct interactions with Ebf, BMP-dependent Smad, retinoic acid receptor, and Notch intracellular domain proteins [5,8,10,12,13]. While dynamically expressed in embryos, postnatal Zfp423 expression attenuates with maturation in many cell lineages and artificially sustained expression in at least the olfactory lineage inhibits normal maturation [4,10,19]. Early observations in the olfactory system by Reed and co-workers indicated that Zfp423 is co-expressed with Ebf1, which Zfp423 inhibits, in immature cells. As these precursors mature, Ebf1 stays on while Zfp423 turns off, promoting Ebf1 transactivation of lineage-specific target genes during differentiation [10]. Conversely, re-expression of Zfp423 reverts the expression of several key genes to an immature state [4]. How this stereotyped progression is regulated is not known, but the sequence of expression states would be consistent with a self-regulating transcriptional network. Our results demonstrate that endogenous Zfp423 directly binds two conserved non-coding sites in its own gene, providing a potential autoregulatory mechanism, in both mouse and human cell culture models; that the intron 5 site acts as an enhancer by classical reporter assays in a P19 cell culture model; and that while this enhancer does not require Zfp423 or its predicted binding site for activity, the enhancer is substantially suppressed by high level expression of ZNF423. These results provide a plausible mechanism for signalinduced progression of precursor cells dependent on Zfp423, using autoregulatory sites to scaffold feed-forward or feedback loops, depending upon the cell state, signaling environment, and the state of co-regulatory factors. The intron 5 enhancer element is functionally blocked by overexpression of ZNF423. This may suggest an autoregulatory negative feedback loop under conditions of high Zfp423 expression, even though knockdown of endogenous Zfp423 RNA (to ,24 of control levels) does not appear to alter enhancer activity. It is possible that the repression caused by exogenous expression is indirect and a consequence of ZNF423 activity on other genes. However, as the enhancer is directly bound by Zfp423 in a conserved fashion and mutation of the presumed binding sites eliminates the effect of overexpression (Figure 5), this seems to us less likely. Alternatively, it is possible that in this experimental system and under these conditions, the level of enhancer occupancy (or its context for other binding partners) is not sensitive to reduced levels, while addition of exogenous ZNF423 alters either the percent occupancy or composition of the binding complex sufficiently to repress activation. For example, while knockdown of Ebf1 reduced reporter expression, simultaneousZfp423 O.
