Cdx1 and Cdx2 are functionally equivalent in vertebral patterning.

The Cdx transcription factors regulate anterior-posterior (AP) vertebral patterning, at least in part, through direct regulation of Hox gene expression. Analysis of allelic series of Cdx mutant mice suggests functional overlap between these family members. However, the lack of a Cdx2 null mutant makes these analyses incomplete. Moreover, Hox proteins are sometimes redundant, making it difficult to discern whether Cdx members regulate identical Hox target genes in a redundant manner, or whether they regulate separate Hox genes which then converge on events related to vertebral patterning. To more directly assess this question, we developed a "knock in" model whereby Cdx2 was substituted for Cdx1. Consistent with functional redundancy Cdx2 "knock-in" mice exhibited perfect complementation of the Cdx1-null phenotype, as evidenced by the lack of skeletal defects or altered expression of Hox genes typically impacted by Cdx1 loss-of-function. It has been proposed that vertebral AP patterning is reliant on a gradient of the sum total of Cdx proteins, a posit that is consistent with functional redundancy between Cdx family members. To further assess this, we generated a gain-of-function model using BAC transgenesis to alter Cdx1 dosage. Cdx1 BAC transgenic mice overexpressed Cdx1 mRNA and protein, and fully complemented the Cdx1 null allele. However, gain of Cdx1 dosage via this BAC transgene in an otherwise wild type background had no discernible effects on vertebral patterning or Hox gene expression, suggesting that a moderate alteration in the Cdx protein gradient is of no consequence.

Cdx1 autoregulation is governed by a novel Cdx1-LEF1 transcription complex.

The Cdx1 gene product is essential for normal anterior-posterior vertebral patterning. Expression of Cdx1 is regulated by several pathways implicated in anterior-posterior patterning events, including retinoid and Wnt signaling. We have previously shown that retinoic acid plays a key role in early stages of Cdx1 expression at embryonic day 7.5 (E7.5), while both Wnt3a signaling and an autoregulatory loop, dependent on Cdx1 itself, are involved in later stages of expression (E8.5 to E9.5). This autoregulation is reflected by the ability of Cdx1 to affect expression from proximal Cdx1 promoter sequences in tissue culture. However, this region is devoid of a demonstrable Cdx response element(s). We have now found that Cdx1 and LEF1, a nuclear effector of Wnt signaling, synergize to induce expression from the Cdx1 promoter through previously documented LEF/T-cell factor response elements. We also found a direct physical interaction between the homeodomain of Cdx1 and the B box of LEF1, suggesting a basis for this synergy. Consistent with these observations, analysis of Cdx1 Wnt3a(vt) compound mutants demonstrated that Wnt and Cdx1 converged on Cdx1 expression and vertebral patterning in vivo. Further data suggest that Cdx-high-mobility group box interactions might be involved in a number of additional pathways.

Cyclin D1-dependent induction of luminal inflammatory breast tumors by activated notch3.

Accumulating evidence suggests that Notch3 (N3) is involved in breast cancer development, but its precise contributions are not well understood. Here, we report that pregnant mice expressing an activated intracellular form of N3 (N3(IC)) exhibit a cyclin D1-dependent expansion of premalignant CD24(+) CD29(low) luminal progenitors with enhanced differentiation potential in vitro and in vivo. Parous mice developed luminal mammary tumors in a cyclin D1-dependent manner. Notably, mice expressing higher levels of N3(IC) exhibited tumors resembling inflammatory breast cancer that frequently metastasized. N3(IC)-induced tumors contained a large percentage of tumor-initiating cells, but these were reduced significantly in tumors derived from N3(IC) transgenic mice that were heterozygous for cyclin D1. After transplantation in the presence of normal mammary cells, N3(IC)-expressing tumor cells became less malignant, differentiating into CK6(+) CK18(+) CK5(-) alveolar-like structures akin to expanded luminal progenitors from which they were likely derived. Taken together, our results argue that activated N3 signaling primarily affects luminal progenitors among mammary cell subsets, with more pronounced levels of activation influencing tumor type, and provide a novel model of inflammatory breast cancer.

Wnt signaling is a key mediator of Cdx1 expression in vivo.

In the mouse, Cdx1 is essential for normal anteroposterior vertebral patterning through regulation of a subset of Hox genes. Retinoic acid (RA) and certain Wnts have also been implicated in vertebral patterning, although the relationship between these signaling pathways and the regulation of mesodermal Hox gene expression is not fully understood. Prior work has shown that Cdx1 is a direct target of both Wnt and retinoid signaling pathways, and might therefore act to relay these signals to the Hox genes. Wnt and RA are believed to impact on Cdx1 through an atypical RA-response element (RARE) and Lef/Tcf-response elements (LRE), respectively, in the proximal promoter. To address the roles of these regulatory motifs and pathways, we derived mice mutated for the LRE or the LRE plus the RARE. In contrast to RARE-null mutants, which exhibit limited vertebral defects, LRE-null and LRE+RARE-null mutants exhibited vertebral malformations affecting the entire cervical region that closely phenocopied the malformations seen in Cdx1-null mutants. Mutation of the LRE also greatly reduced induction of Cdx1 by RA, demonstrating a requirement for Wnt signaling in the regulation of this gene by retinoids. LRE and LRE+RARE mutants also exhibited vertebral fusions, suggesting a defect in somitogenesis. As Wnt signaling is implicated in somitogenesis upstream of the Notch pathway, it is conceivable that Cdx1 might play a role in this process. However, none of the Notch pathway genes assessed was overtly affected.

Cdx4 is a direct target of the canonical Wnt pathway.

There is considerable evidence that the Cdx gene products impact on vertebral patterning by direct regulation of Hox gene expression. Data from a number of vertebrate model systems also suggest that Cdx1, Cdx2 and Cdx4 are targets of caudalizing signals such as RA, Wnt and FGF. These observations have lead to the hypothesis that Cdx members serve to relay information from signaling pathways involved in posterior patterning to the Hox genes. Regulation of Cdx1 expression by RA and Wnt in the mouse has been well characterized; however, the means by which Cdx2 and Cdx4 are regulated is less well understood. In the present study, we present data suggesting that Cdx4 is a direct target of the canonical Wnt pathway. We found that Cdx4 responds to exogenous Wnt3a in mouse embryos ex vivo, and conversely, that its expression is down-regulated in Wnt3a(vt/vt) embryos and in embryos cultured in the presence of Wnt inhibitors. We also found that the Cdx4 promoter responds to Wnt signaling in P19 embryocarcinoma cells and have identified several putative LEF/TCF response elements mediating this effect. Consistent with these data, chromatin immunoprecipitation assays from either embryocarcinoma cells or from the tail bud of embryos revealed that LEF1 and beta-catenin co-localize with the Cdx4 promoter. Taken together, these results suggest that Cdx4, like Cdx1, is a direct Wnt target.

Retinoic acid regulates a subset of Cdx1 function in vivo.

Hox gene products are key players in establishing positional identity along the anteroposterior (AP) axis. In vertebrates, gain or loss of Hox expression along the AP axis often leads to inappropriate morphogenesis, typically manifesting as homeotic transformations that affect the vertebrae and/or hindbrain. Various signalling pathways are known to impact on Hox expression, including the retinoid signalling pathway. Exogenous retinoic acid (RA), disruption of enzymes involved in maintaining normal embryonic RA distribution or mutation of the retinoid receptors (RARs and RXRs) can all impact on Hox expression with concomitant effects on AP patterning. Several Hox loci have well characterized RA response elements (RAREs), which have been shown to regulate functionally relevant Hox expression in the neurectoderm. A similar crucial function for any RARE in mesodermal Hox expression has, however, not been documented. The means by which RA regulates mesodermal Hox expression could therefore be either through an undocumented direct mechanism or through an intermediary; these mechanisms are not necessarily exclusive. In this regard, we have found that Cdx1 may serve as such an intermediary. Cdx1 encodes a homeobox transcription factor that is crucial for normal somitic expression of several Hox genes, and is regulated by retinoid signalling in vivo and in vitro likely through an atypical RARE in the proximal promoter. In order to more fully understand the relationship between retinoid signalling, Cdx1 expression and AP patterning, we have derived mice in which the RARE has been functionally inactivated. These RARE-null mutants exhibit reduced expression of Cdx1 at all stages examined, vertebral homeotic transformations and altered Hox gene expression which correlates with certain of the defects seen in Cdx1-null offspring. These findings are consistent with a pivotal role for retinoid signalling in governing a subset of expression of Cdx1 crucial for normal vertebral patterning.