This project was initiated by the observation that a large percentage of mice in one of six transgenic lines with cardiac-specific overexpression of human CYP2J2 exhibited head swelling followed by rapid neurological deterioration and death in young adulthood. We hypothesized that the transgene had interrupted the coding or regulatory region of an important gene. We identified the 5 prime and 3 prime genomic sequences adjacent to the single transgene insertional site and found them to be highly related to a human chromosome 12 sequence that contained the RFX4 locus. The transgene was inserted into an intron of the RFX4 gene and this insertion prevented expression of a novel variant transcript (termed RFX4_v3) which led to the development of congenital hydrocephalus. We found that wild type (+/+) and transgene-interrupted alleles (+/- for one interrupted allele and -/- for two interrupted alleles) could be readily distinguished by Southern blotting and PCR. Brains from hemizygous (+/-) mice expressed approximately 50% of normal levels of the RFX4_v3 transcript and exhibited severe congenital hydrocephalus of the lateral and third ventricles associated with failure of formation of the subcommissural organ (SCO), leading to stenosis of the aqueduct of Sylvius. In contrast, the homozygous null (-/-) mice had a severe defect in telencephalon formation that led to gross prenatal brain malformations and death in the perinatal period. Indeed, investigation of -/- mice at E12.5 showed that they had lost critical midline structures including the interhemispheric fissure resulting in the formation of a single central ventricle instead of two lateral ventricles. Although the spinal cords were also malformed, facial structures, retinas, olfactory epithelium and all other non-brain tissues examined were unaffected in the -/- mice. The human and mouse RFX4_v3 cDNAs were cloned and found to be 96% identical, indicating that this variant transcript was highly conserved between these two species. Analysis of genomic sequences revealed that the mouse RFX4_v3 transcript was composed of both unique and shared exons with other RFX4 transcripts. The RFX4_v3 transcript was expressed only in brain and initially appeared between E7.5 and E9.5. In situ hybrization revealed that RFX4_v3 expression was highly dynamic during brain development. Importantly, abundant expression of RFX4_v3 was found in the region of the developing SCO in the caudal diencephalon at E14.5. Together, these data indicate that RFX4_v3 is critical for normal brain development. Moreover, the failure of formation of the SCO in +/- mice suggests that this unusual brain organ is extremely sensitive to normal developmental expression of RFX4_v3. To identify the differentially expressed genes in the brains of RFX4_v3 wild-type and mutant mice, RNAs were isolated from the embryonic day 10.5 (E10.5) heads and hybridizations were performed with the Agilent mouse oligo microarrays. Four pairs of individual RNAs and one pair of pooled RNAs were used in the present studies and each pair of samples was hybridized to two oligonucleotide microarrays with each RNA sample labeled with each fluorophore to account for fluorophore incorporation bias. Thirty-four genes were up-regulated and 75 genes were down-regulated in at least 4 out of 5 pairs of samples. Twenty-four genes in the outlier lists were chosen for validation by real-time PCR based on i) relatively large fold changes by microarray analysis, and ii) potentially important functions during brain development. Real-time PCR results indicated that 22 out of 24 genes were confirmed to express differentially with similar or larger fold changes compared to microarray data. The high correlation between microarray data and real-time PCR results demonstrates the utility of using microarrays in identifying potential downstream target genes of transcription factors during brain morphogenesis. Some of the validated genes are well-known regulators for neuronal development. For example, Cx3cl1 is the direct transcriptional target of RFX4_v3 through a specific X-box (X-box1) in its promoter, the responsive element for RFX proteins. To identify the potential interacting partners for RFX4_v3, we performed yeast two-hybrid screening. Nine candidate interactors were identified, including G-protein pathway suppressor 2 (GPS2). GPS2 and RFX4_v3 are predominantly, if not exclusively, expressed in the nuclei. Both GPS2 and RFX4_v3 messages are present in most portions of the adult mouse brain, suggesting that the two proteins could bind to each other. Co-immunoprecipitation assays indicate that physical interactions between GPS2 and RFX4_v3 do indeed occur in vitro. GPS2 could also potentiate the RFX4_v3 activation on the Cx3cl1 promoter through X-box 1, suggesting the interaction is also functionally significant for RFX4_v3 activity in brain development. Based on these data, we conclude that GPS2 interacts with RFX4_v3 to modulate transactivation of a variety of genes involved in brain morphogenesis including Cx3Cl1. Examination of other potential downstream genes may lead to better understanding of the molecular mechanisms underlying the function of RFX4_v3 in brain development.
Showing the most recent 10 out of 13 publications
