|
Figure 1. An extremely conserved region in c-fos first intron: evidence for a nested promoter.A. c-fos gene conservation profile between the mouse gene and 6 other indicated species (VISTA genome server). The baseline corresponds to 50% identity for human, pig and chicken genomes, and to 0% identity for xenopus, fugu, and zebrafish genomes in 100 bp windows. Peaks culminating at more than 70% identity (in 100 bp windows) are painted in pink by the software. Note that the 3â² half of first intron is more conserved than surrounding exons in the four top species. Confirmed start site for canonical mRNA and putative intronic mRNA appear respectively as blue and black broken arrows. A CpG island (green box) extends from the canonical promoter to the putative intronic promoter. Blue boxes represent exons, black mRNAs correspond to the putative mRNA starting in the conserved region and three ESTs previously sequenced and mapping to the same region (UCSC genome server). DNase I hypersensitive sites from Renz et al. are shown as black downward arrows. B. Nucleotide alignment of the most conserved part of c-fos first intron (from +619 to +849 relative to mouse canonical TSS) from 5 species (pig, human, mouse, chicken, xenopus). Asterisks depict nucleotides identical in all 5 species. Conservation culminates in motifs resembling a TRE, a CRE, and a TATA box (shown in blue). Red and green sequences respectively correspond to primers 1 and 2 used in figure 6.
|
|
Figure 2. The intronic TRE and CRE sites bind members of the CREB and AP-1 families of transcription factors.EMSA using HeLa nuclear extracts and probes corresponding to the TRE and CRE sites of c-fos first intron. A. Competition analysis with non-radioactive probes wt (lanes 3, 4) or mutated on the TRE (lanes 5, 6), the CRE (lanes 7, 8), or both (lanes 9, 10). Lane 1 shows the probe without extract, lane 2 with extract but without competing cold probe. B. Disruption or supershift of the protein/DNA complexes. The probe (lane 1) was incubated with HeLa nuclear extract (lanes 2 to 9) and antibodies to c-Fos (lane 3), c-Jun (lane 4), ATF-a (lane 5), ATF-1 (lane 6), JunB (lane 7), CREB-1 (lane 8), and CREM (lane 9).
|
|
Figure 3. The c-fos conserved intronic region drives luciferase expression and responds to the CREB and AP-1 pathways.A. 5â² part of mouse c-fos gene (up) and our reporter construct (bottom), fiL (fos intron Luciferase). B. The fos intronic region suffices to drive Luciferase activity in NIH3T3 cells in transient transfection. The promoterless pGL2 basic vector background activity is shown as a control. Error bars represent standard deviation, nâ=â7. C. The fIL construct responds to the calcium and cAMP pathways, but not to PMA. NIH3T3 transfected with fiL were treated with the indicated drugs. Error bars correspond to standard deviation, nâ=â6. D. Co-transfection with CREB or AP-1 expression vectors stimulates fIL promoter activity. Error bars correspond to standard deviation, nâ=â5. Statistical analysis of variance (ANOVA) was performed (*: p<0.05; **: p<0.01 relative to control).
|
|
Figure 4. Transgenic analysis of the putative promoter shows expression restricted to the spinal cord and mammary bud of mouse embryos.A. 5Ⲡpart of mouse c-fos gene (up) and our NLS-containing, betagalactosidase reporter construct (bottom), fiZ (fos intron lacZ). 7 Transgenic mouse lines were created with fiZ and transgenic embryos were stained for betagalactosidase activity. Here are shown transgenic embryos from mouse line #60 at day 10.5 (B), 11.5 (C, K), 12.5 (D, G, H), and 13.5 (F, I, J). B, C, D, F: whole-mount embryos showing the spinal cord staining starting day 11.5 p. c. and the mammary gland anlage staining starting 12.5 p. c. G. Close-up view of the developing mammary buds showing the stained ring corresponding to mesenchymal cells. H, I, J: sagittal frozen sections showing the nuclear staining in the mesenchymal part of the mammary bud, but not in the central, epithelial part. K. Transverse frozen section on a 11.5 d. p. c. embryo showing the extremely restricted, ventral spinal cord staining. E: wild-type, e13.5 embryo as a control. Scale bars: 1 mm (B, C, D, E, F, G) then 150 µm (H, I, J, K).
|
|
Figure 5. Brain-specific but variegated expression in newborn and adults from three different transgenic mouse lines.Newborn pups (A, D, G) or adult brains (B, C, E, F, H, I) from mouse transgenic lines 3 (A, B, C), 12 (D, E, F), and 21 (G, H, I) were stained for betagalactosidase activity, along with wild-type controls (J, K). Newborn pups (A, D, G, H) and adult brains (C, F, I, K) were cut sagitally or transversely, respectively, for X-Gal penetration and structure identification. Note the different patterns of expression in different mouse lines. Asterisks show a territory of endogenous, cytoplasmic betagalactosidase expression, the nasal pits, also seen in the wild-type controls (J). Scale bars: 5 mm (A, B, D, E, G, H, J) or 1 mm (C, F, I, K).
|
|
Figure 6. Validation of the expression profile and start site region of c-fos intronic mRNA by RT-QPCR.A. Schematic representation of the primers used in this RT-PCR study. Boxes represent exons, larger boxes are translated regions, arrows depict primers. The 3â² PCR primer was always RT1. 5â² primers in grey lie upstream the putative TSS, whereas primers in black are downstream. B. RT-PCR mapping of the c-fos intronic mRNA start site region. RNA extracted from E12.5 mouse embryo mammary buds was reverse transcribed with primer RT1, then cDNAs were amplified with primers used for detection of the intronic mRNA (primer 2 lane d, primer 3 lane e) and canonical pre-mRNA (primer 1, lane f). Minus RT control ensures that no genomic DNA contaminant was amplified with any primer (-RT: lanes a, b, c), while PCR on genomic DNA (genDNA: lanes g, h, i) shows that primer 1 is PCR-competent (lane i). The DNA ladder (last lane) shows fragments of 600, 500, 400, 300, and 200 bp. nâ=â4. C. Quantitative PCR-mediated mapping of the intronic mRNA start site on cDNA from adult mouse cortex. The five different 5â² primers described earlier were used along with RT1 for amplification. Signal was normalized according to standard curves elaborated for each primer pair, allowing a measure of the targeted RNA independently of primer efficiency. The scale shows the amount of RNA in fg per 10 µl of reaction. Error bars represent standard deviation, nâ=â4. D. RT-QPCR analysis of fos intronic mRNA expression in adult tissues. Quantitative PCR was ran with primers 2 and RT1 on cDNA prepared from dissected candidate adult mouse tissues (from left to right: spinal chord, mammary gland, cerebellum, cortex, skin). Normalization was done over the expression of beta-2 microgobulin. Error bars represent standard deviation, nâ=â4. Statistical analysis of variance (ANOVA) was performed relative to primer 0.8 (C) or to spinal chord level (D). *: p<0.05; **: p<0.01.
|
