XB-ART-46722Dev Dyn June 1, 2013; 242 (6): 604-13.
Expression of Ski can act as a negative feedback mechanism on retinoic acid signaling.
Retinoic acid signaling is essential for many aspects of early development in vertebrates. To control the levels of signaling, several retinoic acid target genes have been identified that act to suppress retinoic acid signaling in a negative feedback loop. The nuclear protein Ski has been extensively studied for its ability to suppress transforming growth factor-beta (TGF-β) signaling but has also been implicated in the repression of retinoic acid signaling. We demonstrate that ski expression is up-regulated in response to retinoic acid in both early Xenopus embryos and in human cell lines. Blocking retinoic acid signaling using a retinoic acid antagonist results in a corresponding decrease in the levels of ski mRNA. Finally, overexpression of SKI in human cells results in reduced levels of CYP26A1 mRNA, a known target of retinoic acid signaling. Our results, coupled with the known ability of Ski to repress retinoic acid signaling, demonstrate that Ski expression is a novel negative feedback mechanism acting on retinoic acid signaling.
PubMed ID: 23441061
Article link: Dev Dyn
Grant support: Canadian Institutes of Health Research
Genes referenced: ski actl6a cyp26a1 gapdh id3
Article Images: [+] show captions
|Fig. 1. Endogenous expression of ski in Xenopus. Ski expression was first detectable in a ring around the blastopore at stage 10/11. At stage 20, strong expression was observed in bands in the forming head in the anterior (ant.) view There is also staining on the ventral mid- line immediately posterior to the cement gland at that stage. In a dorsal view, strong staining can be seen in the anterior spinal cord (black arrow) and the staining can be seen even fur- ther posterior at stage 22. Strong staining was also observed in the migrating anterior neural crest (red arrowheads). At tail bud stages (St. 30) low levels of expression are seen over much of the embryo including somites and higher expression was still viewed in discrete regions including the otic vesicle (yellow arrow).|
|ski (v-ski sarcoma viral oncogene homolog) gene expression in Xenopus laevis embryo, assayed via in situ hybridization, NF stage 11, vegetal view, dorsal up.|
|ski (v-ski sarcoma viral oncogene homolog) gene expression in Xenopus laevis embryo, assayed via in situ hybridization, NF stage 20, anterior view, dorsal up.|
|ski (v-ski sarcoma viral oncogene homolog) gene expression in Xenopus laevis embryo, assayed via in situ hybridization, NF stage 25, lateral view, anterior left, dorsal up.|
|ski (v-ski sarcoma viral oncogene homolog) gene expression in Xenopus laevis embryo, assayed via in situ hybridization, NF stage 30, lateral view, anterior left, dorsal up.|
|Exposure to retinoic acid (RA) causes an increase in ski expression. A: When Xenopus embryos were exposed to 1mM retinoic acid from stage 14 to 20, stage 20 to 26, or from stage 26 to 32, there was an increase in ski expression as assayed by whole-mount in situ hybridization. If embryos were exposed to 1 mM retinoic acid antagonist expression of ski was reduced. This was particularly evident in the pharyngeal region (red arrows). B: Quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) for expression of ski in Xenopus embryos treated as mentioned previously between stages 140. Ski is significantly up-regulated in response to RA and down-regulated in embryos treated with RA antagonist. C: Positive control quantifying expression of downstream RA signaling target cyp26a1 in response to RA or RA antagonist treatment showing the expected increase and decrease (respectively) in expression. qRT-PCR is normalized to h4 mRNA levels. mRNA levels were normalized to the h4 housekeeping gene. Error bars represent standard error of the mean. n 1⁄4 5. *P < 0.05 **P < 0.01.|
|Fig 3. Different regions of the embryo responded differently to changes in retinoic acid (RA) signaling. Addition of RA to embryos resulted in a general increase in staining by in situ hybridization, suggesting that the most cells are responding to the increased signaling. There was a much greater response at specific sites. In particular, there is a gap in staining in the neural tube (black arrowheads) roughly corresponding to the hindbrain. Addition of RA resulted in that region showing very strong expression of ski as assayed by in situ hybridization. The posterior spinal cord (blue arrow) does not normally have strong expression of ski but does so when RA is added. When embryos are treated with an RA antagonist, the strong staining in the brain (red arrow) and spinal cord (yellow arrows) is much reduced. Note that staining in the somites is also markedly reduced in the antagonist treated embryos. DMSO, dimethyl sulfoxide.|
|Fig. 5. Retinoic acid (RA) signaling directly alters ski expression in vivo. A: Whole-mount in situ hybridization of embryos pretreated with 20 min of 1 mM cycloheximide to block translation followed by 1 hr treatment with 1 mM RA. In a dorsal view, embryos treated from St. 146 showed increased staining with RA even in the presence of cycloheximide. Embryos treated from St. 268 (side view) showed the same result. B: Quanti- tative reverse transcriptase-polymerase chain reaction (qRT-PCR) of similarly treated embryos between stages 14 and 16 showed that ski expres- sion is up-regulated by RA even in the presence of cycloheximide. C: qRT-PCR analysis of cyp26a1, a known direct target of RA signaling, was used as a positive control for the RA and cycloheximide treatments. Error bars represent standard error of the mean. n1⁄45. *P < 0.05, **P < 0.01, ***P < 0.001.|
|Fig. 7. Overexpression of both Ski and the Ski-APRG mutant causes decreased expression of CYP26A1. A: Western blot (anti-FLAG) confirming protein synthesis from the Ski and Ski-APRG constructs after transfection into HaCaT cells. B: Quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) quantifying the levels CYP26A1 mRNA shows a decreased level of expression after transfection with either Ski construct. C: qRT-PCR of similarly transfected cells quantifying the levels of ID3 mRNA, a target of bone morphogenetic protein (BMP) signaling, showing that there is decreased levels of expression with Ski over expression but not with over expression of the Ski-APRG construct. Error bars represent standard error of the mean. mRNA levels were normalized to the GAPDH mRNA levels. n 1⁄4 3. *P < 0.05, **P < 0.01.|