While microRNAs have already been established as key players in a host of regulatory and developmental processes, a new study from researchers at Ohio State University has linked a single one of these small, non-coding RNAs to proper embryonic tissue segmentation.
According to the study’s senior author, Susan Cole, the findings were particularly interesting because “we were able to look at a single microRNA/target pair, and [discovered that] interrupting that specific … interaction could have a profound phenotypic effect.”
Cole’s research, which was detailed in Developmental Cell, focused on somites — the embryonic precursors that give rise to the axial skeleton and skeletal muscles, and establish the segmental vertebrate body plan. The study examined how the development of somites is controlled by the oscillating expression of genes including one called Lunatic fringe, or Lfng, which encodes a glycosyltransferase that modulates the Notch signaling pathway.
“This is a situation where embryos need to … coordinate temporal information and spatial information to create somites,” she explained to Gene Silencing News.
During somitogenesis, the expression levels of Lfng and other genes oscillate in the presomitic mesoderm, turning on and off as part of a “segmentation clock” that controls the somite formation.
“You can think of it as analogous to a circadian oscillation, but really fast,” Cole said, noting that in humans it occurs every five hours, in chickens every 90 minutes, and in mice every two hours.
Previous studies have shown that the loss of Lfng expression or its sustained, non-oscillatory activity “perturbs somite formation and patterning, presumably by altering its oscillatory expression,” the study’s authors wrote in Developmental Cell.
And while it is known that cyclic Lfng expression is regulated at the transcriptional level, little is known about the post-transcriptional mechanisms that drive its oscillations.
“There has been a lot of mathematical modeling of this clock,” Cole said. “What many models have suggested is that the RNA turnover of all these oscillatory transcripts has to be very tightly regulated.” Importantly, even if Lfng expression is turned off, if its RNA transcript remains, the oscillatory activity would be lost.
“So we knew there must be mechanisms to destabilize that transcript,” Cole said.
At the same time, others have demonstrated that the Lfng 3’ UTR is evolutionarily conserved and suggested that it regulates RNA half-life, according to the paper. And given the role of miRNAs in directing post-transcriptional repression of genes by “promoting RNA turnover and/or by decreasing translational efficiency of their target transcripts,” Cole and her colleagues hypothesized that miRNAs might be controlling the gene’s oscillatory expression in the segmentation clock.
To test this, they examined evolutionary conservation of the Lfng 3’ UTR across vertebrates, looking for conserved miRNA-targeting sites. They then assessed the expression of candidate miRNAs in the presomitic mesoderm of mice.
One in particular, miR-125a-5p, was expressed at higher levels in these cells than in mature somites, indicating that its expression is enriched in the presomitic mesoderm where Lfng requires a short RNA half-life.
After conducting experiments in mouse and chicken cells to show that the miRNA could target the Lfng 3’ UTR, the team then turned to in vivo testing.
They introduced morpholino-based miR-125a-5p antagonists into chick embryos and found that, 24 hours after electroporation, they found that regions where the miRNA had been inhibited displayed severely disrupted formation and patterning of mature somites.
“Interestingly, phenotypes were observed even in cases where electroporation levels were comparably low,” the researchers wrote. “This is unlikely to be due to non-cell-autonomous effects of the morpholino. Instead, it reflects one of the roles of the segmentation clock, which is to synchronize the oscillations of neighboring cells.
“Mosaic regions of the embryo containing wild-type cells mixed with cells that have altered clock function are predicted to exhibit phenotypes at the tissue level due to lack of cell:cell synchronization,” an effect recently confirmed in mouse embryos chimeric for wild-type and Lfng-null cells, they added.
Similar phenotypes were observed when a target protector was used to prevent the binding of endogenous miR-125a-5p binding.
“Together, these findings indicate that miR-125a-5p activity is required for normal formation and patterning of epithelial somites.”
Additional experimentation confirmed that miR-125a-5p is required for the normal cyclic expression of Lfng, and that the interaction between the two is required for normal somitogenesis and segmentation clock function via a feedback loop that affects the transcription of other clock components.
Cole noted that all vertebrates have this clock, but it is unclear whether the way it functions in chickens holds up in other species, namely mammals. As such, she and her lab are currently examining the regulatory process in mice.
“The possibility that the clock function of mir-125a-5p might not be conserved between mouse and chicken would not be surprising given that different organisms can utilize completely distinct sets of protein components in their segmentation clocks,” the researchers noted in Developmental Cell. “Thus, it is possible that different miRNA/transcript pairs are important in the mouse segmentation clock, or that regulation occurs via translational efficiency rather than through effects on transcript stability.”
“Testing of these models will require targeted mutation of the mir-125a-5p binding sites in the Lfng 30 UTR at the endogenous locus, to examine whether these binding sites are required for normal mouse segmentation,” they wrote. “However, it is attractive to hypothesize that different mechanisms of post-transcriptional clock regulation could contribute to the differences in clock period observed among distinct vertebrate species.”
While the findings are very early stage, Cole suggested that her work could prove clinically relevant, noting that perturbations of the segmentation clock have been linked to congenital birth defects of the skeleton, while recent papers have proposed that minor perturbations could contribute to the curvature of the spine seen in scoliosis.