Insights into the biological functions of microexons

The recent finding that dozens of highly conserved tiny microexons exist in our genes and show striking neuronal-specific regulation suggests that microexons play crucial roles in brain development and function. However, these roles are largely unknown. By knocking out the major regulator responsible for the neuronal-specific regulation of microexons, nSR100/SRRM4, we obtained insights into the biological functions of microexons in mammalian brain development. Misregulation of microexons is associated with a plethora of nervous system defects, including neurite outgrowth, cortical layering, timing of neuronal differentiation, and axon guidance. Strikingly, the inclusion of a single six-nucleotide microexon is able to rescue neurite growth defects of cultured knock-out neurons.

HFSP Long-Term Fellow Manuel Irimia and colleagues
authored on Thu, 01 October 2015

In a recent report (Irimia et al. Cell 2014), we discovered hundreds of microexons (exons as short as 3-27 nucleotides) in mammalian genomes. This finding alone came as a big surprise, since exons – the bits of ‘meaningful’ sequence that form the messenger RNA from genes – need to be recognized as such by the cell machinery within a sea of intronic sequence, they were believed to be relatively long. However, an even bigger surprise was that the majority of microexons have a strong tendency to be expressed only in the neurons, and that this regulatory pattern is highly conserved across vertebrate evolution. Furthermore, microexons are not distributed randomly among or within coding regions of genes: they tend to fall in the surfaces of interaction domains of proteins that play crucial roles in neuronal differentiation and function (including neuritogenesis, synapse structure, vesicle biology, etc.). Therefore, they can modulate how these important proteins interact with each other specifically in neurons. Finally, microexons are significantly misregulated in the brains of some autistic patients, and affect genes such as SHANK2, CASK or CADPS that were previously linked genetically to autism spectrum disorder. However, despite these multiple lines of evidence suggesting microexons play crucial functions in mammalian nervous systems, a more direct proof for their biological importance was still missing.

Figure: Coronal cortical section of a nSR100 knockout mouse at E18.5 stage. Red=Tbr1+ (early born neurons), Green=Satb2+ (Tbr1 negative-Satb2 positive are late-born neurons), Blue=DAPI. Credit: Mathieu Quesnel-Vallières.

A recent study published in the journal Genes & Development provides this proof. In our previous study, the neuronal-specific splicing factor SRRM4/nSR100 was identified as a major direct regulator of the inclusion switches of microexons during neuronal differentiation. Remarkably, not only does nSR100 regulate the majority of neural microexons, but it also shows a striking specificity for microexons and other relatively short exons. Therefore, to shed light on the functions of this splicing factor as well as on the roles of microexons in vivo, Mathieu Quesnel-Vallières (main author of this study, under the co-supervision of Dr. Blencowe and Dr. Cordes) set out to knock out (KO) nSR100 in mice.

As expected, comparative RNA-seq analysis of E18.5 cortex and hippocampus from KO and wild type mice showed that dozens of neural microexons were misregulated in vivo after nSR100 depletion. In fact, microexons and other short exons (<50 nucleotides) accounted for the vast majority of misregulated alternative splicing events. Furthermore, nSR100-deficient mice displayed several phenotypic alterations. First, most homozygous KO animals died soon after birth, due to breathing defects. Suggestively, severe defects in neurite outgrowth of motoneurons innervating the diaphragm were observed, involving reduced number and length of secondary branches. Moreover, these were not specific to phrenic nerve innervations, but were also observed in other nerves. Second, homozygous (and to a lesser extent heterozygous) mutants displayed aberrant cortical layering, characterized by an increase in the number of early born neurons (a thicker layer VI) and a reduction in late born neurons. This was accompanied by a reduction in the pool of Pax6-positive progenitors and total number of differentiated NeuN-positive neurons. Pulse-labeling of early-born neurons further revealed that the timing of neurogenesis was impaired in mutant mice. Third, nSR100 ablation caused a large fraction of axons to fail crossing the midline in the corpus callosum, highlighting for the first time a role for a splicing factor in axon guidance. Finally, primary cultures of neurons showed reduced neurite lengths in the KO cells.

However, despite the remarkable neurodevelopmental defects in KO mice and that these mice show a strong misregulation of microexons, more direct evidence for the involvement of microexons in these processes was needed. One of the most useful experiments to demonstrate the importance of specific targets in the phenotype(s) observed in a KO of a regulatory gene is to perform a rescue experiment of that phenotype with single targets. Therefore, we set out to test whether the inclusion of a misregulated six-nucleotide neural microexon in the mRNAs of Unc13b, a gene that has been previously involved in neuritogenesis, could rescue the growth defects of neurites observed in primary cultures. For this, KO and wild type neuron cultures were transfected with two different Unc13b constructs, only diverging in the inclusion or exclusion of the six-nucleotide microexon. Strikingly, whereas the KO neurons expressing the isoform without the microexon showed the same neurite length as sham-transfected KO neurons, the neurons expressing the inclusion isoform showed the same neurite length as the wild type neurons, thus fully rescuing the growth defect. That is, a difference of only two residues in the Unc13b protein has a dramatic functional impact. In conclusion, our study provides strong evidence that microexons play crucial roles in vivo.


[1] Essential roles for the splicing regulator nSR100/SRRM4 during nervous system development. Quesnel-Vallières, M., Irimia, M., Cordes, S.P., Blencowe, B.J. 2015. Genes Dev, 29:746-59.

Other references

[2] A highly conserved program of neuronal microexons is misregulated in autistic brains. Irimia, M., Weatheritt, R.J., Ellis, J., Parikshak, N.N., Gonatopoulos-Pournatzis, T., Babor, M., Quesnel-Vallières, M., Tapial, J., Raj, B., O’Hanlon, D., Barrios-Rodiles, M., Sternberg, M.J.E., Cordes, S.P., Roth, F.P., Wrana, J.L., Geschwind, D.H., Blencowe, B.B. 2014. Cell, 159:1511-23.

Pubmed link [1]

Pubmed link [2]