Abstract
Small non‐coding miRNA appear to be vital in brain development and function by organising complex patterns of gene expression. These molecules are important for the regulation of synaptically localised mRNAs that encode proteins involved in neurotransmission and behaviour. In this issue of EMBO Reports, Lackinger et al 1 demonstrate that a large cluster of miRNAs, that emerged in placental mammals, functions as a repressor of social behaviour. This discovery has significant implications for our understanding of the brain, behaviour and in particular psychiatric syndromes, which have been shown to display alterations of these molecules.
Subject Categories: Neuroscience, RNA Biology
Post‐transcriptional regulation of gene expression is a significant feature of metazoans that enabled them to develop complex cellular architectures and multicellularity. While this evolutionary advance is supported by an array of RNA binding proteins, the emergence of small trans‐acting RNAs, which function as nucleic acid “guide” sequences, was a very significant innovation. These molecules, particularly miRNA, increased their numbers and influence in vertebrates to regulate even greater levels of complexity. The reliance on miRNA for higher functions continued to grow in mammals and particularly in eutherians that are developmentally nurtured in the maternal uterus via the placenta. A large expansion of miRNA genes occurred in placental mammals at a single locus (14q32 in humans). This region encodes an enormous set of 38 miRNAs known as the miR379‐410 cluster. These are thought to have a diverse array of functions, but they have already featured strongly in neural activity and neurobehavioural phenotypes, in particular miR‐134, which is one of the best‐characterised brain‐associated miRNAs. miR‐134 was shown previously by Schratt et al 2 to be enriched in the synapto‐dendritic compartment of rat hippocampal neurons and to negatively regulate dendritic spine development. This landmark study has also shown that this response is mediated by the miR‐134 target Limk1, a protein kinase. Later, part of the same team showed that transcription of the entire miR379‐410 cluster was regulated by synaptic activity, which is associated with changes in Mef2 transcription factor binding in rat neurons 3. This suggested that the regulation of the entire cluster might be highly coordinated with several members participating in neural function. When the locus was deleted in a mouse knockout model, the loss of miR379‐410 expression was associated with abnormal emotional response, including anxiety in unfamiliar environments 4. The deletion had no apparent impact on cognitive function, nor did it have any impact on social behaviour in sociability tests. This lack of change in social behaviour was contrary to the expectation of the investigators, as the miRNA cluster resides within a paternally imprinted domain. This directs monoallelic expression in a parent of origin‐specific manner, which has been suggested previously to be important for regulating social behaviour 5.
Now, Lackinger et al 1 use a similar model to further investigate the neural and behavioural function of the miR379‐410 cluster. Because the paternal allele is silenced by imprinting, it was only necessary to produce heterozygous knockout of the maternal allele to achieve complete loss of expression. The authors tested a range of behavioural phenotypes and, in contrast to Marty et al 4, observed changes in social behaviour. Firstly, knockout mouse pups separated from their mothers showed increased ultrasonic vocalisations. This communication stimulates maternal search/retrieval behaviour and is attributed to social function. Secondly, juvenile knockout mice spent more time in reciprocal interactions and made more ultrasonic vocalisations compared to their wild‐type littermates. Finally, a three‐chamber test was also performed to examine social approach to a “stranger” mouse over an inanimate object. Again, the preference for the stranger was significantly elevated in adolescent knockout mice. All these behaviours suggest enhanced social functions in knockout mice 1.
These phenotypes were supported at the cellular level by an increase in dendritic spine density and increased frequency of miniature excitatory post‐synaptic currents (mEPSC) in hippocampal neuron cultures from knockout mice. Also, dendritic spine volume and mEPSC amplitude were reduced, collectively suggesting that mutant animals have increased excitatory synaptic transmissions 1. To explore the impact of miR379‐410 knockout at the molecular level, RNA extracted from isolated hippocampi were subjected to RNAseq and compared to tissue from the wild‐type littermates. miR379‐410 targets and pathways were analysed, revealing gene ontology clusters associated with glutamatergic and GABAergic ionotropic receptor complexes 1. miRNA motif analysis highlighted a group of 5 miRNA from the cluster with enrichment of modulated target genes. Interestingly, the seed sequences in these five miRNAs shared some sequence homology, suggesting that there is substantial functional overlap and possibly convergence in the most active members of the cluster in this tissue. Several of these up‐regulated target genes were further validated by qPCR, including Cnih2, Dlgap3, Prr7, Src, Lzts2, Mpp2, Shank1, Shank3 and Shb. Putative miRNA interactions with a number of the ionotropic glutamate receptor genes were also confirmed by reporter gene assays in transfected primary cortical neurons in vitro.
In summary, Lackinger et al provide strong support for the hypothesis that the miR379‐410 cluster, or at least five key miRNAs, are involved in repressing neurotransmission, perhaps (among other targets) through the attenuation of post‐synaptic ionotropic glutamate receptor translation (Fig 1). This seems to be important for regulating social behaviour, as the knockout animals displayed hyper‐social phenotypes, which could have implications for several psychiatric disorders. Copy number variation in this imprinted locus has been reported in the literature and was found associated with significant physical, cognitive and other behavioural phenotypes 6. We also observed changes in the expression of miRNAs from the miR379‐410 cluster in schizophrenia in both post‐mortem cortex and peripheral blood mononuclear cells (PBMCs) 7, 8. While these alterations were not associated with any specific genetic anomaly in PBMCs, enrichment of miR379‐410 miRNA was associated with environmental exposures in rats, including maternal immune activation and adolescent cannabinoid treatment 9. Perhaps even more important, ASD and Angelman syndrome with their social avoidant and hyper‐engagement phenotypes, respectively, are strong candidates for an involvement in this regulatory system. In this regard, the miR379‐410 cluster has been implicated previously with these syndromes through Ube3A1, a long non‐coding transcript variant of the ubiquitin ligase 3A, which is up‐regulated in autism spectrum disorders and maternally deleted in Angelman syndrome 10. Paradoxically, the transcript variant was shown to function as an efficient competing endogenous RNA (ceRNA) capable of sponging a large number of miRNA from the miR379‐410 cluster in dendrites of rat hippocampal neurons 10. Therefore, it is not clear how the loss of expression of this expression in Angelman syndrome, leading to a rise in functional miRNA, is consistent with the knock out model used here, as both conditions are characterised by hyper‐social behaviour. Similarly, the elevation of Ube3A1 in ASD also seems to be inconsistent with a reduction in the propensity for social interactions. Finally, Marty et al 4 did not observe any difference in terms of sociability between maternal miR379‐410 knock out animals and wild‐type littermates, although there were methodological differences, which could account for these inconsistencies. There is no doubt that more work will be needed to resolve these questions completely.
Figure 1. Excitatory pyramidal neurons of the hippocampus under the influence of the miR379‐410 microRNA cluster.

The human chromosome 14 is shown inside the nuclear membrane, with the region 14q32.2 expressing the polycistronic primary miRNA transcript Mirg. After cleavage by the microprocessor complex, the precursor miRNA “hairpins” from the cluster are exported into the cytoplasm and interact with proteins in the RNA‐induced silencing complex and their target transcripts. This complex is then able to recruit the CCR4_NOT complex promoting deadenylation (Pacman motif Y), decapping and degradation (Pacman motif X) of mRNAs via the exonuclease XRN1. This prevents the synthesis of target transcripts that encode the ionotropic glutamate receptor complex, leading to a reduction of synaptic neurotransmission and connectivity in circuits facilitating social behaviour. When the expression of the miR379‐410 microRNA cluster is reduced in psychiatric syndromes, or in a knockout mouse model, target transcripts escape post‐transcriptional gene regulation and are available to bind ribosomes and template protein synthesis, including the subunits of the ionotropic glutamate receptor complex. These become inserted into the post‐synaptic density and enhance synaptic neurotransmission.
We are just beginning to understand the powerful role of post‐transcriptional mechanisms in behavioural neuroscience, with tiny RNAs welding more influence over our thinking than anyone would have imagined. These molecules have tremendous functional independence and can curate translation at the most distal synapse to facilitate input restricted and activity‐dependent remodelling of neural circuits 11. We may never fully comprehend the complexity that can be managed in this way, but we must continue to explore and seek opportunities for intervention, as it is clear (as exemplified by the work of Lackinger et al) that these mechanisms are extremely important for the brain, behaviour and psychiatric syndromes.
Acknowledgement
MC is supported by an NHMRC Senior Research Fellowship (1121474).
EMBO Reports (2019) 20: e47663
See also: M Lackinger et al (February 2019)
References
- 1. Lackinger M, Sungur AO, Daswani R et al (2019) EMBO Rep 20: e46429 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Schratt GM, Tuebing F, Nigh EA et al (2006) Nature 439: 283–289 [DOI] [PubMed] [Google Scholar]
- 3. Fiore R, Khudayberdiev S, Christensen M et al (2009) EMBO J 28: 697–710 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Marty V, Labialle S, Bortolin‐Cavaille ML et al (2016) Hum Mol Genet 25: 728–739 [DOI] [PubMed] [Google Scholar]
- 5. Isles AR, Davies W, Wilkinson LS (2006) Philos Trans R Soc Lond B Biol Sci 361: 2229–2237 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Zollino M, Seminara L, Orteschi D et al (2009) Am J Med Genet A 149A: 1116–1124 [DOI] [PubMed] [Google Scholar]
- 7. Beveridge NJ, Gardiner E, Carroll AP et al (2010) Mol Psychiatry 15: 1176–1189 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Gardiner E, Beveridge NJ, Wu JQ et al (2012) Mol Psychiatry 17: 827–840 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Hollins SL, Zavitsanou K, Walker FR et al (2014) Transl Psychiatry 4: e452 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Valluy J, Bicker S, Aksoy‐Aksel A et al (2015) Nat Neurosci 18: 666–673 [DOI] [PubMed] [Google Scholar]
- 11. Kiltschewskij D, Cairns MJ (2019) Nucleic Acids Res 47: 533–545 [DOI] [PMC free article] [PubMed] [Google Scholar]
