“RASopathic” astrocytes constrain neural plasticity

See allHide authors and affiliations

Science  08 May 2015:
Vol. 348, Issue 6235, pp. 636-637
DOI: 10.1126/science.aab3738

Over the past decade, mutations in genes encoding RAS family members, other components of an intracellular signaling cascade that RAS controls, and proteins that modify the cascade have been recognized as causes of developmental syndromes. Collectively, these syndromes are often referred to as “RASopathies.” Not surprisingly, RASopathies have numerous manifestations, including propensity to cancer, craniofacial abnormalities, cardiac defects, cutaneous abnormalities, neurodevelopmental delay, and varying degrees of cognitive dysfunction. Uncovering the causes and developing treatments for the neurodevelopmental abnormalities are a challenge because of the myriad cellular elements in the brain and the complexity of nervous system function. A recent study by Krencik et al. (1) takes a major step toward identifying the cellular pathology underlying Costello syndrome, a RASopathy that is characterized by delayed development, craniofacial and heart problems, and cognitive impairment. The latter appears to be linked to abnormal development and function of a population of nonneuronal cells (astrocytes) in the brain.

The extracellular signal–regulated kinase/mitogen-activated protein kinase (ERK/MAPK) signaling cascade (also known as the RAS-RAF-MEK-ERK pathway) is among the most important cellular pathways, transducing effects of external signals and regulating key cellular responses such as proliferation, differentiation, and morphological development. Interestingly, most RASopathies exhibit gain of function in ERK signaling with exact clinical manifestations varying with the specific mutation (2). Dysregulation of ERK signaling in RASopathy mouse models leads to premature differentiation and overproduction of astrocytes (3, 4). Astrocytes and other glial cells, including oligodendroglia, are thought to represent more than 50% of the cellular elements in the human brain and are increasingly recognized as having important roles in regulating the development of brain circuits (5, 6).

Altering neural plasticity.

Astrocytes derived from fibroblasts of Costello syndrome patients harboring the HRASG12S mutation exhibit premature differentiation, increased proliferation, and larger size compared to controls. They also express more extracellular matrix components, resulting in the formation of perineuronal nets around interneurons (in a mouse model of Costello syndrome). Premature formation of perineuronal nets may accelerate neuron maturity and “close” the critical period of plasticity.


Costello syndrome is caused by at least 15 different mutations in the gene encoding the RAS family member HRAS (7). Disease-associated mutations cause constitutive activation of this protein. Krencik et al. explored the function of one particular mutation, HRASG12S (in which glycine at amino acid 12 is changed to serine) in gliogenesis. From dermal fibroblasts derived from Costello syndrome patients, the authors generated human induced pluripotent stem cells (iPSCs). These iPSCs were then induced to give rise to neuroepithelial cells, the precursor to both neurons and glia. When these neuroepthelial cells were exposed to epidermal growth factor and fibroblast growth factor 2 (FGF2), which act through the RAS-ERK signaling pathway, they expressed astrocytic markers indicative of glial differentiation. Fewer control neuroepithelial cells expressed astrocytic markers, indicating that mutant HRAS leads to premature astrocyte differentiation.

Astrocytes expressing HRASG12S were apparently healthy, hyperproliferative, markedly larger in size than controls, and also displayed more elaborate neural processes. The hypertrophic nature was also observed when these astrocytes were cultured on mouse hippocampal organotypic slices (a more realistic environment than dissociated cells), as well as in a mouse model in which a different Costello syndrome mutant, HRASG12V (where glycine at amino acid 12 is changed to valine), was expressed in astrocytes. These findings are in line with a recent study in the fly Drosophila melanogaster showing that FGF signaling elicits similar morphological changes in astrocytes. Importantly, astrocytes are closely associated with developing synapses (astrocytes can create a tripartite structure at neuronal synapses) (8). The implication is that hypertrophy of astrocytes could produce an enlarged brain and might account for the macrocephaly seen in many Costello syndrome patients.

Astrocytes derived from Rett syndrome and Down syndrome patients have been observed to have growth-inhibiting effects on developing neurons (9, 10). By contrast, astrocytes expressing the mutant HRASG12S secreted factors that were notably more potent than those released by control astrocytes in promoting neuronal process outgrowth and synaptic development (1). During normal mammalian brain development, astrocytes release constituents of the extracellular matrix that influence maturation of synapses and other neurodevelopmental processes. Consistent with premature astrocyte differentiation, transcripts and proteins of extracellular matrix components (such as proteoglycans and fibrous collagens) were markedly increased in astrocytes derived from HRASG12S iPSCs compared to controls. The expression of regulators and effectors of the RAS-ERK signaling pathway and regulators of synaptogenesis was also increased. The transcription factor SNAI2 is required for the increased expression of the extracellular matrix constituents.

Perineuronal net.

A perineuronal net of extracellular matrix protein (magenta) surrounds an interneuron in the visual cortex of a mouse model of Costello syndrome. The elaborate processes of astrocytes are shown in black.


A critical function of glial-secreted proteoglycans is to form “perineuronal nets” that influence neural plasticity regulated by experience (11) (see the first figure). They are found around neuronal cell bodies and neurites and stabilize established neuronal connections by restricting neuronal morphological plasticity. Importantly, Krencik et al. found that higher amounts of perineuronal nets accumulate around inhibitory neurons expressing parvalbumin in the visual cortex of the HRASG12V mouse model of Costello syndrome (compared to controls) at 14 days after birth. Parvalbumin-expressing neurons control the “critical period” in early brain development, a window of dynamic synaptic rearrangement and change in synaptic efficacy (11). Accelerated deposition of perineuronal nets around these neurons could enhance their maturation, resulting in premature critical-period closure and altered neuronal output (see the second figure). Indeed, the authors observed changes in synaptic potentials onto cortical excitatory neurons in the HRASG12V mouse model at later developmental stages. Although this study did not investigate potential alterations of brain functions, it is plausible that disrupting synaptic development during the critical period may produce long-term effects on neuronal circuits. Overall, the results of Krencik et al. suggest that “RASopathic” astrocytes may have potent effects on neural plasticity that contribute to cognitive dysfunction. A variety of standard paradigms could be used to test this idea, including formation of ocular dominance columns, tests of the physiological effects of monocular deprivation, and studies of patterning in the sensory cortex.

The results of Krencik et al. are important not only for understanding pathogenesis, but also for developing effective treatments. Owing to the importance of the RAS-ERK pathway in cancer, many inhibitors of pathway components have been developed for treating various malignancies. In mouse models of RASopathies, MEK (MAPK kinase) inhibitors reversed cardiac defects (12) and some abnormalities of brain development, including increased gliogenesis (3). The use of RAS and MEK inhibitors in a Noonan syndrome mouse model demonstrated reversibility of synaptic dysfunction without obviously altering the behavior of control animals (13). However, there are many reasons for caution, including that RAS-ERK signaling regulates higher-order processes involved in learning (14), and that other signaling pathways are likely activated downstream of pathogenic RAS (15). The implication that there is premature closure of critical periods in brain development due to hyperactive HRAS provides another plasticity paradigm that might be further explored in preclinical studies.

References and Notes

  1. Acknowledgments: W.D.S. is supported by NIH grant RO1 NS031768.
View Abstract

Stay Connected to Science

Navigate This Article