Astrocytes control the critical period of circuit wiring

One of the most extraordinary qualities of the mammalian nervous system is its ability to change with experience and throughout its life span. Mammalian brain plasticity is thought to be mainly mediated by neurons. Increased plasticity during specific windows of time during development called “critical periods” allows neuronal circuitry to be shaped. How this phase ends, however, has not been clear. On page 77 of this issue, Ribot et al. (1) show that an unsuspected cellular player—astrocytes—control when experience-dependent wiring of brain circuits is permitted in the developing primary visual cortex (V1). This finding points to possible similar roles of astrocytes or other nonneuronal cells in other neural circuits.

The primary visual cortex has long served as a model system to study brain plasticity, since the pioneering work by Hubel and Wiesel in the 1960s, when they showed that the V1 circuit is powerfully shaped by the visual experience during development (2). Their seminal studies in kittens revealed that, in response to transient eyelid closure to provoke monocular deprivation (blocking visual stimulation through one eye), the V1 circuits remodel to shift the preference of cortical neurons for the eye that remains open. This results in the so-called ocular dominance (3, 4). Notably, this influence of sensory activity on the organization of neural circuits is restricted to a critical period (4), which highlights the importance of early life experiences for the optimal functioning of the brain. Anomalous critical periods are also largely detrimental and associated with various neurodevelopmental disorders (5). Hence, how the critical period of ocular dominance plasticity is opened and closed is of fundamental importance for understanding brain development and function.

A new and fruitful development in this area of investigation has been the mouse model (6). Ribot et al. report that the ocular dominance plasticity in mice is determined by astrocytes. These nonneuronal cells have long been associated with housekeeping functions in the brain, such as regulation of the extracellular ionic environment, reuptake and recycling of neurotransmitters, and structural support (7). However, more recently, astrocytes have also been shown to control synapse formation and connectivity (8), synaptic transmission and plasticity (9), and even animal behavior (10). Ribot et al. found that grafting immature astrocytes from newborn mice in the V1 of adult mice enhanced the ocular dominance plasticity that occurred after visual stimulation of one eye. The ~200 genes differentially expressed in immature and mature astrocytes include the gene encoding connexin 30 (Cx30). Cx30 is a subunit of a gap junction channel—a specialized intercellular connection between cells. The authors observed that the expression of Cx30 in the V1 peaked approximately when the critical period for ocular dominance plasticity ended. This prompted the authors to assess plasticity in a mouse model genetically engineered to lack Cx30. Although ocular dominance plasticity peaked at about postnatal day 28 (P28) in wild-type mice, it continued to increase in mice lacking Cx30 until P50, indicating impairment in the closure of the critical period.

Electrophysiological recordings of excitatory and inhibitory synaptic transmission in cortical slices revealed that mice lacking Cx30 had reduced inhibitory transmission. Moreover, perineuronal nets were smaller in these animals. Perineuronal nets are a highly organized form of extracellular matrix that contains chondroitin sulfate proteoglycans. They tend to coalesce around inhibitory neurons (11) and are thought to contribute to the closure of ocular dominance plasticity (12). Altogether, these results indicate that astrocytes control the visual critical period by promoting the maturation of inhibitory circuits through signaling pathways that involve Cx30.

What about a relevant signaling pathway associated with Cx30? Ribot et al. discovered that Cx30 is physically associated with the protein-phosphorylating enzyme ROCK2 (Rho-associated coiled-coil–containing protein kinase 2). The expression of the small guanosine triphosphatase (GTPase) RhoA, ROCK2, and the extracellular matrix–degrading enzyme matrix metalloproteinase 9 (MMP9) were all increased by either monocular deprivation or the lack of Cx30, indicating a common signaling pathway. The authors therefore propose that astrocytes control the visual critical period by promoting the maturation of inhibitory circuits through signaling pathways that involve Cx30 and inactivation of RhoA and MMP9. This promotes the formation of perineuronal nets, the enhancement of inhibitory transmission, and the closure of ocular dominance plasticity (see the figure).

Astrocytes influence plasticity.

During development of the mammalian brain’s primary visual cortex, astrocytes regulate the so-called critical period during which plasticity allows the neural network to form. This depends on a signaling pathway controlled by connexin 30.

ROCK2, Rho-associated coiled-coil–containing protein kinase 2; MMP9, matrix metalloproteinase 9

Cx30 is a member of a large family of proteins that form intercellular channels that enable the direct transfer of ions and molecules between adjacent cells, but whether a Cx30-RhoA-ROCK2 signaling pathway involves ion and molecule permeation into astrocytes remains unknown. Moreover, several human deafness diseases have been associated with Cx30 mutations (13). It is unknown whether any changes in critical-period plasticity are found in these patients. Notably, astrocytes in the fruit fly Drosophila melanogaster regulate the maturation of the motor circuit and are essential for proper critical-period closure (14). In this case, interaction between the cell adhesion proteins neuroligin and neurexin is the likely signaling pathway. Thus, there may be a diversity of molecular and signaling pathways in which astrocytes influence the use-dependent plasticity of neural circuits during development.