The eye tells the brain when to plasticize

The classic Nobel Prize-winning studies of David Hubel and Torsten Weisel showed how the proper maturation of the developing visual cortex is critically dependent upon visual information received from the eyes. In what would today be considered highly unethical experiments, Hubel and Weisel sewed shut one eye of newborn kittens. They found that this monocular deprivation had dramatic effects on the visual part of the brain: the columns of cortical tissue that normally receive inputs from the closed eye failed to develop, while those that receive inputs from the other eye were significantly enlarged. The kittens also failed to develop visual cortical areas which normally receive inputs from both eyes, and as a result did not have binocular vision.

The work also showed that this could be reversed, as long as the deprived eye was opened within a specific time window. Thus, the concept of the critical period was established, a concept which remains infulential to this day. These findings have important implications for a human condition called amblyopia, or "lazy eye", in which vision in one eye is poor during early childhood, due to, for example, cataracts or short- or long-sightedness. The child's visual system develops abnormally, in exactly the same way as Hubel and Weisel's one-eyed kittens, unless the condition is detected early on. It is easily corrected by temporarily covering the good eye, and so forcing use of the "lazy" eye, such that the visual inputs and neuronal circuitry in the visual cortex develop as they should.

Since the work of Hubel and Weisel, the visual cortex has become the best characterised example of experience-dependent synaptic plasticity, and modern techniques have led to some understanding of the cellular mechanisms by which visual experience induces changes in the brain. Under normal circumstances, the pathway from eye to brain, like other parts of the nervous system, forms an overabundance of synaptic connections. Light entering the eye activates cells in the retina, which generate nervous impulses that are propagated, via the lateral geniculate nucleus in the thalamus, to the visual cortex at the back of the brain. This neuronal activity is essential for strengthening those connections which have formed correctly and for pruning or completely eliminating inappropriate connections.

A new study, published today in the journal Cell, provides further insight into plasticity in the visual cortex, and reveals a surprising new mechanism by which changes in the connectivity of visual cortical circuitry are induced. It shows that a protein which is involved in earlier stages of neural development, and which is usually thought of as being restricted to the nucleus of nerve cells, can regulate plasticity after being transferred, through at least two synapses, from cells in the retina to neurons in the visual cortex. The discovery of this novel mechanism, which was previously thought not to occur in the nervous system, could eventually lead to therapies for various conditions of impaired vision.

The work was led by Takao Hensch of the Laboratory for Neuronal Circuit Development at the RIKEN Brain Science Institute in Japan. In recent years, Hensch and his colleagues have focused their work on the role of inhibitory interneurons in visual cortical plasticity. These cells do not project to other parts of the cortex, but instead have local projections which remain entirely within circuits, and which release the inhibitory neurotransmitter gamma-amino butyric acid (GABA). The interneurons are involved in a process called lateral inhibition, whereby activity in one cell leads to decreased activity in adjacent cells.

The work from Hensch's lab, and from several other labs, shows that inhibitory interneurons have a central role to play in the development of visual cortical circuitry. Lateral inhibition is crucial for the development of ocular dominance columns, those regions of the visual cortex which receive inputs from the left or right eye, and whose development is affected by monocular deprivation. Mutant mice lacking the enzyme GAD65, which is required for the synthesis of GABA, do not exhibit experience-dependent cortical plasticity. The critical period for ocular dominance column development is also delayed in these animals. Furthermore, infusion of a benzodiazepene (an anti-anxiety drug which activates GABA receptors) into the kitten visual cortex leads to an approximately 30% increase in ocular dominance column width, and bring the critical period forward; conversely, infusion of a GABA receptor blocker reduces columnar width.

Thus, plasticity in the visual cortex and the timing of the onset of the critical period is regulated by the inhibitory interneurons. So too are the decline of the critical period and the age-related reduction of plasticity. One type of inhibitory interneuron, called the basket cell, projects a horizontal plexus of axons which spans multiple ocular dominance columns in such a way as to shape their spacing during development. With age, the basket cells become enveloped in a network of extracellular matrix molecules, a process coincides with the decline of plasticity. Plasticity involves the reorganization of dendritic spines, the tiny finger-like projections on excitatory neurons which receive synaptic inputs from other cells; a neuron can sprout new spines in response to activity, or it can eliminate old ones in the absence of activity. The network of matrix molecules thus inhibits plasticity by reducing spine motility, and plasticity can be reactivated by treatment with an enzyme which breaks down the network.

Knowing that visual cortical plasticity is dependent on the maturation of inhibitory circuits, Hensch's team examined the role of the transcription factor Otx2 in specifying basket cell function. Transcription factors are proteins which bind to DNA in the nucleus of a cell, and activate specific sets of genes which give each type of neuron its unique identity. Otx2 is expressed early in neural development, and is involved in demarcating those territories of the embryonic nervous system that will eventually form the forebrain and cerebellum. It is also found in the retina, lateral geniculate nucleus and visual cortex long before the appearance of the basket cells. Until now, however, a possible role for Otx2 in visual cortical plasticity has not been investigated.

First of all, the researchers used antibody staining to determine the distribution of Otx2 protein in the mouse visual cortex. At the same time, they visualized basket cells using an antibody against a calcium binding protein called parvalbumin (PV), which is expressed by this class of interneuron. At postnatal day 19, before the onset of experience-dependent plasticity, there was very little Otx2 or PV staining. But from postnatal days 28-30, when plasticity is peaking, and onwards, the concentration of the protein increased and the staining continued to intensify. In the majority of cases, the staining of the two antibodies overlapped, and Otx2 was localized to the nuclei of basket cells. Both Otx2 and PV levels in visual cortex were found to increase in response to neuronal activity, and to decrease when the mice were reared in darkness.

Together, these findings suggest that neuronal activity evoked by visual experience leads to Otx2 expression in basket cells, which is required for the maturation of these cells. This was confirmed by another experiment in which recombinant human Otx2 was delivered directly into the visual cortex of mice reared in the dark. This resulted in an increase in PV-positive cells on the side of the brain to which the exogenous protein was added, but not on the other side, which was untreated. Furthermore, delivery of Otx2 prior to the onset of plasticity at postnatal day 19 accelerated the maturation of basket cells, as shown by a significant increase in the intensity of PV staining.

As would be expected if plasticity is dependent upon the maturation of inhibitory circuits, delivery of Otx2 at postnatal day 19 also hastened the onset of the critical period. Conversely, mice engineered to lack Otx2 throughout the cortex did not exhibit plasticity, and had reduced numbers of basket cells in the cortex.

Surprisingly, however, Otx2 mRNA was not detected in the visual cortex during the period of plasticity, nor was green fluorescent protein whose gene was under the control of the DNA sequences which normally drive Otx2 expression. In other words, the Otx2 protein is not being synthesized in the visual cortex, even though it is required for the development of the inhibitory interneurons. Instead, Otx2 synthesis is restricted to the upstream parts of the visual pathway - the retina, lateral geniculate nucleus and the superior colliculus, a part of the midbrain involved in generating fast eye movements called saccades.

Thus, Otx2 must be delivered to the visual cortex from some other part of the brain. To investigate the potential routes of Otx2 transport, Hensch and his colleagues injected Otx2 tagged with a molecule called biotin into the eyes of their mice. When the visual cortex was examined later, biotinylated Otx2 protein molecules were found within the nuclei of cells which also expressed PV. However, the retina continued to synthesize the protein even in dark-reared animals. Finally, it was found that injection of an anti-Otx2 antibody into either the retina or the visual cortex prevented plasticity, but "knockdown" of Otx2 synthesis by RNA inhibition only occurred when the interfering RNAs were injected into the retina.

Thus, Otx2 synthesized in the retina is transported to the visual cortex in response to visual experience; the protein is taken up by the nucleus, where it activates a set of basket cell-specific genes which are necessary for the maturation of inhibitory circuits. Otx2 and related proteins are known to contain domains that mediate secretion and internalization, and these are likely to be involved in the transfer of the protein from one cell to another along the visual pathway. The precise mechanism by which this occurs is likely to be the subject of future investigations.

Upon the publication of Hubel and Weisel's work, it came to be believed that deprivation of visual experience beyond the critical period led to irreversible changes in the visual cortex. But this has recently been challenged, and the current study adds weight to the notion that visual cortical plasticity can be induced in adult life. It also suggests that a treatment based on Otx2 could one day be available. Such a treatment could in theory promote cortical plasticity after the critical period in amblyopic children, and possibly in elderly patients with failing eyesight.

Related:

• An overview of corticogenesis
• Hubel's Eye, Vision and Brain online
• AMPA receptors and synaptic plasticity
• Synapse proteomics and brain evolution

Sugiyama, S. et al (2008). Experience-Dependent Transfer of Otx2 Homeoprotein into the Visual Cortex Activates Postnatal Plasticity. Cell 134: 508-520. DOI: 10.1016/j.cell.2008.05.054