Ординатура / Офтальмология / Английские материалы / Eye, Retina, and Visual System of the Mouse_Chalupa, Williams_2008

CHINFEI CHEN

What is synaptic remodeling?

Synapses, the connections between neurons of the nervous system, have a remarkable ability to change in strength over a wide range of time scales. This process allows an organism to adapt to its external environment throughout life. During development, the large-scale rearrangement of initially redundant neuronal connections involves a combination of strengthening, weakening, elimination, and reformation of synapses. Such synaptic remodeling—the making and breaking of connections in a neuronal circuit—is crucial for proper development of the nervous system. Developmental synaptic remodeling occurs throughout the animal kingdom, in peripheral and central nervous systems, and the mechanisms underlying this process are of great interest to researchers (Katz and Shatz, 1996; Constantine-Paton and Cline, 1998).

The visual system is a powerful model for studying synaptic remodeling. The visual system is one of the best understood sensory systems and an elegant example of brain adaptation during development (Hübel and Wiesel, 1979; Hübel, 1988). Visual information is initially encoded in the retina, and the output of this processing is specific firing patterns of retinal ganglion cells (RGCs). This information is then transmitted to thalamic relay neurons in the lateral geniculate nucleus (LGN), also known as the visual thalamus, via the retinogeniculate synapse. Relay neurons in turn project to the primary visual cortex and subsequently to neighboring regions of the cortex involved in higher-order processing. At all levels of the visual system, synaptic features can be modified in response to changes in neural activity. In the retina, refinement of RGC dendritic arborization is disrupted on visual deprivation (Tian and Copenhagen, 2003). In the LGN, normal developmental segregation of retinal axons into eye-specific layers is disrupted with blockade of neuronal activity (Sretavan and Shatz, 1986; Shatz and Stryker, 1988). In the visual cortex, changes in sensory experience have profound effects on the structure and function of ocular dominance columns (Wiesel and Hubel, 1963a, 1965; Blakemore et al., 1978). Thus, the visual system is a

useful model for studying synaptic remodeling over development and in response to changes in the external environment (Hubel, 1988). This review focuses largely on synaptic remodeling of the mouse LGN, where the ease of monitoring morphological changes in retinal projections and recording functional changes at synaptic connections, combined with the power of mouse genetics, has provided insight into this developmental process.

Synapses of the mouse lateral geniculate nucleus

The mouse, a nocturnal animal, has lower visual acuity than the cat, ferret, or primate. Yet examination of the cellular and synaptic organization of neurons at different levels of the mouse visual system reveals many similarities with other mammals. For example, different regions of the mouse LGN are innervated exclusively by one or the opposite eye, much like the eye-specific layers found in monkey, ferret, and cat LGN (Rakic, 1976; Linden et al., 1981; Shatz, 1983; Godement et al., 1984). Similar to the ferret, mice have their eyes positioned on the side of their head, and thus their visual system is more monocular than binocular. The majority of the mouse LGN receives retinal projections from the opposite (contralateral) eye, with only a small region in the LGN representing ipsilateral projections. This bias toward contralateral projections persists in the visual cortex, which has a relatively small binocular region, a territory where cortical pyramidal cells can be driven by visual stimuli in either eye.

At the cellular level, electron microscopy (EM) studies of the mouse LGN reveal a similar architecture to that of the cat and monkey (Szentagothai, 1963; Colonnier and Guillery, 1964; Peters and Palay, 1966; Rafols and Valverde, 1973; Lieberman, 1974). Two general classes of neurons are seen in the mouse LGN: (1) excitatory thalamic relay neurons that project to the visual cortex and (2) intrinsic interneurons. As in other mammals, thalamic relay neurons outnumber the intrinsic interneurons by approximately four to one (Steriade et al., 1997). While thalamic relay neurons are further classified based on their summation response to

visual stimulation into X and Y cells in cat or primates (Cleland et al., 1971; Hoffmann et al., 1972; Shapley et al., 1981), this specialization is still unclear in the mouse. However, the majority of mouse relay neurons recorded in vivo exhibit linear spatial summation responses most similar to the characteristics of X cells (Grubb and Thompson, 2003).

RGC axons synapse on dendritic shafts as well as dendritic spines of thalamic relay neurons in the mouse (Rafols and Valverde, 1973; Lieberman, 1974). They also terminate on presynaptic dendritic structures of intrinsic interneurons that, in turn, form dendrodendritic synapses with thalamic relay neurons. This unusual configuration of synaptic structures, called triads, is thought to play a role in a fast ionotropic negative feedback circuit in cat and monkey (Famiglietti, 1970; Pasik et al., 1973; Hamos et al., 1985; Koch, 1985). An in vitro study from mouse LGN recently demonstrated activation of inhibitory current in relay neurons within 1 ms of glutamatergic retinal axon activation, consistent with a fast inhibitory output of triadic synapses (Blitz and Regehr, 2005).

At the synaptic level, the retinogeniculate synapse is glutamatergic, containing both AMPA and NMDA receptors (Chen and Regehr, 2000). Thus, neurotransmission in the mouse is similar to that in the ferret (Mooney et al., 1993), cat (Kemp and Sillito, 1982), monkey (Molinar-Rode and Pasik, 1992), and rat (Salt, 1986). Elegant EM studies in cats demonstrate that the majority of retinal axons terminate in the proximal third of the dendritic tree, while glutamatergic corticothalamic projections synapse more distally, on the thalamic relay neuron (Hamos et al., 1985; Wilson et al., 1984). Although an analogous study is lacking in the mouse, patch-clamp recordings of excitatory postsynaptic currents (EPSCs) at retinogeniculate synapses demonstrate a very

rapid AMPA receptor EPSC decay time course (τ = 1–2 ms), consistent with a proximal location of the synapses (Chen and Regehr, 2000). In contrast, EPSCs from corticothalamic synapses exhibit slower decay kinetics, which most likely reflects their distal position in the relay neuron dendritic tree (Golshani et al., 1998).

Robust synapse remodeling in the mouse lateral geniculate nucleus

Our current understanding of developmental synaptic remodeling in the LGN, shaped from work from many laboratories, is that synapse development and maturation can be separated into at least three distinct phases: (1) coarse mapping and rearrangement of RGC axons in the LGN, (2) fine-scale refinement of neuronal circuitry and functional maturation of synaptic connections, and (3) maintenance of mature synaptic connections (figure 35.1). In the first phase, axons from presynaptic RGCs must map to the postsynaptic relay neurons, distinguishing their proper targets from inappropriate targets. At the retinogeniculate synapse, presynaptic neurons can be labeled in retina, and thus the first phase of remodeling has been elegantly visualized in many species (Rakic, 1976; Linden et al., 1981; Shatz, 1983; Jeffery, 1984; Ziburkus and Guido, 2006), including the mouse (Godement et al., 1984; Muir-Robinson et al., 2002; Jaubert-Miazza et al., 2005). These studies demonstrate that the mapping of retinal axons is initially diffuse, and then, over development, there is a large-scale rearrangement of RGC axon terminals into their eye-specific layers (Sretavan and Shatz, 1984, 1986; see also chapters 28 and 34).

By the time the bulk of eye-specific segregation is complete (P0 in cat, P8–P12 in ferret, mouse, and rat), synaptic contacts have formed and are functional (Shatz and

Figure 35.1 Timeline of three phases of synapse remodeling in the mouse lateral geniculate nucleus. The initial phase involves retinal axon mapping and coarse rearrangement of axons into eye-specific layers. In the second phase, synaptic connectivity

refines, as many synapses are functionally eliminated while others strengthen. During the third phase, synaptic maintenance depends on vision. Time windows for different forms of activity and the cortical critical period are superimposed for comparison.

Kirkwood, 1984; Ramoa and McCormick, 1994; Chen and Regehr, 2000; Jaubert-Miazza et al., 2005; Ziburkus and Guido, 2006). However, a number of features suggest that these synapses are immature at this time (Ramoa and McCormick, 1994; Chen and Regehr, 2000). Synaptic strength is weak, as assessed by the amplitude of the AMPA receptor currents, while the AMPAR/NMDAR current ratio is low. Moreover, the time course of NMDA receptor decay kinetics is slow, consistent with a lack of NR2A receptor, a subunit of NMDAR that is inserted as the synapse matures (Carmignoto and Vicini, 1992; Hestrin, 1992; Monyer et al., 1992, 1994; Vicini et al., 1998). Studies using in vitro preparations of the mouse LGN demonstrate that subsequent synaptic development after axons have segregated into eye-specific layers involves an increase in AMPAR/NMDAR current ratio, an acceleration of NMDAR decay time course, and intense remodeling of the connectivity between RGC and LGN relay neurons (Chen and Regehr, 2000; Jaubert-Miazza et al., 2005).

In a region of the mouse LGN that receives mainly projections from the contralateral retina by P7, 10–15 RGC inputs are estimated to innervate a given LGN neuron between P9 and P11, while the average synaptic strength is weak (Chen and Regehr, 2000). Over the next 2 weeks, spanning the time of eye opening (P12–P14), the number of inputs decreases, or prunes, down to one to three inputs, while the average strength of the remaining inputs increases 20-fold (figure 35.2). Insofar as sharpening of LGN receptive fields (RFs) occurs after eye opening in the cat, ferret, and monkey (Daniels et al., 1978; Blakemore and Vital-Durand, 1986; Tavazoie and Reid, 2000), it is plausible that the large-scale synaptic rearrangement described in mice contributes to this sharpening process. Surprisingly, although the bulk of synaptic remodeling during this second phase of remodeling occurs around eye opening, it does not depend on vision. Instead, spontaneous activity is the driving force of synaptic remodeling during this developmental phase (Hooks and Chen, 2006).

The third phase of remodeling of the retinogeniculate synapse is notable for its dependence on vision (Hooks and Chen, 2006). Dark rearing from birth in mice does not appear to alter the normal developmental remodeling of the retinogeniculate synapse. In contrast, deprivation after a week of vision results in a dramatic change in connectivity. When mice are dark reared from P20 for more than a week (also referred to as late dark rearing), the number of inputs increases from about 3–4 at P20 to about 10 at P27–P32. Concurrently, the strength of the average retinal input decreases by more than half. Thus, visual experience is necessary to maintain LGN circuitry during the third phase of synaptic remodeling. It is currently unclear whether this rearrangement is a reflection of reactivation of previously “eliminated” synapses or the sprouting of de novo synapses.

Figure 35.2 Role of activity in retinogeniculate development. Shown are changes in the strength and number of RGC axons that innervate an LGN relay neuron over development. Visual deprivation from birth does not appear to disrupt the normal developmental process. However, after exposure to visually evoked activity, the synaptic connections become dependent on vision.

Activity and the three phases of synapse remodeling

Studies in mice thus reveal three phases of plasticity that appear to be regulated by different mechanisms. All are activity dependent, but the relationship between activity and plasticity is different in each case. Both activity and molecular cues play an important role in axon mapping in the first phase of remodeling (Katz and Shatz, 1996). Careful studies involving pharmacological manipulations in the eye and the use of mouse mutants have demonstrated that segregation of these fibers is dependent on retinal waves, synchronous bursts of neuronal activity that march across the retina with a periodicity of 1–2 minutes (Galli and Maffei, 1988; Meister et al., 1991; Wong et al., 1993; Feller et al., 1997; Penn et al., 1998; Muir-Robinson et al., 2000), see also chapters 28 and 34. In addition to retinal waves, axon guidance cues, notably the ephrins, play an important role in axon mapping (Feldheim et al., 1998, 2000; Lyckman et al., 2001; Huberman et al., 2005). Eye-specific layering defects are greater in ephrin-A2, -A3, and -A5 knockout mice that have disruption of retinal waves than in the same mice with normal retinal waves or in mice with abnormal retinal waves (Pfeiffenberger et al., 2005, 2006). Thus, retinal waves and axon guidance cues appear to have distinct roles in dictating the final location of RGC axon terminals.

The second phase of synapse remodeling, between P9 and P20 in mice, is also dependent on spontaneous activity, although the relationship between activity and remodeling is not linear. Inhibiting retinal activity for 12 hours per day over 4 days did not disrupt synaptic remodeling (unpublished observations). Instead, continuous blockade of all activity in the retina over 4 days is necessary for significant retardation of developmental synapse pruning and strength-

ening (Hooks and Chen, 2006). Thus, the threshold of activity necessary for the second phase of remodeling at the retinogeniculate synapse is very low.

How the first developmental phase of eye-specific segregation relates to the functional synapse plasticity seen during the second phase is unclear. Specific features of retinal waves appear to drive the morphological rearrangement of retinal axon arbors to the proper region of the mouse LGN (Torborg et al., 2005; Demas et al., 2006; but see also Huberman et al., 2003; Grubb and Thompson, 2004; Huberman and Chapman, 2006). However, it is unclear which aspect of these patterns is important for the subsequent functional remodeling of synapses. Indeed, genetically altered mice lacking the nicotinic acetylcholine receptor β2 subunit (nAChR-β2; beta 2 mice) exhibit disrupted retinal waves and eye-specific segregation (Muir-Robinson et al., 2002), but only mild abnormalities in LGN RF properties (Grubb et al., 2003; Cang et al., 2005). At the synaptic level, preliminary studies from our laboratory show that developmental input pruning and strengthening of the retinogeniculate synapse are not significantly disrupted at mature ages (>P25) in either beta 2 mice or ephrin-A2/A5 knockout mice (unpublished observations; Hooks et al., 2004). Thus, the second phase of synaptic remodeling is not simply a continuation of the rearrangement of retinal axon fibers seen before P8. The mechanisms underlying the first and second phases of developmental remodeling are likely to be different.

Two aspects of the third phase of plasticity are activity dependent. During this period, maintenance of synaptic connections requires visually evoked activity. Thus, expression of this phase of synaptic remodeling is vision dependent. However, the expression of synaptic plasticity during the third phase is not dependent on cumulative levels of activity. In chronically dark-reared animals, a slowing of NMDA receptor decay time course occurs, but only after P21 in age; deprivation after P21 results in the same effects. Thus, deprivation experiments in mice clearly demonstrate a specific window of development when synaptic function becomes sensitive to visual deprivation.

The dependence on vision occurs only with a history of prior visual experience, suggesting that patterned visually evoked activity, but not spontaneous activity, is necessary for induction of the third phase of synaptic remodeling. It is unclear whether the feature of vision that is important for induction is a specific pattern of activity or whether a threshold level of absolute activity is necessary. The sensory conditions for chronic and late dark-reared animals are identical after P20, yet the response of the retinogeniculate synapse is quite distinct. Thus the difference between the two conditions lies in the period between eye opening and P20. In mice, retinal waves persist for a few days after eye opening, begin to break down by P15, and have completely disappeared by P21 (Demas et al., 2003). Thus there is a short

time between eye opening and P21 when both retinal waves and visually evoked activity are seen by LGN relay neurons. Dark rearing during this period does not significantly disrupt retinal wave activity yet may reduce the total level of activity. It is tempting to hypothesize that the induction of visiondependent synaptic remodeling depends on the total level of activity.

Plasticity in the lateral geniculate nucleus:

Comparing the mouse to other species

Data from the mouse suggest there is a great deal of functional plasticity in the LGN, especially during the third phase of remodeling. Is there plasticity in the LGN of other species? Chronic dark-rearing studies in kittens from birth to 4 months of age (a time that corresponds to the end of the cortical critical period in this species) revealed normal RF responses from X cells (Kratz et al., 1979). However, there appeared to be a reduction in the number of physiologically identifiable Y cells in both the monocular and binocular regions of the LGN. These findings were consistent with morphological studies showing an increase in neurons containing laminated bodies, thought to represent X cells, in dark-reared cats (LeVay and Ferster, 1977; Kalil and Worden, 1978). Because neurons of the mouse LGN have not been classified into X and Y cell categories, direct comparisons of the effects of chronic dark rearing in mouse and cat are difficult. If, however, there is a predominance of X cells in mice (Grubb and Thompson, 2003), the lack of dramatic synaptic remodeling at the mouse retinogeniculate synapse in response to chronic dark rearing would be consistent with that previously described in cat. In ferrets, dark rearing at P16 through closed eyelids resulted in convergence of ONand OFF-center responses. Whether abnormalities in ON/OFF segregation correspond directly to changes in retinogeniculate connectivity or to changes in RGC dendritic arborization (or both) is not clear (Ackerman et al., 2002; Tian and Copenhagen, 2003).

There are remarkably few studies in the LGN examining the physiological consequences to dark rearing during the cortical critical period in cat or monkey that could be compared directly to the third phase of synaptic remodeling in mice. Most studies in these larger mammals have involved monocular or binocular deprivation. The classic work of Wiesel and Hubel demonstrated that only 20% of relay neurons displayed irregular RF properties in the LGN in monocularly deprived kittens during the period when ocular dominance plasticity is robust (Wiesel and Hübel, 1963a, 1963b). However, these studies focused mainly on the midline area of the LGN, representing the region of the visual field that is projected on both left and right retinas (Hübel, 1988). It is unclear whether recordings in the LGN representing the temporal part of the visual field, presumably receiving mon-

ocular innervation, would have revealed a different physiology. Interestingly, a study examining monocular deprivation of the rabbit, a species in which more than 90% of the LGN receives monocular innervation, demonstrated a specific developmental time window during which visual deprivation results in disrupted RF properties (Baumbach and Chow, 1978). Moreover, these authors described a later phase of development when vision is needed for maintenance of the RF.

Surprisingly, binocular eyelid suturing in cats elicited different responses than dark rearing or monocular deprivation. For example, monocular deprivation results in a reduction of the size of geniculate neuron somata in the binocular segment of the LGN (Wiesel and Hübel, 1963a), although relay neurons in the monocular segment of the same layer (the region that receives innervation from the deprived eye but not the nondeprived eye) are unaffected (Guillery, 1972). In contrast, binocular suturing results in milder effects on soma size in the binocular region of the LGN, with a greater effect on the monocular segment of the LGN (Wiesel and Hübel, 1965; Guillery, 1972; Hickey et al., 1977). Finally, dark rearing does not significantly affect geniculate soma size. Although it is still unclear why there should be a difference in results of binocular eye suturing and dark rearing, recent data from ferret raise the possibility that light perceived through closed eyelids may represent a differentiating factor (Akerman et al., 2002).

Comparison of remodeling in the lateral geniculate nucleus to other regions of the visual system

The Vision-Sensitive Phase of Lateral Geniculate Nucleus Remodeling and the Cortical Critical

Period The vision-dependent period of synaptic remodeling during development of the mouse retinogeniculate synapse corresponds strikingly to the cortical critical period in mouse (Gordon and Stryker, 1996; Hensch, 2004), despite differences in the form of visual deprivation (dark rearing vs. monocular deprivation). Similar to the findings in the LGN, visual deprivation elicits a change in the response of cortical neurons only during a specific developmental window. The analogous change in the binocular region of the visual cortex involves a shift in the responsiveness of cortical neurons to stimulation of one eye or the other, that is, a shift in the ocular dominance (OD) preference of cortical neurons. It is still unclear how the plasticity between the LGN and cortex relate to each other during this developmental period. Both feedback and feedforward processes may influence the response of the LGN and cortex to visual deprivation.

In the mouse visual cortex, changes in OD plasticity depend strongly on the balance of excitatory and inhibitory circuit interactions (Hensch, 2005). The maturation of in-

hibitory circuits plays an important role in this balance. Manipulations of the level of GABAergic inhibition in the visual cortex can alter the onset of the critical period, consistent with the idea that the induction of OD plasticity is dependent on the maturation of inhibitory synapses (Hensch et al., 1998; Fagiolini and Hensch, 2000). In the future, it will be interesting to determine whether activation of the inhibitory circuitry in the LGN is also important in the induction of the third phase of remodeling in the LGN.

In contrast to the weakening and change in connectivity of retinogeniculate synapses in response to late dark rearing, intracortical glutamatergic synapses onto pyramidal neurons in the monocular region of the visual cortex are unchanged in response to 3 days of deprivation during the cortical critical period (Maffei et al., 2006). Instead, synapses between inhibitory fast-spiking basket cells and star pyramidal neurons strengthen, consistent with a role for LTP of inhibitory circuits in the expression of OD plasticity. However, longer periods of monocular deprivation (40 days) result in the shrinkage of thalamocortical axon arbors serving the deprived eye and further shift OD preference (Antonini et al., 1999). Moreover, some forms of OD plasticity have been shown to be NMDA receptor dependent (Sawtell et al., 2003). This finding, along with studies that demonstrate occlusion of long-term synaptic depression (LTD) in the binocular visual cortex by previous monocular deprivation, suggests that LTD of excitatory synapses may also be involved in synapse remodeling in the cortex (Heynen et al., 2003; Crozier et al., 2007).

The Superior Colliculus Many axons of RGCs bifurcate and innervate both the LGN and superior colliculus (SC; Illing, 1980; Yamadori et al., 1989). Despite their having the same presynaptic neurons, however, there are notable differences in the developmental remodeling of the retinal synapses onto the two subcortical regions. In genetically altered mice with disrupted retinal axon mapping, such as those lacking ephrin-A2/A3/A5, nAChR-β2, serotonergic receptor 5-HT1B, or the serotonin transporter, mapping defects are greater in the SC than in the LGN (Upton et al., 1999, 2002; Feldheim et al., 2000; Pfeiffenberger et al., 2005, 2006). Consistent with axon mapping defects, RFs in the mice with disrupted retinal waves (beta 2 mice) are more distorted in the SC than in the LGN (Grubb et al., 2003; Chandrasekaran et al., 2005; Mrsic-Flogel et al., 2005).

In vitro studies of rat SC demonstrate that glutamatergic inputs onto superficial collicular neurons also prune at the synaptic level during development (Lu and ConstantinePaton, 2004). In contrast to the LGN, acceleration of this normal process occurs within 24 hours after eye opening. However, in the brain slice preparation of the superficial laminae of the rodent colliculus, it is difficult to selectively stimulate inputs from the cortex, retina, and brainstem.

Thus, the specific class of inputs innervating the collicular neurons that exhibit sensitivity to visual activity during development is still not clear. Moreover, the bulk of pruning still occurs in visually deprived rats, although with a lag when compared with rats reared in a 12-hour light-dark cycle (Lu and Constantine-Paton, 2004). Thus, vision accelerates the time course of the pruning in the colliculus, although the process can occur at a slower rate without sensory experience. These differences between the LGN and the SC during the first and second phases of synapse remodeling suggest that the rules governing developmental synaptic remodeling may also be cell-type specific for RGC targets.

Despite differences in synaptic maturation between different visual system areas, similarities also exist with respect to the role of vision. As for the LGN, the role of sensory experience in the development of the SC varies among species. Chronic visual deprivation does not disrupt the average RF properties of neurons in hamster SC (Rhoades and Chalupa, 1978; Chalupa and Rhoades, 1978) or rabbit (Chow and Spear, 1974). In contrast, binocular eyelid closure in kittens results in significant changes in direction sensitivity in collicular neurons, consistent with a reduction in the Y indirect and direct pathways (Hoffmann and Sherman, 1975). In addition, a recent study examining a number of developmental time points during prolonged chronic dark rearing revealed that collicular RFs gradually become larger (Carrasco et al., 2005). Thus, the role of vision in the maintenance of synapses or synaptic circuits may be a common theme at the retinogeniculate, retinotectal, and cortical synapses of the visual system across species (Crair et al., 1998; Carrasco et al., 2005; Hooks and Chen, 2006).

Genes involved in synapse remodeling

There is a great deal of interest in identifying the genes involved in synaptic remodeling. The power of mouse genetics can be harnessed to this end. Because this field is quite extensive, I cannot do justice to all the candidate genes for developmental synaptic remodeling. Thus, in this chapter, I highlight only some genes that are proposed to play a role in synapse remodeling in the LGN.

Two recent studies using unbiased differential gene expression screens in the visual cortex identified groups of activity-dependent genes that had relatively little overlap (Majdan and Shatz, 2006; Tropea et al., 2006). However, a number of themes emerged from these studies. Both groups found that although there is a set of genes that appear to be regulated by general changes in activity, there are also sets of genes regulated by reducing visually evoked activity in both eyes (dark rearing) that differed from those identified by altering the balance of activity between the two eyes (monocular deprivation or enucleation). Moreover, certain genes are regulated only during a particular developmental

window, such as during the cortical critical period, and still others require previous visual experience for normal regulation of gene expression. In the future, it will be interesting to assess whether synaptic remodeling in the different regions of the visual system, in particular the visual cortex and LGN, share common regulatory gene mechanisms.

Other activity-dependent gene expression screens have identified candidates that may contribute to the large-scale synaptic remodeling observed at the retinogeniculate synapse. One gene of interest, cpg15 (candidate plasticity gene 15), was identified from a forward genetic screen in the rat hippocampus after seizure induction (Nedivi et al., 1993). Cpg15 is an activity-dependent gene expressed in the visual system of the cat, rat, mice, and tadpole (Nedivi et al., 1996, 2001; Corriveau et al., 1999) that encodes a secreted protein that binds to the extracellular membrane via glycosylphosphatidylinositol linkage (Naeve et al., 1997). In Xenopus laevis, infection of cpg15 in tectal neurons enhances dendritic arbor growth (Nedivi et al., 1998). In the rat, cpg15 is expressed at high levels in the LGN in the first 2 postnatal weeks and then declines with age. At a corresponding developmental period when cgp15 levels are high in cat (prenatally), infusion of the sodium channel inhibitor tetrodotoxin (TTX) into the cerebrospinal fluid or injection of TTX into the eye does not alter cgp15 levels in the contralateral LGN. In contrast, monocular TTX injections at a time corresponding to the onset of the cortical critical period (P18+ in the cat) results in a decrease in LGN cpg15 levels (Corriveau et al., 1999). Expression of cpg15 also occurs in the cortex, lagging developmentally behind that in the LGN. In rat, cpg15 levels are detected at P10 and increase between the second and third postnatal weeks before decreasing in the adult (Nedivi et al., 1996). At the peak of the cortical critical period, cgp15 expression gradually becomes dependent on retinal-driven action potentials. Dark rearing during that time decreases the peak levels of cpg15, but, more interestingly, it prevents the normal developmental decline in cpg15 levels in adulthood. Moreover, previous visual experience during a specific developmental window is important for proper regulation of cpg15 expression (Lee and Nedivi, 2002). Thus, cpg15 exhibits an expression pattern that shares many developmental features of the sensitive/critical period in both LGN and cortex. It will be interesting to examine whether this gene plays a role in glutamatergic synapse remodeling.

Another screen for genes whose expression in the cat LGN changes when spontaneous activity is blocked with intracranial infusion of TTX identified the class I major histocompatibility complex (MHC I) antigen (Corriveau et al., 1998), a protein previously shown to play a role in cell-mediated immune recognition. TTX injected monocularly in the cat reduced the expression of MHC I mRNA. In mice, MHC I is present in the LGN at P6, and the expression decreases by P40. In addition to the visual system,

MHC I is also found in the hippocampus and cortex. Genetically modified mice deficient in either MHC I or a subunit of the MHC receptor (CD3ζ) demonstrate defects in eyespecific layer formation (Huh et al., 2000). Moreover, hippocampal LTP in these mutant mice is enhanced, while LTD is absent. These findings suggest that MHC I plays a role in the developmental remodeling of neuronal axon morphology as well as functional synapses. It will be interesting to determine whether MHC I plays a role in remodeling of the retinogeniculate synapse, and if so, what phase of remodeling it regulates.

Another group of molecules of interest are the neuronal pentraxins, a family of synaptic proteins that have homology to acute phase proteins (pentraxins) of the immune system. Neuronal pentraxins (NP1 and NP2) are present in the mouse LGN at P7, a developmental time when eye-specific layering is nearly complete. The levels of NP1 decrease by P14. In contrast, NP2 is still highly expressed in retinal axons, and NP receptor expression increases during this developmental period. Although NP1/2 knockout mice exhibit abnormal segregation of eye-specific layers, the number of synaptic contacts, assessed in cultures of purified RGCs from these mutant animals, was not different from that seen in cultures from wild-type mice. However, the normal developmental increase in the frequency of mEPSCs did not occur in cultures from the knockout mice, suggesting that although synapses were present, they did not mature normally (Bjartmar et al., 2006).

With growing evidence of different phases of developmental synapse remodeling, it will be important to determine the stage of remodeling each identified candidate gene regulates. For example, neuronal pentraxin affects the early stage of eye-specific layering and possibly synapse maturation (the first and second phase of remodeling). In contrast, mutant mice lacking paired-immunoglobulin-like receptors (PirB mice), a class I MHC receptor, exhibit normal developmental changes of eye-specific inputs in the LGN and cortex (Syken et al., 2006). However, in PirB mice, OD plasticity, as measured indirectly by arc labeling (Tagawa et al., 2005), persists past the normal critical period (Syken et al., 2006). These findings suggest that PirB plays a role in the cortical critical period. It will be interesting to examine whether PirB also plays a role in the third phase of synaptic remodeling at the retinogeniculate synapse.

Conclusion

Our understanding of synaptic remodeling in the CNS continues to grow. Future work will benefit from the power of well-established mouse models of visual development and from mouse genetics. With increasing evidence of distinct phases of synaptic maturation, it is likely that future studies will be able to identify distinct phases of active pruning and

maintenance that appear to be governed by different genes and forms of activity.

A number of questions have yet to be answered. One basic question is, what is the purpose of developmental synaptic remodeling? Although some synaptic connections in the CNS exhibit large-scale changes over the course of development, others develop with striking specificity and do not exhibit an early period of refinement (Callaway and Lieber, 1996; Bender et al., 2003; Bureau et al., 2004). One proposed model suggests that transient connections in the LGN are used to establish fine-tuned, oriented RFs in the visual cortex (Tavazoie and Reid, 2000). An alternative model arises from an observation applicable to many areas of the brain, namely, that the ability to adapt to changes is enhanced by previous experience (Knudsen et al., 2000; Hofer et al., 2006b). Perhaps the “memory” of previously pruned synaptic connections provides a scaffold for potential changes in the adult brain (DeBello et al., 2001; Linkenhoker et al., 2005; Hofer et al., 2006a). This may be a principle that is generalizable to the entire CNS.

acknowledgments Work was supported by the National Eye Institute (grant no. EY013613) and by the Children’s Hospital, Boston, Mental Retardation and Developmental Disabilities Research Center (grant no. PO1 HD18655). I thank B. M. Hooks, Xiaojin Liu, James Choi, Alan Mardinly, Brett Carter, and Whitney Blair for their contributions.

REFERENCES

Akerman, C. J., Smyth, D., and Thompson, I. D. (2002). Visual experience before eye-opening and the development of the retinogeniculate pathway. Neuron 36:869–879.

Antonini, A., Fagiolini, M., and Stryker, M. P. (1999). Anatomical correlates of functional plasticity in mouse visual cortex. J. Neurosci. 19:4388–4406.

Baumbach, H. D., and Chow, K. L. (1978). Receptive field development in the dorsal lateral geniculate nucleus in rabbits subjected to monocular eyelid suture. Brain Res. 159:69–83.

Bender, K. J., Rangel, J., and Feldman, D. E. (2003). Development of columnar topography in the excitatory layer 4 to layer 2/3 projection in rat barrel cortex. J. Neurosci. 23:8759–8770.

Bjartmar, L., Huberman, A. D., Ullian, E. M., Renteria, R. C., Liu, X., Xu, W., Prezioso, J., Susman, M. W., Stellwagen, D., et al. (2006). Neuronal pentraxins mediate synaptic refinement in the developing visual system. J. Neurosci. 26:6269–6281.

Blakemore, C., Garey, L. J., and Vital-Durand, F. (1978). The physiological effects of monocular deprivation and their reversal in the monkey’s visual cortex. J. Physiol. 283:223–262.

Blakemore, C., and Vital-Durand, F. (1986). Organization and post-natal development of the monkey’s lateral geniculate nucleus. J. Physiol. 380:453–491.

Blitz, D. M., and Regehr, W. G. (2005). Timing and specificity of feed-forward inhibition within the LGN. Neuron 45:917–928.

Bureau, I., Shepherd, G. M. G., and Svoboda, K. (2004). Precise development of functional and anatomical columns in the neocortex. Neuron 42:789–801.

Callaway, E. M., and Lieber, J. L. (1996). Development of axonal arbors of layer 6 pyramidal neurons in ferret primary visual cortex. J. Comp. Neurol. 376:295–305.

Cang, J., Kaneko, M., Yamada, J., Woods, G., Stryker, M. P., and Feldheim, D. A. (2005). Ephrin-As guide the formation of functional maps in the visual cortex. Neuron 48:577–589.

Carmignoto, G., and Vicini, S. (1992). Activity-dependent decrease in NMDA receptor responses during development of the visual cortex. Science 258:1007–1011.

Carrasco, M. M., Razak, K. A., and Pallas, S. L. (2005). Visual experience is necessary for maintenance but not development of receptive fields in superior colliculus. J. Neurophysiol. 94:1962– 1970.

Chalupa, L. M., and Rhoades, R. W. (1978). Directional selectivity in hamster superior colliculus is modified by strobe-rearing but not by dark-rearing. Science 199:998–1001.

Chandrasekaran, A. R., Plas, D. T., Gonzalez, E., and Crair, M. C. (2005). Evidence for an instructive role of retinal activity in retinotopic map refinement in the superior colliculus of the mouse. J. Neurosci. 25:6929–6938.

Chen, C., and Regehr, W. G. (2000). Developmental remodeling of the retinogeniculate synapse. Neuron 28:955–966.

Chow, K. L., and Spear, P. D. (1974). Morphological and functional effects of visual deprivation on the rabbit visual system.

Exp. Neurol. 42:429–447.

Cleland, B. G., Dubin, M. W., and Levick, W. R. (1971). Sustained and transient neurones in the cat’s retina and lateral geniculate nucleus. J. Physiol. 217:473–496.

Colonnier, M., and Guillery, R. W. (1964). Synaptic organization in the lateral geniculate nucleus of the monkey. Z. Zellforsch. Mikrosk. Anat. 62:333–355.

Constantine-Paton, M., and Cline, H. T. (1998). LTP and activity-dependent synaptogenesis: The more alike they are, the more different they become. Curr. Opin. Neurobiol. 8:139–148.

Corriveau, R. A., Huh, G. S., and Shatz, C. J. (1998). Regulation of class I MHC gene expression in the developing and mature CNS by neural activity. Neuron 21:505–520.

Corriveau, R. A., Shatz, C. J., and Nedivi, E. (1999). Dynamic regulation of cpg15 during activity-dependent synaptic development in the mammalian visual system. J. Neurosci. 19:7999– 8008.

Crair, M. C., Gillespie, D. C., and Stryker, M. P. (1998). The role of visual experience in the development of columns in cat visual cortex. Science 279:566–570.

Crozier, R. A., Wang, Y., Liu, C. H., and Bear, M. F. (2007). Deprivation-induced synaptic depression by distinct mechanisms in different layers of mouse visual cortex. Proc. Natl. Acad. Sci. U.S.A. 104:1383–1388.

Daniels, J. D., Pettigrew, J. D., and Norman, J. L. (1978). Development of single-neuron responses in kitten’s lateral geniculate nucleus. J. Neurophysiol. 41:1373–1393.

DeBello, W. M., Feldman, D. E., and Knudsen, E. I. (2001). Adaptive axonal remodeling in the midbrain auditory space map. J. Neurosci. 21:3161–3174.

Demas, J., Eglen, S. J., and Wong, R. O. (2003). Developmental loss of synchronous spontaneous activity in the mouse retina is independent of visual experience. J. Neurosci. 23:2851–2860.

Demas, J., Sagdullaev, B. T., Green, E., Jaubert-Miazza, L., McCall, M. A., Gregg, R. G., Wong, R. O., and Guido, W. (2006). Failure to maintain eye-specific segregation in nob, a mutant with abnormally patterned retinal activity. Neuron 50: 247–259.

Fagiolini, M., and Hensch, T. K. (2000). Inhibitory threshold for

critical-period activation in primary visual cortex. Nature 404: 183–186.

Famiglietti, E. V., Jr. (1970). Dendro-dendritic synapses in the lateral geniculate nucleus of the cat. Brain Res. 20:181–191.

Feldheim, D. A., Kim, Y. I., Bergemann, A. D., Frisen, J., Barbacid, M., and Flanagan, J. G. (2000). Genetic analysis of ephrin-A2 and ephrin-A5 shows their requirement in multiple aspects of retinocollicular mapping. Neuron 25:563–574.

Feldheim, D. A., Vanderhaeghen, P., Hansen, M. J., Frisen, J., Lu, Q., Barbacid, M., and Flanagan, J. G. (1998). Topographic guidance labels in a sensory projection to the forebrain. Neuron 21:1303–1313.

Feller, M. B., Butts, D. A., Aaron, H. L., Rokhsar, D. S., and Shatz, C. J. (1997). Dynamic processes shape spatiotemporal properties of retinal waves. Neuron 19:293–306.

Galli, L., and Maffei, L. (1988). Spontaneous impulse activity of rat retinal ganglion cells in prenatal life. Science 242:90–91.

Godement, P., Salaun, J., and Imbert, M. (1984). Prenatal and postnatal development of retinogeniculate and retinocollicular projections in the mouse. J. Comp. Neurol. 230:552–575.

Golshani, P., Warren, R. A., and Jones, E. G. (1998). Progression of change in NMDA, non-NMDA, and metabotropic glutamate receptor function at the developing corticothalamic synapse. J. Neurophysiol. 80:143–154.

Gordon, J. A., and Stryker, M. P. (1996). Experience-dependent plasticity of binocular responses in the primary visual cortex of the mouse. J. Neurosci. 16:3274–3286.

Grubb, M. S., Rossi, F. M., Changeux, J. P., and Thompson, I. D. (2003). Abnormal functional organization in the dorsal lateral geniculate nucleus of mice lacking the beta 2 subunit of the nicotinic acetylcholine receptor [see comment]. Neuron 40:1161– 1172.

Grubb, M. S., and Thompson, I. D. (2003). Quantitative characterization of visual response properties in the mouse dorsal lateral geniculate nucleus. J. Neurophysiol. 90:3594–3607.

Grubb, M. S., and Thompson, I. D. (2004). The influence of early experience on the development of sensory systems. Curr. Opin. Neurobiol. 14:503–512.

Guillery, R. W. (1972). Binocular competition in the control of geniculate cell growth. J. Comp. Neurol. 144:117–129.

Hamos, J. E., Van Horn, S. C., Raczkowski, D., Uhlrich, D. J., and Sherman, S. M. (1985). Synaptic connectivity of a local circuit neurone in lateral geniculate nucleus of the cat. Nature 317:618–621.

Hensch, T. K. (2004). Critical period regulation. Annu. Rev. Neurosci. 27:549–579.

Hensch, T. K. (2005). Critical period plasticity in local cortical circuits. Nat. Rev. Neurosci. 6:877–888.

Hensch, T. K., Fagiolini, M., Mataga, N., Stryker, M. P., Baekkeskov, S., and Kash, S. F. (1998). Local GABA circuit control of experience-dependent plasticity in developing visual cortex. Science 282:1504–1508.

Hestrin, S. (1992). Developmental regulation of NMDA receptormediated synaptic currents at a central synapse. Nature 357: 686–689.

Heynen, A. J., Yoon, B. J., Liu, C. H., Chung, H. J., Huganir, R. L., and Bear, M. F. (2003). Molecular mechanism for loss of visual cortical responsiveness following brief monocular deprivation. Nat. Neurosci. 6:854–862.

Hickey, T. L., Spear, P. D., and Kratz, K. E. (1977). Quantitative studies of cell size in the cat’s dorsal lateral geniculate nucleus following visual deprivation. J. Comp. Neurol. 172:265–282.

Hofer, S. B., Mrsic-Flogel, T. D., Bonhoeffer, T., and Hubener,

M. (2006a). Lifelong learning: Ocular dominance plasticity in mouse visual cortex. Curr. Opin. Neurobiol. 16:451–459.

Hofer, S. B., Mrsic-Flogel, T. D., Bonhoeffer, T., and Hubener, M. (2006b). Prior experience enhances plasticity in adult visual cortex. Nat. Neurosci. 9:127–132.

Hoffmann, K. P., and Sherman, S. M. (1975). Effects of early binocular deprivation on visual input to cat superior colliculus.

J. Neurophysiol. 38:1049–1059.

Hoffmann, K. P., Stone, J., and Sherman, S. M. (1972). Relay of receptive-field properties in dorsal lateral geniculate nucleus of the cat. J. Neurophysiol. 35:518–531.

Hooks, B. M., and Chen, C. (2006). Distinct roles for spontaneous and visual activity in remodeling of the retinogeniculate synapse. Neuron 52:281–291.

Hooks, B. M., Feldheim, D., Flanagan, J. G., and Chen, C. (2004). Mechanisms of synaptic remodeling. Paper presented at a satellite symposium meeting of ARVO, “Mouse Visual System,” Fort Lauderdale.

Hübel, D. H. (1988). Eye, brain, and vision. Scientific American Library. Distibuted by W. H. Freeman, New York.

Hübel, D. H., and Wiesel, T. N. (1979). Brain mechanisms of vision. Sci. Am. 241:150–162.

Huberman, A. D., and Chapman, B. (2006). Making and breaking eye-specific projections to the lateral geniculate nucleus. In R. Erzurumlu, W. Guido, and Z. Molnar (Eds.), Development and plasticity in sensory thalamus and cortex (pp. 247–270). New York: Springer Science + Business Media.

Huberman, A. D., Murray, K. D., Warland, D. K., Feldheim, D. A., and Chapman, B. (2005). Ephrin-As mediate targeting of eye-specific projections to the lateral geniculate nucleus. Nat. Neurosci. 8:1013–1021.

Huberman, A. D., Wang, G. Y., Liets, L. C., Collins, O. A., Chapman, B., and Chalupa, L. M. (2003). Eye-specific retinogeniculate segregation independent of normal neuronal activity. Science 300:994–998.

Huh, G. S., Boulanger, L. M., Du, H. P., Riquelme, P. A., Brotz, T. M., and Shatz, C. J. (2000). Functional requirement for class I MHC in CNS development and plasticity. Science 290:2155–2159.

Illing, R. B. (1980). Axonal bifurcation of cat retinal ganglion cells as demonstrated by retrograde double labelling with fluorescent dyes. Neurosci. Lett. 19:125–130.

Jaubert-Miazza, L., Green, E., Lo, F. S., Bui, K., Mills, J., and Guido, W. (2005). Structural and functional composition of the developing retinogeniculate pathway in the mouse. Vis. Neurosci. 22:661–676.

Jeffery, G. (1984). Retinal ganglion cell death and terminal field retraction in the developing rodent visual system. Brain Res. 315:81–96.

Kalil, R., and Worden, I. (1978). Cytoplasmic laminated bodies in the lateral geniculate nucleus of normal and dark reared cats.

J. Comp. Neurol. 178:469–485.

Katz, L. C., and Shatz, C. J. (1996). Synaptic activity and the construction of cortical circuits. Science 274:1133–1138.

Kemp, J. A., and Sillito, A. M. (1982). The nature of the excitatory transmitter mediating X and Y cell inputs to the cat dorsal lateral geniculate nucleus. J. Physiol. 323:377–391.

Knudsen, E. I., Zheng, W., and DeBello, W. M. (2000). Traces of learning in the auditory localization pathway. Proc. Natl. Acad. Sci. U.S.A. 97:11815–11820.

Koch, C. (1985). Understanding the intrinsic circuitry of the cat’s lateral geniculate nucleus: Electrical properties of the spine-triad arrangement. Proc. R. Soc. Lond. B. Biol. Sci. 225:365–390.

Kratz, K. E., Sherman, S. M., and Kalil, R. (1979). Lateral

geniculate nucleus in dark-reared cats: Loss of Y cells without changes in cell size. Science 203:1353–1355.

Lee, W. C., and Nedivi, E. (2002). Extended plasticity of visual cortex in dark-reared animals may result from prolonged expression of cpg15-like genes. J. Neurosci. 22:1807–1815.

LeVay, S., and Ferster, D. (1977). Relay cell classes in the lateral geniculate nucleus of the cat and the effects of visual deprivation.

J. Comp. Neurol. 172:563–584.

Lieberman, A. R. (1974). Comments on the fine structural organization of the dorsal lateral geniculate nucleus of the mouse. Anat. Embryol. (Berl.) 145:261–267.

Linden, D. C., Guillery, R. W., and Cucchiaro, J. (1981). The dorsal lateral geniculate nucleus of the normal ferret and its postnatal development. J. Comp. Neurol. 203:189–211.

Linkenhoker, B. A., von der Ohe, C. G., and Knudsen, E. I. (2005). Anatomical traces of juvenile learning in the auditory system of adult barn owls. Nat. Neurosci. 8:93–98.

Lu, W., and Constantine-Paton, M. (2004). Eye opening rapidly induces synaptic potentiation and refinement. Neuron 43:237– 249.

Lyckman, A. W., Jhaveri, S., Feldheim, D. A., Vanderhaeghen, P., Flanagan, J. G., and Sur, M. (2001). Enhanced plasticity of retinothalamic projections in an ephrin-A2/A5 double mutant. J. Neurosci. 21:7684–7690.

Maffei, A., Nataraj, K., Nelson, S. B., and Turrigiano, G. G. (2006). Potentiation of cortical inhibition by visual deprivation. Nature 443:81–84.

Majdan, M., and Shatz, C. J. (2006). Effects of visual experience on activity-dependent gene regulation in cortex. Nat. Neurosci. 9:650–659.

Meister, M., Wong, R. O., Baylor, D. A., and Shatz, C. J. (1991). Synchronous bursts of action potentials in ganglion cells of the developing mammalian retina. Science 252:939–943.

Molinar-Rode, R., and Pasik, P. (1992). Amino acids and N- acetyl-aspartyl-glutamate as neurotransmitter candidates in the monkey retinogeniculate pathways. Exp. Brain Res. 89:40–48.

Monyer, H., Burnashev, N., Laurie, D. J., Sakmann, B., and Seeburg, P. H. (1994). Developmental and regional expression in the rat brain and functional properties of four NMDA receptors. Neuron 12:529–540.

Monyer, H., Sprengel, R., Schoepfer, R., Herb, A., Higuchi, M., Lomeli, H., Burnashev, N., Sakmann, B., and Seeburg, P. H. (1992). Heteromeric NMDA receptors: Molecular and functional distinction of subtypes. Science 256:1217–1221.

Mooney, R., Madison, D. V., and Shatz, C. J. (1993). Enhancement of transmission at the developing retinogeniculate synapse. Neuron 10:815–825.

Mrsic-Flogel, T. D., Hofer, S. B., Creutzfeldt, C., CloezTayarani, I., Changeux, J. P., Bonhoeffer, T., and Hübener, M. (2005). Altered map of visual space in the superior colliculus of mice lacking early retinal waves. J. Neurosci. 25:6921– 6928.

Muir-Robinson, G., Hwang, B. J., and Feller, M. B. (2002). Retinogeniculate axons undergo eye-specific segregation in the absence of eye-specific layers. J. Neurosci. 22:5259–5264.

Naeve, G. S., Ramakrishnan, M., Kramer, R., Hevroni, D., Citri, Y., and Theill, L. E. (1997). Neuritin: A gene induced by neural activity and neurotrophins that promotes neuritogenesis. Proc. Natl. Acad. Sci. U.S.A. 94:2648–2653.

Nedivi, E., Fieldust, S., Theill, L. E., and Hevron, D. (1996). A set of genes expressed in response to light in the adult cerebral cortex and regulated during development. Proc. Natl. Acad. Sci. U.S.A. 93:2048–2053.