Synaptic plasticity and the organization of behaviour after early and late brain injury

Synaptic plasticity and the organization of behaviour after early and late brain injury

Bryan Kolb

Abstract Hebb proposed that synaptic change underlies behavioural and cognitive plasticity. When applied to recovery from brain injury, the general hypothesis is that if there is recovery following brain injury, then there ought to be a correlated synaptic change, which is presumed to be responsible for recovery. In contrast, if recovery fails to occur, or expected recovery is blocked in some manner, then the synaptic change will likely not be present. Systematic study of functional recovery and synaptic change following brain injury at different ages supports these predictions. Good recovery is always correlated with enhanced connectivity whereas poor recovery is always correlated with an absence of reorganized connectivity. Furthermore, factors that stimulate recovery, such as neurotrophins or experience, stimulate synaptic change and functional recovery. Factors that retard recovery, such as depletion of neuromodulators, also block synaptic change. These results thus support Hebb’s general idea that synaptic plasticity is related to behavioural change.

Introduction

Hebb proposed that synaptic change underlies behavioural and cognitive flexibility. There was little evidence of this in 1949, and although the concept of the “Hebb synapse” has become popular in the last decade, there is still rather little direct evidence that synaptic plasticity is associated with processes such as learning or memory, except perhaps in the simplest model systems. Nonetheless, the concept has considerable appeal and remains a viable hypothesis for behavioural flexibility. I will present evidence from studies of the anatomical bases of recovery from brain injury that support the general hypothesis that synaptic change can support behavioural change.

ASSUMPTIONS

As I begin, I must first admit to several biases (see also Kolb, Forgie, Gibb, Gorny, & Rowntree, 1998). First, I assume that the nervous system is conservative. Thus, general mechanisms that are used for one type of behavioural change, such as learning and memory, may also be used for other types of behavioural change, such as in recovery from brain injury. (This assumption does not preclude separate mechanisms too, but it provides a rationale for what to look for and where to look for it.) Studies of functional recovery have the advantage that perturbations of the brain at different times from birth through to ageing produce consistent differences in behavioural outcome ranging from almost complete recovery of function after cortical injury to truly devastating behavioural loss. This variation in behaviour is useful for it is reasonable to suppose that these behavioural differences are related to differences in anatomico-physiological response(s) as well.

Second, I am a “cortex chauvinist” and assume that the changes in the cerebral cortex form the principal mechanism for cognitive change. This assumption comes from several lines of evidence. For instance, it is generally agreed that the relative increase in cortical volume across mammalian evolution is associated with increased cognitive capacity. It follows that changes in cognitive functions in a particular mammal likely will involve changes in cortical structure or organization. Furthermore, studies of decorticated rats show that although they are capable of a remarkable behavioural repertoire (e.g., Whishaw, 1990), there is virtually no recovery or sparing of function from decortication under conditions that would normally lead to marked recovery after restricted cortical removals. For example, Whishaw and I have found that whereas removal of frontal, motor, or parietal cortex at 7-10 days of age allows dramatic sparing of function relative to similar injury in adulthood, there is no sparing at all after complete neonatal neodecortication at 7 days (Kolb & Whishaw, 1981a, b). Further, there is no obvious evidence of anatomical reorganization in the neonatal decorticates (Kolb et al., 1986). Finally, there are marked interspecies differences in the details of cortical organization, such as in old-world and new-world monkeys, and it has been assumed that these differences reflect the clear differences in perceptual and cognitive abilities (Kaas, 1987).

Third, I assume that the most likely mechanism of cortical plasticity will be found at the synapse and that synaptic changes can be measured either pre- or postsynaptically. Traditionally, the emphasis in the literature has been upon the presynaptic side. For example, in the now well-known studies of the effects of unilateral entorhinal cortex lesions in rats, various investigators have shown a major reorganization of the remaining hippocampal afferents (e.g., Steward, 1991). Similar inferences have been made in other models, such as in studies of cholinergic outgrowth after cortical injury (e.g., Cuello 1994), collateral sprouting after peripheral nerve crush (Diamond, 1988), and terminal sprouting after various types of central injuries (e.g., Flohr, 1988). One difficulty with studying presynaptic changes is that they are very difficult to locate unless one knows a priori where to look. In addition, once found, they are difficult to quantify, which is a requisite if one is to correlate with behaviour.

An alternate way to look at synaptic change is to study the postsynaptic side. This requires that the cell body and dendritic tree be stained, such as in a Golgi-type stain. Since the dendritic surface receives more than 95% of the synapses on a neuron, it is therefore possible to infer changes in synapse number from measurements of dendritic extent and spine density. One clear advantage of this measure is that one need not know a priori where to look since it is possible to stain, and examine, the entire cortical mantle. In addition, analysis requires only a light microscope (and a lot of time!). Golgi analysis of the postsynaptic side has proven useful in several types of studies of cortical plasticity (see Figure 1). For example, various groups have shown that housing animals in “enriched” environments leads to increased dendritic outgrowth (e.g., Greenough, Black, & Wallace, 1987). Similarly, training animals in specific tasks leads to dendritic changes in specific populations of neurons (e.g., Greenough, Larsen, & Withers, 1985). One inescapable conclusion of postsynaptic studies is just how plastic dendritic (and presumably synaptic) structure is. One example is especially intriguing. Purves and Hadley (1985) were able to label cells in vivo in the dorsal root ganglion of mice. The dendritic field was mapped. The same cell was relabeled at different times ranging from a few days to weeks and it was possible to see obvious qualitative changes in dendritic extent, which can be taken as at least suggestive evidence of synaptic plasticity. Perhaps the most surprising aspect of the Purves and Hadley study was that the dendritic morphology was so changeable in absence of any particular training. One could reasonably expect even greater change in an animal that was given special somatosensory-related training or possibly peripheral nerve injury.

As we began our studies of synaptic change following cortical injury, we chose to use the Golgi analysis since (1) we did not really know where to begin to look and (2) it was important that we be able to quantify our observations. When we began we had only a single previous study to go on and, unfortunately, this study was not encouraging if one hoped to see dendritic outgrowth correlated with recovery of function. Thus, Jones and Thomas (1962) showed that transection of the olfactory tract produced atrophy of the dendrites of pyriform cortex cells. Nonetheless, they had shown that dendritic analysis of a cortical area could show an effect of a distal injury, which was at least mildly encouraging.

BRAIN DAMAGE AS A MODEL OF CORTICAL PLASTICITY

One of the major accomplishments of neuropsychology in the 1800s was the establishment that language was normally located in the left hemisphere. It therefore came as a major surprise when it was reported as early as 1868 that damage to the left hemisphere of children did not necessarily lead to permanent language deficits. In fact, apparently even Broca was aware of this possibility as he wrote: “I am convinced that a lesion of the left third frontal convolution, apt to produce lasting aphemia (aphasia) in an adult, will not prevent a small child from learning to talk” (Finger & Almli, 1988, p. 122). Because there was little doubt that language ought to have been affected by the injuries, it was reasonable to presume that there had been a fundamental change in the cortical organization of these children. Studies a century later have shown this to be a correct assumption. For example, Rasmussen and Milner (1977) showed that depending upon the precise age at injury, the language zones of the left hemisphere could move either to the right hemisphere or to an alternate region of the damaged left hemisphere. Because similar reorganization is not observed in adults, even after decades of recovery time, it follows that the cerebral response to injury in children must differ in some fundamental way from that in adults. This possibility was first studied systematically in the laboratory by Kennard in the 1930s. She made unilateral motor cortex lesions in infant and adult monkeys. The impairments in the infant monkeys were milder than those in the adults, which led Kennard to hypothesize that there had been a change in cortical organization in the infants, which supported the behavioural recovery. In particular, she hypothesized that if some synapses were removed as a consequence of brain injury, “others would be formed in less usual combinations” and that “it is possible that factors which facilitate cortical organization in the normal young are the same by which reorganization is accomplished in the imperfect cortex after injury” (Kennard, 1942, p. 239). Although Kennard had much to say regarding the limitations of functional recovery after early brain injury (see a review by Finger & Almli, 1988), it was her demonstration that the consequences of motor cortex lesions in infancy were less severe than similar injury in adulthood that is usually associated with her name, and is commonly referred to as the “Kennard principle.”

Kennard was aware that early brain damage might actually produce more severe deficits than expected, but it was Hebb (1947, 1949) who focused on this issue. “It appears …, that an early injury may prevent the development of some intellectual capacities that an equally extensive injury, at maturity, would not have destroyed… Physiologically, the matter may be put as follows: some types of behaviour that require a large amount of brain tissue for their first establishment can then persist when the amount of available tissue has been decreased. This of course is consistent with the theory of cell-assemblies… It has been postulated that, with the enlargement of synaptic knobs, the number of fibers necessary for transmission at the synapse decreases. In the first establishment of an assembly, then, more fibers are necessary than for its later functioning” (Hebb, 1949, p. 292-293).

The difference between the view of Kennard and Hebb is important in the current context for it provides an important starting point for studies looking for a relationship between synaptic change and behaviour. Thus, whereas Kennard hypothesized that recovery from early brain damage was associated with a reorganization into novel neural networks, Hebb postulated that the failure to recover was correlated with a failure of initial organization. Hebb also made an important point regarding recovery from cerebral injury in adulthood. He expected that recovery would be possible since he believed that it would take less neural tissue to support at least certain types of behaviour, once learned, than for the initial learning. Hebb recognized, however, that the functioning of partially damaged cell assemblies might be less reliable and more easily subject to disruption.

We are left with several behavioural and anatomical predictions that follow from both Kennard’s and Hebb’s hypotheses. First, from Kennard, we should be able to demonstrate functional recovery of some types of behaviours after injury during development. Second, from Hebb, we should be able to show a worsened effect from brain injury at other times during development. Third, from Hebb, there should be some recovery after injury in adulthood, but this recovery may be specific to particular environmental constraints. Finally, it follows from both Kennard and Hebb that we ought to be able to identify a synaptic correlate of these three functional outcomes and, as a corollary, if we are able to manipulate the synaptic correlate, function should also be affected. In sum, studies of the functional and anatomical consequences of cerebral injury at different ages ought to provide a window on the general mechanism by which synaptic plasticity is related to behavioural and cognitive flexibility.

Functional Recovery after Cortical Injury Neuropsychological studies play an especially important role in the study of recovery from brain damage because the functional outcome from injury will define the times that we can reasonably expect to see synaptic change. The simple correlation between behaviour and anatomy in an isolated situation is meaningless unless conditions that influence behaviour also influence anatomy. I will therefore begin by outlining some of the variables that appear to constrain behaviour after cortical injury. With these variables in mind, I shall then consider the anatomical correlations and how they respect the behavioural outcomes.

EARLY AGE AND BRAIN DAMAGE

As predicted from both Kennard and Hebb, age is the variable that has the greatest effect upon behavioural outcome. The precise relationship between age and behaviour is not so easily predicted, however. Consider the following example. My colleagues and I removed the frontal cortex of rats at various postnatal ages, and then studied the behaviour of the animals in adulthood (e.g., Kolb, 1987; Kolb & Whishaw, 1981b). Figure 2 illustrates the Morris water task. Animals are required to find a partially submerged platform in a large tank of water. Normal rats learn this task quickly and can find the platform within a few seconds, regardless of the starting location. Rats with frontal cortex removals in adulthood are impaired at this task, but can eventually master it (e.g., Kolb, Sutherland, & Whishaw, 1983). In contrast, rats with cortical lesions in the first 5 or so days of life are severely impaired and never learn the location of the platform (Figure 3). Surprisingly, however, rats with lesions around 7-10 days of age show remarkably normal behaviour relative to littermate controls or adult operates. Hence, it appears that the Kennard effect is true at about 10 days of age whereas the Hebb prediction holds for earlier lesions. This result is not specific to a particular cortical injury as we have seen this outcome from prefrontal, motor, parietal, and occipital lesions (e.g., Kolb, 1987; Kolb, Holmes, & Whishaw, 1987; Kolb, Ladowsky, Gibb, & Gorny, 1996). A parenthetical comment is in order at this point. There is a temptation to see birth as a constant time when comparing results across species such as rats and monkeys. Birth date is a red herring here, however, because the rat brain is embryologically much younger than the cat or monkey. Thus, evidence of sparing after cortical injury in newborn monkeys does not impugn the finding that newborn rats are devastated by cortical injuries. The appropriate rat-monkey comparison is probably more like 10 postnatal days in the rat versus birth day in the monkey (see Figure 4).

In the course of studying the functional effects of early cortical injury in rats, we noticed that brain size in adulthood was directly related to the age at injury: the earlier the injury, the smaller the brain. Thus, rats with perinatal lesions have very small brains whereas those with lesions at day 10 have larger brains. Curiously, however, the day-10 brains still are markedly smaller than the brains of rats with lesions later in life, such as day 25, even though the behavioural outcome is far better (Kolb & Whishaw, 1981a). Therefore, it must be the organization of the brain rather than its size that predicts recovery in the day-10 animal. As our evidence of an “optimal” time for good functional outcome after cortical injury accumulated, it was our tacit assumption that we were observing sparing of function. That is, like Kennard and many others before us, we assumed that animals with injuries sustained before specific behavioural functions were present simply developed the behaviours as though there had been no injury. This phenomenon is usually called sparing of function to distinguish it from recovery of function, which represents a return of a behaviour that was lost and then improved. We were stimulated to test this hypothesis directly by the evidence by Rudy, Stadler-Morris, and Albert (1987) showing that infant rats could learn various spatial navigation tasks in early adolescence. We assumed that rats with frontal lesions around day 10 would perform as well on these tasks relative to intact littermates as they do in adulthood, whereas rats with similar lesions around the day of birth would be relatively as poor at this age as when they were assessed later in life. From this result we could predict that the neural changes supporting sparing were present early. Our behavioural prediction proved to be incorrect, however. Rather, what we found was that rats with frontal lesions at days 1 or 10 were equally impaired relative to littermate controls when assessed at 20 days of age, but that when the animals were assessed at 60 days of age the day-10 animals showed the improved performance that we had found earlier (Kolb & Gibb, 1993). In other words, the animals showed recovery rather than sparing. This result is important for our search for neural correlates for it implied that the synaptic changes underlying the behavioural recovery were slow in developing and that they should not be present at 20 days but should be present at 60 days.

Although we had shown that brain injury during the first week of life in the rat was particularly disruptive of later behaviour, we wondered whether brain injury earlier yet in life would have even more dire consequences. To test this we made prenatal lesions of the presumptive frontal cortex on embryonic day 18. Hicks (1960) had shown that irradiating the developing rat brain on day 18 had severe consequences upon brain and behavioural development and we anticipated a similar result from our focal ablations. Again, we were wrong. Animals with prenatal lesions were indistinishable from control animals when we studied them in adulthood (Kolb, Cioe, & Muirhead, 1998). However, like rats with cortical lesions on postnatal day 10, the prenatal operates had smaller than normal brains. Again, it was not the size of the brain that was critical but its synaptic organization.

RECOVERY AFTER CORTICAL INJURY IN ADULTHOOD

Having found that there are periods in development when behavioural outcome is particularly poor or good leads to the question of what happens after injury in adulthood. Over the past decade there have been various claims of rather dramatic improvement in behavioural outcome after particular treatments such as amphetamine (e.g., Feeney & Sutton, 1987) or cortical transplants (e.g., Kolb & Fantie, 1994) but there have been few systematic studies of the evolution of behavioural outcome after cortical injury. One of the fundamental problems in such studies is a definition of what constitutes recovery (see reviews by Fantie & Kolb, 1991; Kolb, 1990, 1992; Stein & Glasier, 1992). For our current purposes I shall define recovery as a change in behaviour over time and I shall illustrate an example. Figure 5 illustrates the change in forepaw reaching ability in rats with restricted motor cortex injuries. The animals were first trained to reach through bars to obtain small pieces of food. They were then given motor cortex lesions and were retested at different intervals over the following three months (Rowntree & Kolb, 1997). The results showed a slow improvement in performance in the “endpoint” measure of success in reaching. The performance of the animals never returned to normal levels and, perhaps more importantly, when similar animals have been examined in other experiments by Whishaw and his colleagues, it has been shown clearly that the animals make markedly different movements (Whishaw, Pellis, Gorny, & Pellis, 1991). Thus, the apparent recovery in the reaching accuracy reflects not only some behavioural recovery but also behavioural compensation. Nonetheless, there is a behavioural change over time and this change is presumably associated with some kind of neural modification. We have found parallel results in the Morris water task performance of rats with frontal cortex removals in adulthood. Rats with large frontal lesions are very poor at this task, in large part because their initial strategy of swimming around the perimeter of the pool is completely ineffective in finding the platform, which is located well away from the pool wall. Thus, when rats are trained a month or so after cortical injury they often fail to find the platform and swim around and around the pool until rescued. In contrast, however, when the same animals are trained six months later they are able to learn the task and can eventually perform surprisingly efficiently, although their navigation is never as accurate as control rats. Thus, we have evidence of behavioural improvement in a “motor task” such as forepaw reaching as well as in a “cognitive task” such as the performance of a spatial navigation task. In neither case are the animals as proficient as intact control animals, but there is a clear improvement over time. One explanation for the behavioural improvement is that the animals have adapted to the cortical injury by using other, remaining cortical circuitry. The question is whether this circuitry has been modified in order to facilitate this behavioural improvement.

RECOVERY IN THE AGEING BRAIN

There is considerable anecdotal evidence that aged humans show poorer recovery from brain damage than do younger subjects. There is little direct evidence of this, however. One difficulty is that it is difficult to equate the nature of cerebral injury in young and elderly people. A second is that older people may not have the same level of preinjury functioning as younger people so recovery must be seen as relative to the starting point. One of the few studies that has directly compared the effects of similar brain injury at different ages was a comparison of recovery from missile wounds to the brain in WWII soldiers. Teuber (1975) showed that men aged 17-20 showed better recovery than those aged 21-25, who in turn showed better recovery than those aged 26 or older. Twenty-six-year-olds are hardly aged but the marked decline in recovery associated with age in adult men is suggestive (and depressing!).

In order to examine the possibility that aged animals might not fare as well as younger ones, we compared the effects of frontal injury in rats that were young middle age (6 months) to those that sustained frontal injuries late in their life span (24 months). The results were dramatic as the old animals performed much worse than the younger animals, suggesting that their brain was less able to compensate (Kolb & Gibb, unpublished observations). One real difficulty, however, was that not only did the animals perform poorly in neuropsychological tests, but their general health appeared more affected by the cerebral injury, much as one might expect in aged humans. One way to get around this problem is to approach the study of ageing and brain damage in a different way. One can induce the brain damage earlier in life and then observe the functional changes, if any, as animals age. Such an experiment was done by Schallert (1983). He gave rats unilateral lateral hypothalamic lesions in adulthood and followed their sensorimotor responses for the next two years. He found that although the animals initially showed a dramatic reduction in responsitivity to the side contralateral to the lesions, they gradually recovered over the next 60 days. Over time, however, the tactile asymmetry reemerged in the elderly brain-injured rats. A parallel finding has been reported by Corkin (1989) who showed that the veterans studied by Teuber also showed an unexpected decline in recovered cognitive abilities as they aged (but see Newcombe, 1996). One conclusion from these studies is that the neural modifications that supported the relatively good behavioural outcome earlier in the life of these animals were either susceptible to ageing or perhaps were similar to the ones normally used in ageing, and the animals thus were unable to make these changes during ageing.

EXPERIENCE AND RECOVERY

It is widely assumed that the outcome of brain damage is influenced by experience, such as in rehabilitation programs. Indeed, there is a developing field of “cognitive rehabilitation,” especially in the treatment of closed head injuries. Although there are certainly anecdotal studies of semimiraculous recoveries after various rehabilitative programmes, there is little systematic study of what type of rehabilitation will be most effective, when the treatment ought to be instituted, or what type of injury is most affected by recovery. We have begun to examine this issue both to determine if and when treatment might be most effective but also because evidence of a positive effect of treatment at one time and not at another implies that the brain has responded differently in the two cases.

In our first studies we took advantage of the extensive literature showing that raising rats in relatively complex and stimulating environments, such as large pens filled with various toys that were changed regularly, had a beneficial effect on the performance of cognitive tasks relative to animals housed in relatively impoverished conditions such as laboratory cages. It was our expectation that following brain damage housing in the complex environments would enhance recovery regardless of the age at injury. We were mistaken. We compared the benefits of three months of postoperative housing in rats with frontal lesions at day 1, 5, 7, 10, or 120 (adulthood). There were very small (if any) benefits in the performance of the Morris water task in the animals with lesions on postnatal days 7 or 10. These animals showed significant recovery without the complex housing and the treatment did not provide a further benefit. In contrast, the animals with the earliest injuries, which were the ones with the least spontaneous recovery, showed dramatic recovery after the complex housing. Thus, it appears that the environmental “therapy” was most effective in the animals with the worst behavioural outcome. This result has clinical implications for it might be predicted that children with the earliest cortical injuries will show the most benefit from behavioural therapies. It might also follow that older subjects, who also show poor behavioural outcome, would also show a disproportionately large beneficial effect of the therapy.

One additional experiment is relevant here. We had shown in a series of studies that merely stroking rat pups with a little paint brush for 15 minutes three times a day during the first two weeks of life could stimulate changes in the brain and could improve skilled motor learning in adulthood (Kolb, Gibb, & Gorny, 1999). We therefore wondered if this treatment might stimulate recovery in rats with the earliest cortical injuries. It did. Animals with frontal lesions on postnatal day 1 showed marked recovery of function with just two weeks of paintbrush stimulation. This result showed that behavioural therapies introduced very early in life could have major consequences for later behaviour.

SUMMARY

I have assumed that at least some similar mechanisms are likely to be present in most instances of functional recovery after cortical injury. If this is so, then our behavioural studies have imposed several constraints on our search for a synaptic correlate of functional recovery. First, a synaptic change should be found in animals with prenatal lesions or postnatal lesions at 7-10 days of age as well as in young adulthood, but this change should not be present in animals with cortical injuries sustained in the first few days of life or in aged animals. Second, the synaptic changes that support recovery should develop slowly over time in both the young (day-10) and adult animals. Third, if behavioural therapies act on the same synaptic processes, then paint brush stroking or complex housing ought to have greater effects upon the brains of the perinatal operates because they showed the greatest behavioural advantage. Fourth, because the presence of even small injuries contralateral to a large cortical injury appears to compromise recovery, one might expect that the synaptic changes are reduced in animals with bilateral as opposed to unilateral injuries. All of these predictions prove to be correct.

Synaptic Change after Cortical Injury

As I have indicated, we chose to focus upon the changes in dendrites as a measure of synaptic change. Thus, at appropriate postinjury times the brains of rats with cortical injuries were removed and stained with Golgi-Cox (Gibb & Kolb, 1998). The pyramidal neurons of layer II/III or v in different cortical regions are identified and drawn (see Figure 6). The extent of dendritic arborization can be quantified in several different ways. Consider the following botanical analogy. If one examines a tree it can be seen that for a given species there is a typical pattern of branch growth. Thus, there might be 20 primary branches originating from the trunk. Each of these primary branches will have secondary branches, which in turn have branches (tertiary) and so on to the last, or terminal, branches. Now consider what might happen if the tree were treated in some way, such as with fertilizer. The tree could grow longer existing branches, new primary branches, new terminal branches, etc. The growth of new branches is unlikely to begin at the trunk, however, and it is more likely to occur toward the tips of the branches. Regardless of the particular change, however, there would be room for more leaves. Alternately, another treatment, such as drought, could lead to branch death or leaf loss. A similar logic can be used in the analysis of branches in neurons. In this case the goal is to estimate the total space for synapses, and to consider the topology of the changes in arborization.

Figure 1 illustrates a typical pyramidal cell, showing the characteristic apical dendrites, which orient towards the cortical surface and tend to cross different cortical laminae, as well as the basilar dendrites, which emanate from the cell body and tend to stay within the same laminae as the cell body itself. The neuron in Figure 1 also shows the dendritic spines visible on the branches. If we return to our botanical analogy, we could imagine the “spine-equivalents” to be leaves. In this case we would likely predict that the density of leaves would vary with the location on the branch. In the case of a tree, one would probably predict relatively low numbers of leaves on the trunk and on the branches close to the trunk, and higher numbers of leaves towards the branch tips. This can be described as “leaf density,’ which could be quantified easily. A similar logic could be used with the dendritic spines. Spine density can be calculated for different parts of the neuron, which would both imply synapse numbers as well as being a potential measure of treatment effects on the neuron.

Both branch and spine measures can be compared with behavioural outcome. There are two important advantages to dendritic measures in behavioural studies. First, they can be used to infer changes in synaptic organization following a treatment. This could be done more directly by counting synapses but this is very very laborious and impractical for large samples of subjects. Second, it is possible to correlate behavioural outcome with the anatomical measures both in groups and in individual animals. This is very difficult to do using other measures, such as changes in presynaptic terminals, which are not easily quantified.

Studies of dendritic changes in brain-damaged animals have shown two general types of outcomes. First, when cells are deafferented they lose synapses and this is correlated with a loss of spines and a retraction of dendritic arbor. Second, when cells establish new synapses they show an increase in dendritic spines and arbor (e.g., Steward & Rubel, 1993). Our expectation was that if synaptic plasticity underlies behavioural change then several things should occur. First, behavioural recovery, such as in the Kennard effect, ought to be correlated with increased dendritic arbor and/or spines. A corollary of this would be that a failure to show the Kennard effect ought to be correlated with an absence of the dendritic change, or perhaps even atrophy as was found in the Jones and Thomas (1962) study. Second, the temporal changes in behaviour should be correlated with the temporal changes in dendrites. Third, treatments that alter the behavioural outcome should have a corresponding effect on the dendritic morphology and vice versa. All of these predictions are confirmed.

DENDRMIC CHANGE AFTER EARLY CORTICAL INJURY

Recall that we found a different behavioural outcome at different developmental ages. Thus, damage on the day of birth led to a miserable behavioural outcome, whereas damage at 10 days of age led to a good behavioural outcome with significant functional recovery. We therefore expected to see some type of difference in the dendritic response to injury at day 1 and day 10. Indeed, we found enhanced dendritic arborization in remaining cortex of adult rats which had sustained frontal lesions at day 10 but not in those with injuries on day 1 (Figure 6). Rats with lesions at day 10 showed a significant increase in both dendritic length and spine density from day 20 to 60, whereas normal control animals or rats with day-1 lesions did not. In addition, there was a dramatic increase in spine density in the day-10 operates compared to a decrease in the day-1 operates. One difficulty with experiments that alter behaviour differentially at different ages is that the behaviour could change the brain. In this case an anatomical difference between day-10 and day-1 lesions is not the cause of the behaviour but vice versa. In order to assure ourselves that this was not likely, we prepared animals with serial lesions in which one hemisphere was damaged at one day of age and the other at 10 days of age. The dendritic changes in the two hemispheres was visibly different, as illustrated in Figure 6. We therefore concluded that the different behavioural outcomes in the day-1 and day-10 operates might be due to a difference in synaptic organization of the remaining cortex (Kolb, Gibb, & van der Kooy, 1994).

At this point we really had no idea when the compensatory changes might be occuring during development. However, as I described earlier, we had shown behaviourally that there was no sparing at 20 days of age but there was functional recovery at 60 days. If our hypothesis that the observed synaptic changes were supporting the functional restitution, then we ought to find a dendritic difference between the day-1 and 10 operates sacrificed at 60 days but not at 20 days. This is what we found. It thus appears that the dendritic response is slow to occur after the injury but its appearance correlates with behaviour. In order to test our results we decided to make lesions on postnatal day 7, a time at which there is some recovery of function but it is far more incomplete than after day-10 lesions. Our dendritic analysis showed that animals with day-7 lesions did not have increased dendritic arborization but they did have an increase in dendritic spines. Thus, it appears that there may be two independent mechanisms of synaptic change after early cortical lesions: increased dendritic arborization and increased spine density.

DENDRITIC CHANGE AFTER CORTICAL INJURY IN ADULT ANIMALS

Our first experiments looked at the changes in the motor cortex of rats with unilateral injuries. As I noted earlier, such rats show some behavioural improvement, which stabilizes at about three weeks after injury. We sacrificed animals three weeks after such lesions and examined the branching pattern of interneurons in layers II and III of the adjacent sensorimotor cortex on the lesion side and the same tissue on the intact side. Our analysis of these cells showed an increase in branching proximal to the injury but no change in the intact hemisphere. Thus, once again there appeared to be a relationship between functional recovery and synaptic change (Kolb & Whishaw, 1988). More recently, Jones and Schallert (1992) approached this same model somewhat differently. They noticed that rats with unilateral motor cortex lesions strongly favoured the use of the paw ipsilateral to the injury and they wondered if this behavioural change might be correlated with a change in the output neurons. Hence, they looked at the large pyramidal cells in layer v, which are the output cells, and found a increase in branching in the good hemisphere. Importantly, the increase in dendritic branching in the normal hemisphere declined over time as the animals recovered use of the other paw. Taken together, our results and those of Jones and Schallert suggest that dendritic change can support recovery in adult rats, much as it can in infant rats. This hypothesis is supported by one additional experiment in which we gave rats with large frontal lesions 7 days, 1 month or 4 months to recover from frontal lesions and then measured dendritic arborization (e.g., Kolb & Gibb, 1991). There were two principal results. First, there was neither functional recovery nor dendritic growth after either 7 days or 1 month. Second, in contrast, 4 months allowed at least some functional restitution and a small, but significant, increase in dendritic arbor (Figure 7). Although the dendritic response was much smaller than the changes after day10 lesions, this is consistent with the functional outcome, which is reassuring! Third, in contrast to the effects of day10 frontal lesions, which led to enhanced dendritic arborization throughout the remaining hemisphere, adult frontal lesions only increased dendritic arborization in the adjacent sensorimotor cortex. There was no detectable change in posterior cortex, even with extended recovery periods.

DENDRITIC CHANGE AFTER CORTICAL INJURY IN AGEING ANIMALS

The first question to address in the ageing animals was whether there might be significant change in the dendritic arborization of normal aged animals relative to younger ones. Coleman and Buell (1985) predicted that since there is continual neuron death during adulthood one mechanism for compensation might be an increased dendritic arborization in ageing animals. We tested this possibility by looking in the somatosensory cortex of middle-aged (6 months) and aged (24 + months) rats. The results confirmed Coleman and Buell’s prediction as there was a clear increase in dendritic arbor in the normal-aged animal. Armed with this result we then made frontal lesions in similar-aged rats but, in contrast to young adult or day-10 rats with frontal lesions, we did not find an increase in dendritic branching but actually found a decrease. As I noted earlier, such animals have a very poor behavioural outcome, which is consistent with the failure to show synaptic compensation for the lesion. These results are intriguing for they suggest that: (1) the brain may use a similar mechanism for ageing and recovery from brain injury; and, (2) there is a limit to the plasticity in the cortex.

EFFECTS OF EXPERIENCE

Like many others before us (e.g., Greenough et al., 1987), we have consistently shown that housing animals in enriched environments leads to changes throughout much of cortex increase in dendritic arborization in both visual and somatosensory cortex (e.g., Kolb & Elliott, 1987). It is of some interest, therefore, to ask what effect enriched experience might have on remaining cortical regions in rats with cortical injuries. Our results from ageing animals suggested that there might be a limit to dendritic growth so it was possible that cells that showed increased branching in response to injury might be unable to show a further increase with experience. On the other hand, cells that failed to change after the injury might show a response to enrichment. In fact, these were the findings. Thus, we have seen already that rats with adult frontal lesions and prolonged recovery have an increase in dendritic branching in the sensorimotor cortex adjacent to the injury, but no response in the posterior cortex. Housing such rats in enriched environments caused an increase in branching in the posterior cortex but had no detectable effect on the cells in the sensorimotor cortex, which had already changed in response to the lesion (Kolb & Gibb, 1991). This result suggests that one can enhance branching with either injury or environmental treatment, but not both. If this is true, then one implication is that, functionally, it might be of most benefit to place brain-injured subjects into the enriched environments soon after brain injury so that the injuryinduced changes can be modified by the environment. It may be that once the brain has been altered by the response to injury it is resistant to shaping by environmental events. Note, however, that this assumes that the details of synaptic change from the two factors are similar, but this need not be so. Thus, although it is reasonable to suppose that the brain may use the same general mechanism to accomplish synaptic plasticity in different circumstances, the details of the synaptic changes must certainly be different. In one case there is a synaptic change in response to experience, which presumably represents some kind of “trace” of the experience, and in the other there is a synaptic change in response to injury, which probably represents an alteration of cortical function to at least partly accommodate for the loss of tissue. It is therefore difficult to imagine that the details of synaptic change could be identical in the two cases. In fact, we have evidence that the details might be quite different. Consider the following experiment. Rats were given frontal lesions at day 7 and placed in complex environments or standard cages for 4 months after weaning. Analysis of the dendritic arbor showed once again that quantitatively the lesion and enrichment had similar effects upon dendritic branching. However, to our surprise, they had opposite effects upon spine density: the lesions increased spine density whereas the experience decreased it (Kolb, Buday, Gorny, & Gibb, 1992)! Although this particular result was unexpected, it is sensible to expect that there will many ways to remodel neural circuitry, depending upon the requisite outcome.

THE EFFECTS OF NORADRENALINE DEPLETION

Having shown a correlation between dendritic plasticity and functional recovery led us to wonder what would happen if we could somehow block either the anatomical or functional changes. We would have to expect that if our brainbehaviour correlation meant anything, then if we blocked one the other would follow. One way to approach this problem is to deplete the cortex of “factors” that might be necessary for cortical plasticity. Our first candidate was noradrenaline. The possibility that cortical noradrenaline might promote plasticity in the developing neocortex was first demonstrated by Kasamatsu and Pettigrew (1976). These authors found that depletion of NA from the kitten cortex by intraventricular 6-hydroxydopamine (6HDA) infusion blocked the shift in ocular dominance that is observed in visual cortical neurons after monocular deprivation during development. Although the findings of Kasamatsu and his colleagues have proven controversial, they have stimulated research into the involvement of NA is other situations in which the cortex is modified by experience (e.g., O’Shea, Saari, Pappas, Ings, & Stange, 1983; Uylings et al., 1978; Whishaw et al., 1986) or by cortical injury in infancy (e.g., Sutherland et al., 1982).

Our hunch that cortical NA might influence recovery from early cortical injury proved correct as depletion of cortical NA prior to postnatal day 7 frontal lesions completely blocked sparing of function (Kolb & Sutherland, 1992). Analysis of the dendritic arborization showed that the absence of NA in the operated rats reduced dendritic branching by about 30% relative to saline-treated operates. Thus, we were able to show that removal of cortical noradrenaline blocked functional recovery and this was correlated with a dramatic stunting of dendritic branching. One puzzling result of the noradrenaline experiments was our failure to find a behavioural consequence of NA depletion in unoperated rats. Surely we would have expected some type of behavioural consequence of such a major intervention. Further analysis of the dendritic results found a solution. Unoperated rats that had NA depletion had a significant increase in spine density in cortical pyramidal neurons relative to saline-treated controls. This increased spine density was not present in rats with frontal lesions, again consistent with their failure to show functional recovery.

EFFECTS OF NERVE GROWTH FACTOR INFUSION

Our demonstration that blockade of dendritic growth with noradrenaline depletion was associated with poor functional outcome led us to wonder if it might be possible to enhance dendritic growth and improve functional outcome by a different form of exogenous chemical treatment. Many recent studies have suggested that intraventricular treatment with nerve growth factor (NGF) can enhance recovery from cortical injury (e.g., Fischer, Bjorklund, Chen, & Gage, 1991). We therefore gave rats large unilateral motor cortex lesions and studied the recovery of forelimb reaching. One group of animals received intraventricular injections of nerve growth factor whereas another received only cerebral spinal fluid (Kolb, Cote, Ribeiro-da-Silva, & Cuello, 1997 ). The results were dramatic: NGF treatment markedly enhanced both recovery, dendritic growth and spine density in remaining cortical motor neurons (Figure 8). Thus, we were able to show that blockade of dendritic growth with 6HDA is correlated with poor functional outcome, whereas stimulation of dendritic growth with NGF is correlated with improved functional outcome.

SUMMARY

Our anatomical results are consistent with the hypothesis that a synaptic change is responsible for functional plasticity after cortical injury. The behavioural data defined the conditions under which the anatomical changes ought to be found and for the most part we have found them. First, behavioural recovery is correlated with changes in both dendritic arborization and spine density. Second, the dendritic and behavioural changes are temporally linked. Third, circumstances that block recovery (ageing or NA depletion) are correlated with a decrease, rather than an increase of dendritic arbor. Fourth, treatments that enhance behavioural recovery are correlated with an increase in dendritic branching.

There are still vast areas of ignorance, however, as we still have little idea what is actually responsible for initiating the putative synaptic changes or, indeed, what the changes we have described actually mean functionally. I now turn to these more general issues.

Remaining Issues

WHAT DO POSTSYNAPTIC CHANGES MEAN?

I have shown that functional recovery is correlated with an increase in dendritic space and in both spine density and total spines, depending upon the particular experiment. The next question is what does this mean? My assumption, which I have not tested directly, is that an increase in dendritic space reflects an increase in synapses on the dendrites. If this is so, and I have no reason to believe that it is not, then it follows that the behavioural flexibility associated with recovery of function after cortical injury is correlated with a change in the synaptic organization of the cortex. We now must ask where the presynaptic elements that are synapsing upon the dendrites are coming from. This is not a simple question but there are two likely candidates. First, there may be a change in the afferents to the cortical neurons. These could originate from other cortical regions, the thalamus, or possibly other subcortical structures. It strikes me that these afferents are most likely to represent an expansion of existing afferents rather than the development of novel projections, but I am aware of no evidence either way. Second, there could be a change in the intrinsic organization of the cortex. That is, the vertical processing within a functional “column” of cortex could be altered. There is some appeal to this possibility because this would appear to be the mechanism that evolution has used. Thus, as the brain expands in evolution, the number of cortical neurons in a column of tissue is held constant but the thickness of the cortex increases considerably. This increase thus represents increased axonal and dendritic material, including synapses. One appeal of this notion is that an increase in intrinsic synapses provide a mechanism for increased numbers of cell assemblies in the cortex.

WHAT IS TURNING ON THE SYNAPTIC CHANGES?

Synaptic changes do not simply happen. What is the trigger that tells the dendrites and the presynaptic terminals to grow, and the synapses to form? And, more importantly perhaps, how do appropriate connections form that are functionally beneficial, rather than random connections that may be dysfunctional? The most obvious place to look for signals is in the developing brain. As the brain develops, various classes of agents play a role in promoting or guiding connectivity. Many of these agents are extra neuronal and include: 1) trophic or growth factors (e.g., NGF), which function to keep neurons alive, to direct or to enhance neurite growth, or to facilitate specific protein production; 2) cell-adhesion molecules (CAMs), which function as a directional signal or marker; and, 3) components of the extracellular matrix (e.g., laminin), which provide a substrate for cell migration or neurite outgrowth. Presumably, several such agents will be involved in the pre- and postsynaptic changes as new synapses will require both the growth of axonal and dendritic elements as well as the presence of specific substrates that allow or facilitate the outgrowth.

The identification of the specific factors that might be involved in the dendritic changes that we have observed seems to be a daunting task. There are, however, a few clues. For example, it is now clear that one response to neuronal injury is the growth and probably proliferation, of astrocytes. In vitro studies have shown that agents manufactured by astroctyes are growth promoting, so we might predict that astrocytes play a major role in stimulating, and perhaps directing, neurite outgrowth and synaptogenesis. Recent work by Cotman and his colleagues (Cotman, Cummings, & Pike, 1993) are instructive. They have shown that a factor known as bFGF (basic fibroblast growth factor) is produced by astrocytes after cerebral injury and that it apparently plays a role in promoting new synapse formation after cortical injury. My colleagues and I are currently exploring the role of astrocytes in our frontal cortex model of recovery and dendritic change. Our prediction is that there will be greater astrocyte activity in rats with lesions around 10 days of age than at other times. Indeed, preliminary data suggest that this is the case as rats with day-10 lesions have far more astrocyte activity than do rats with similar lesions on day 1 (Kolb & Gibb, unpublished observations). This observation is consistent with the general hypothesis of astrocyte-mediated initiation of dendritic outgrowth. Furthermore, in a related study we have been able to demonstrate that blockade of bFGF activity after cortical injury in adulthood not only prevents recovery but also blocks dendritic growth (Rowntree & Kolb, 1997).

THE HEBB SYNAPSE REVIS1TED

I began by introducing Hebb’s proposition that synaptic change underlay behavioural and cognitive flexibility. Although we are far from proving Hebb correct, we have started down a road that looks promising. The major questions facing us over the near future will not be Hebb’s question of whether synaptic change can support behavioural change, but rather one of how the synaptic change is stimulated both at a external (environmental) level as well as at an internal (molecular) level.

This research was supported by an NSERC grant to the author. The author wishes to acknowledge the excellent help of Robbin Gibb and Grazyna Gorny in many of the experiments summarized here. Address for correspondence: Bryan Kolb, Department of Psychology and Neuroscience, University of Lethbridge, Lethbridge, AB T1K 3M4 (Telephone: (403) 329-2405; Fax: (403) 329-2555; e-mail: Kolb@HG.ULETH.CA).

References

Coleman, P. D., & Buell, S. J. (1985). Regulation of dendritic extent in developing and aging brain. In C. W. Cotman (Ed.), Synaptic plasticity (pp. 311-334). New York: Guilford.

Corkin, S. (1989). Penetrating head injury in young adulthood exacerbates cognitive decline in later years. Journal

of Neuroscience, 9, 3876-3883.

Cotman, C. W., Cummings, B. J., & Pike, C. J. (1993). Molecular cascades in adaptive versus pathological plasticity. In A. Gorio, (Ed.), Neuroregeneration (pp. 217-240). New York: Raven.

Cuello, A. C. (1994). Trophic factor therapy in the adult CNS: remodeling of injured basalo-cortical neurons. Progress in Brain Research, 100, 213-221. Diamond, J. (1988). Nerve growth factor and the reinnervation of skin after peripheral nerve lesions. In H. Flohr, (Ed.), Post-lesion neural plasticity (pp. 35-47). Berlin: Springer-Verlag.

Fantie, B. D., & Kolb, B. (1991). The problems of prognosis. In J. Dywan, R. D. Kaplan, & F. J. Pirozzolo, (Eds.), Neuropsychology and the law (pp. 186-238). Berlin: Springer-Verlag.

Fenney, D. M., & Sutton, R. L. (1987). Pharmacotherapy for recovery of function after brain injury. CRC Critical Reviews in Neurobiology, 3, 135-107.

Finger, S., & Almli, C. R. (1988). Margaret Kennard and her “principle” in historical perspective. In S. Finger, T. E. Le Vere, C. R. Almli, & D. G. Stein (Eds.), Brain injury and recovery: Theoretical and controversial issues (pp. 117-132). New York: Plenum.

Fisher, W., Bjorklund, A., Chen, K., & Gage, F. H. NGF improves spatial memory in aged rodents as a function of age. Journal of Neuroscience, 11, 1889-1906. Flohr, H. (1988). Post-lesion neural plasticity. Berlin: Springer-Verlag.

Gibb, R., & Kolb, B. (1998). A method for Golgi-Cox staining of Vibratome cut tissue. Journal of Neuroscience Methods, 79, 1-4.

Greenough, W. T., Black, J .E., & Wallace, C. S. (1987). Experience and brain development. Child Development, 5 539-559.

Greenough, W. T., Larson, J. R., & Withers, G. S. (1985). Effects of unilateral and bilateral training in a reaching task on dendritic branching of neurons in the rat motor-sensory forelimb cortex. Behavioral and Neural Biology, 44, 301-314.

Hebb, D. O. (1947). The effects of early experience on problem solving at maturity. American Psychologist, 2, 737-745.

Hebb, D. O. (1949). The organization of behaviour. New York: Wiley.

Jones, T. A., & Schallert, T. (1992). Overgrowth and pruning of dendrites in adult rats recovering from neocortical damage. Brain Research, 581, 156-160. Jones, W. H., & Thomas, D. B. (1962). Changes in the dendritic organization of neurons in the cerebral cortex following deafferentation. Journal of Anatomy, 96, 375-381.

Kaas, J. H. (1987). The organization of neocortex in mammals: Implications for theories of brain function. Annual

Review of Psychology, 38, 129-151. Kasamatsu, T., Pettigrew, J. D., & Ary, M. (1979). Restoration of visual cortical plasticity by local microperfusion of norepinephrine. Journal of Comparative Neurology, 185, 163-182.

Kennard, M. (1938). Reorganization of motor function in the cerebral cortex of monkeys deprived of motor and premotor areas in infancy. Journal of Neurophysiology, 1, 477-496.

Kennard, M. (1940). Relation of age to motor impairment in man and in subhuman primates. Archives of Neurology and Psychiatry, 44, 377-397.

Kennard, M. (1942). Cortical reorganization of motor

function. Archives of Neurology, 48, 227-240. Kolb, B. (1987). Recovery from early cortical damage in rats. I. Differential behavioral and anatomical effects of frontal lesions at different ages of neural maturation. Behavioural Brain Research, 25, 205-220.

Kolb, B. (1990). Sparing and recovery of function. In B. Kolb & R. Tees (Eds.), The cerebral cortex of the rat (pp. 537-561). Cambridge, MA: MIT.

Kolb, B. (1992). Mechanisms underlying recovery from cortical injury: Reflections on progress and directions for the future. In F. D. Rose & D. A. Johnson (Eds.), Recovery from brain damage (pp. 169-186). New York: Plenum. Kolb, B. (1995). Brain plasticity and behavior. Hillsdale, NJ: Erlbaum.

Kolb, B., Buday, M., Gorny, G., & Gibb, R. (1992). Sex and environment affect behavioral recovery and dendritic growth after neonatal frontal lesions in rats. Society for Neuroscience Abstracts, 18, 869.

Kolb, B., Cioe, J., & Muirhead, D. (1998). Cerebral morphology and functional sparing after prenatal frontal cortex lesions in rats. Behavioural Brain Research, 91, 143-155. Kolb, B., Cote, S., Ribeiro-da-Silva, & Cuello, A. C. (1997). NGF stimulates recovery of function and dendritic growth after unilateral motor cortex lesions in rats. Neuroscience, 76, 1139-1151.

Kolb, B., & Elliott, W. (1987). Recovery from early cortical damage in rats. II. Effects of experience on anatomy and behavior following frontal lesions at 1 or 5 days of age. Behavioural Brain Research, 26, 47-56. Kolb, B., & Fantie, B. (1994). Cortical graft function in adult and neonatal rats. In S. B. Dunnett & A. Bjorklund (Eds.), Functional neural transplantation (pp. 415-436). New York: Raven Press.

Kolb, B., Forgie, M., Gibb, R., Gorny, G.,& Rowntree, S. (1998). Age, experience, and the changing brain. Neurosci ence and Biobehavioral Reviews, 22, 143-159. Kolb, B., & Gibb, R. (1991). Environmental enrichment and cortical injury: Behavioral and anatomical consequences of frontal cortex lesions. Cerebral Cortex, 1, 189-198. Kolb, B., & Gibb, R. (1993). Possible anatomical basis of recovery of spatial learning after neonatal prefrontal

lesions in rats. Behavioral Neuroscience, 107, 799-811. Kolb B., Gibb, R., & Gorny, G. (1999). Experience and the changing brain. I. Experience-dependent changes in dendritic growth vary with age. Manuscript in submission. Kolb, B., Gibb, R., & van der Kooy, D. (1994). Neonatal frontal cortical lesions in rats alter cortical structure and connectivity. Brain Research, 645, 85-97. Kolb, B., Holmes, C., & Whishaw, I. Q. (1987). Recovery from early cortical damage in rats. III. Neonatal removal of posterior parietal cortex has greater behavioral and anatomical effects than similar removals in adulthood. Behavioural Brain Research, 26,119-137. Kolb B., Ladowski, R., Gibb, R., & Gorny, G. (1996). Does dendritic growth underlie recovery from neonatal occipital lesions in rats? Behavioural Brain Research, 77,125-133. Kolb, B., Stewart, J., & Sutherland, RJ. (1997). Recovery of function is associated with increased spine density in cortical pyramidal cells after frontal lesions and/or noradrenaline depletion in neonatal rats. Behavioural Brain Research, 89, 61-70.

Kolb, B., & Sutherland, R. J. (1992). Noradrenaline depletion blocks behavioral sparing and alters cortical morphogenesis after neonatal frontal cortex damage in rats. Journal of Neuroscience, 12, 2221-2330. Kolb, B., Sutherland, R. J., & Whishaw, I. Q. (1983). A comparison of the contributions of the frontal and parietal association cortex to spatial localization in rats. Behavioral Neuroscience, 97, 13-27.

Kolb, B., & Whishaw, I. Q. (1981a). Decortication of rats in infancy or adulthood produced comparable functional losses on learned and species typical behaviors. Journal of Comparative and Physiological Psychology, 95, 468-483. Kolb, B., & Whishaw, I. Q. (1981b). Neonatal frontal lesions in the rat: Sparing of learned but not species-typical behavior in the presence of reduced brain weight and cortical thickness. Journal of Comparative and Physiological Psychology, 95, 863-879.

Kolb, B., & Whishaw, I. Q. (1991). Mechanisms underlying behavioral sparing after neonatal retrosplenial cingulate lesion in rats: Spatial navigation, cortical architecture, and electroencephalographic activity. Brain Dysfunction, 4, 75-92.

Kolb, B., Whishaw, I.Q., & van der Kooy, D. (1986). Brain development in the neonatally decorticated rat. Brain Research, 397, 315-326.

Newcombe, F. (1996). Very late outcome after focal wartime brain wounds. Journal of Clinical and Experimental Neuropsychology, 18, 1-23.

O’Shea, L., Saari, M., Pappas, B. A., Ings, R., & Stange, K. (1983). Neonatal 6-hydroxydopamine attenuates the neural and behavioral effects of enriched rearing in the rat. European Journal of Pharmacology, 92, 43-47. Purves, D., & Voyvodic, J. T. (1987). Imaging mammalian nerve cells and their connections over time in living

animals. Trends in Neurosciences, 10, 398-404. Rasmussen, T., & Milner, B. (1977). The role of early left-brain injury in determining lateralization of cerebral speech functions. Annals of the New York Academy of Sciences, 299, 355-369.

Rowntree, S., & Kolb, B. (1997). Antibodies to bFGF block functional recovery and dendritic compensation after motor cortex lesions. European Journal of Neuroscience, 9, 2432-2442.

Rudy, J. W., Stadler-Morris, S., & Albert, P. (1987). Ontogeny of spatial navigation behaviors in the rat: Dissociation of “proximal:” and “distal” cue-based behaviors. Behavioral Neuroscience, 101, 62-73.

Schallert, T. (1983). Sensorimotor impairment and recovery of function in brain-damaged rats: Reappearance of symptoms during old age. Behavioral Neuroscience, 97, 159-164.

Stein, D. G., & Glasier, M. M. (1992). An overview of developments in research on recovery from brain injury. In F. D. Rose & D. A. Johnson (Eds.), Recovery from brain damage (pp. 169-186). New York: Plenum. Steward, O. (1991). Synapse replacement on cortical neurons following denervation. In A. Peters and E. G. Jones, (Eds.), Cerebral cortex, Volume 9 (pp. 81-131). New York: Plenum.

Steward, O., & Rubel, E. (1993). The fate of denervated neurons: Transneuronal degeneration, dendritic atrophy, and dendritic remodeling. In A. Gorio (Ed.), Neuroregeneration (pp. 37-60). New York: Raven Press.

Sutherland, R. J., Kolb, B., Whishaw, I. Q., & Becker, J. (1982). Cortical noradrenaline depletion eliminates sparing of spatial learning after neonatal frontal cortex damage in the rat. Neuroscience Letters, 32, 125-130. Teuber, H.-L. (1975). Recovery of function after brain injury in man. In Outcome of severe damage to the nervous system, Ciba Foundation Symposium 34. Amsterdam: Elsevier North-Holland.

Uylings, H. B. M., Kuypers, K., Diamond, M. C., Veltman, W. A. M. (1978). Effects of differential rearing on plasticity of dendrites of cortical pyramidal neurons in adult rats. Experimental Neurology, 62, 658-677. Vargha-Khadem, F., Watters, G., & O’Gorman, A. M. (1985). Development of speech and language following bilateral frontal lesions. Brain and Language, 37, 167-183. Whishaw, I. Q. (1990). The decorticate rat. In B. Kolb & R. Tees (Eds.), The cerebral cortex of the rat (pp. 239-268). Cambridge, MA.: MIT Press.

Whishaw, I. Q., Pellis, S. M., Gorny, B. P., & Pellis, V. C. (1991). The impairments in reaching and the movements of compensation in rats with motor cortex lesions: An endpoint, videorecording, and movement notation analysis. Behavioural Brain Research, 42, 77-91.

Whishaw, I. Q., Sutherland, R. J., Kolb, B., & Becker, J. (1986). Effects of neonatal forebrain noradrenaline depletion on recovery from brain damage: Performance on a spatial navigation task as a function of age of surgery and postsurgical housing. Behavioral and Neural Biology, 6, 285-307.

BRYAN KOLB, University of Lethbridge

Copyright Canadian Psychological Association Mar 1999

Provided by ProQuest Information and Learning Company. All rights Reserved