Neural grafting

Neural grafting
One way is to transplant into the damaged brain a new supply of neurons of the missing kind. If the transplanted neurons are themselves taken from a brain at the right stage of development, they will grow in an adult host brain and form new connections, leading at least to a partial restoration of normal function. This has been most convincingly demonstrated in the dopamine system running from the substantia nigra to the caudate-putamen at the base of the forebrain. Destruction of this dopaminergic pathway leads to movement disorders in rats, and to Parkinson’s disease in humans (Hornykiewicz, 1973; Ungerstedt, 1971). Transplants of dopamine cell bodies lead to a clear restoration of some motor functions in the rat (Dunnett et al., 1981), and alleviate some of the symptoms of Parkinson’s disease in human patients (Hagell et al., 1999). If the transplant is made into the site in the substantia nigra where the original dopamine cell bodies would have been located, it will not grow its axons to the original target for the dopamine pathway. So the transplant has to be made into the caudateputamen. This means that the incoming connections to the original dopamine cell bodies will not be made to the transplanted dopamine cell bodies, as these are not where the original cells were located. How do we explain the success of this transplanting procedure? The most plausible explanation is that the implanted dopamine cells make synaptic connections where they can release dopamine in appropriate amounts, although this dopamine release will not be controlled normally by the activities of the remainder of the brain. The success of this procedure also illustrates the key role of dopamine – to enable other neurons in the motor system to operate normally, rather than carrying some specific signals of its own. The dopamine system itself does not seem to give instructions about which muscle to move next.
The capture and transmission of more specific information by transplanted neural tissue has been demonstrated in retina transplant experiments. A retina is transplanted so as to make connections with the brain of a developing rat. In adulthood, the rat can learn to respond to illumination of the transplanted retina as readily as it responds to illumination of its own natural retina (Coffey, Lund & Rawlins, 1989). While these kinds of transplant procedures in no way reconstruct the original circuitry in its entirety, more recent developments offer some hope of coming much closer to this ideal. In these procedures, instead of neurons that have already differentiated into a particular neuronal variety, neuronal stem cells are transplanted. These neuronal precursors have the potential to develop into any kind of neuron. So when they are transplanted into a damaged brain, they migrate to the areas in which cells are missing and form new structures that become integrated into the host brain (Lundberg et al., 1997). It is possible that this kind of technique will result in a far more complete recreation of the missing circuitry (Svendsen & Smith, 1999).
In some birds, the brain region related to memory (the hippocampus) varies in size according to demand. In the Americas, cowbirds parasitize other birds’ nests in the same way that the European cuckoo does, by laying eggs in them to be brought up by foster parents. Successful brood parasites need to know where the hosts’ nests are, and how the egg laying is going in each nest. It is not much use laying your egg after the host’s eggs have been incubated, giving them a head start in the race to hatching, nor is it a good idea to lay your egg before the host has laid any eggs at all. Cowbirds therefore need to keep careful track of what is going on during the breeding season. It is now clear that at this time the cowbird hippocampus increases in size relative to other structures in the brain. In one species of cowbird, only the female keeps track of nest development. In this species, the female’s hippocampus increases in relative size during the breeding season and decreases again afterwards, but the male’s does not. In another species, both male and female keep track of nests, and the hippocampus in both sexes increases for the breeding system. A third species of cowbird is not a brood parasite at all, and in this species the hippocampus shows no sign of growing or shrinking (Reboreda, Clayton & Kacelnik, 1996). In humans, hippocampal damage leads to such a profound amnesia that the patient is more or less incapable of living an independent life. Yet birds appear to need a hippocampus only some of the time. At other times they get rid of it, even though they will have to regrow it next year. Why? It has been suggested to me (by my colleague Professor Sir John Krebs) that the answer lies in energy saving, since the brain uses a great deal of energy, and a reduction in energy load may be vitally important. Small birds in cold climates can lose a significant proportion of their body weight overnight, so even a marginal saving could make a vital difference. Whatever the reason, the bird’s ability provides a striking example of the potential for reforming circuitry in adult brains.
Neurogenesis in adult mammals The potential for neural replacement (neurogenesis) is almost completely absent in the adult mammalian brain – but not entirely absent (Gage, Ray & Fisher, 1995). Intriguingly, one of the two areas of the adult mouse brain that how neurogenesis lies in the dentate gyrus (part of the hippocampus). Elderly mice living in enriched environments show an increase in the numbers of new neurons formed, as well as increased numbers of surviving dentate granule cell neurons (Kemperman, Kuhn & Gage, 1997). Perhaps the mammalian brain more closely parallels the avian brain in its potential for reconstructing central nervous circuitry than we have tended to assume. Recent work shows evidence of neurogenesis in the adult hippocampus of primates, including humans (Eriksson et al., 1998; Kornack & Rakic, 1999). A structural imaging experiment has shown that London taxi drivers – whose job demands an extraordinary knowledge of London streets – have a relatively larger posterior hippocampus compared to age-matched controls (Maguire et al., 2000). The extent of the increase in size correlated with the length of time spent as a taxi driver. It is thus conceivable that constant use of a spatial navigation system has led to growth of the adult human hippocampus, or at least to selective protection from age-related hippocampal shrinkage. One day we may be able to take advantage of this potential, and use it for clinical therapy.

The capacity for neural regeneration

The capacity for neural regeneration
So the eventual wiring of the adult brain in part reflects experiences during the long period of brain development that takes place after birth. And, to some extent, the brain responds to those experiences by making structural changes. However, once the mammalian brain is fully developed, the capacity to form new neurons is drastically reduced, though not totally lost (see below). Even before full development is reached, a lack of input during a critical stage can lead to a permanent loss of appropriate connection. For example, covering one eye during development can distort visual connections, leading to persistent impairment of adult vision that depends on that eye. As little as two weeks of occluded vision can induce these effects in human infants. This has implications for eye surgery procedures in children – for example, placing a patch over the eye after surgery could significantly impair the efficient wiring of the visual system. In this respect, the central nervous system differs from the peripheral nervous system, in which regeneration occurs regularly after injury. Areas of axonal loss can be reinnervated (i.e. the neural connections can be re-established) under some circumstances, to afford a complete recovery of function. But in the central nervous system, spinal cord damage, for instance, leads to permanent paralysis. Christopher Reeve, once the star of Superman films, is now confined to a wheelchair due to spinal cord damage sustained during a riding accident. It is possible that this difference between the central nervous system and the peripheral nervous system lies in the nonneuronal cells that are found alongside neurons. In the central nervous system, these are glial cells; in the peripheral nervous system, they are schwann cells. These non-neuronal cells provide the environment for the neuron, and can clearly secrete a variety of bioactive signalling substances. When peripheral nerves are cut, the portion of axon lying beyond the injury is cleared away, partly by the schwann cells, which form into cylindrical guides along the original path of the axon. New axon processes sprout and spread from the remaining stump, and if one of these processes enters the schwann cell guide tube, then its growth rate increases and it is led along the tube towards the nerve’s original target. Central nervous glial cells do not seem to have this ability to guide regenerating axons.
The question of whether it might be possible to induce the central nervous system to regenerate has taken a new turn since the early 1970s. At this time, it became clear that adult neurons can sometimes form new connections. If one input to a target area is lost, the remaining inputs sometimes send out new branches from their axons to colonize the vacant space (Raisman & Field, 1973). This is not necessarily an advantage. If normal function of the target area depends partly on interactions between two inputs, it may be worse off having a double signal from only one of them than having a normal signal from one and no signal from the other.

NEONATAL BRAIN DEVELOPMENT

NEONATAL BRAIN DEVELOPMENT
When mammals are born, they have to pass through a narrow birth canal, which places a practical limit on neonatal head size. But this does not necessarily limit ultimate brain size, so long as further brain development can take place after birth. This is not too much of a problem for human infants, whose particularly helpless state at birth is made feasible by parental care. Small animals like rats can hide their young safely away in holes. So, like humans, the young can be born immature without incurring excess risk. In contrast, herbivores that inhabit open grassland may need to be able to run with the herd within minutes, or at most a few hours, of their birth. Such creatures could not afford a long postnatal period of general brain development. This problem has been solved in a different way by the kangaroo, which is born very early in its development but remains protected in the safety of its mother’s pouch, where it continues to develop until it is capable of independent movement. Our own protracted postnatal development not only allows us to grow a bigger brain (the adult brain is around four times the size of a new-born baby’s brain). It also ensures that our brain continues to develop while we are interacting with our environment. So each person’s brain will, to a certain extent, be adapted to the circumstances of their lives. Survival of the most useful The first sign of what will become the brain appears very early in human gestation. By the end of the second week, a neural plate made up of precursor neurons can be identified. By the end of the first month, a primitive brain has already formed. Like other parts of the body, the brain develops when cells migrate to the appropriate place. Those cells have to know how to differentiate into the right kinds of eventual cell types, and when to stop differentiating. But brain development requires more than cells simply knowing how to get to the right place and what to do when they are there. For this particular organ, a high level of competitiveness is involved. During development, connections in the brain respond to what is going on. We start off with many more potential neurons than will eventually survive. Neurons compete to make connections with their targets, and it is the connections that are actually used which seem to have a better prospect of survival. Unsuccessful neurons die, through a process of programmed cell death, called apoptosis. If one set of neurons fails to make its normal connections, then another set of opportunist neurons may colonize the vacant space.

The NMDA receptor

The NMDA receptor
One key element in LTP is a particular subtype of glutamate receptor, the NMDA receptor. Calcium entry into the cell is one of the triggers for the development of LTP. The NMDA receptor controls a calcium ion channel that is both transmitter dependent and voltage dependent. This means that even when the NMDA receptor is activated by glutamate, no calcium will pass into a cell through the NMDA-controlled channel unless the target cell has also recently been depolarized. So NMDA dependent LTP can only develop in a cell that has been depolarized and then receives a further input – exactly the conditions that apply during a burst of high frequency stimulation. The mechanism underlying this dual sensitivity to neurotransmitter levels and voltage levels is remarkably simple. It turns out that in cells at their normal resting potential, a positively charged magnesium ion is held in the channel by the electrostatic gradient across the cell membrane. If the NMDA receptor is activated, so that the channel could, in principle, be opened, the inflow of calcium is blocked by the magnesium ion. Once the cell is depolarized, nothing holds the magnesium ion in place, so it can diffuse into the extracellular fluid. If the NMDA receptor is again activated at this point, the ion channel opens and, with the magnesium block removed, calcium can pour into the cell, triggering the series of events that leads to LTP. Of course, as the cell repolarizes, the magnesium ion is drawn back into position once more. The LTP system, particularly in the hippocampus, has been a focus of intense research activity. Rat experiments have shown the blockade of the NMDA receptor by the drug AP5 prevents the development of LTP, and at the same time appears to prevent the normal operation of hippocampus-dependent spatial memory (Morris et al., 1986). More recently still, psychological studies of ‘knockout’ mice, genetically engineered so they can no longer show LTP in the hippocampus, have also shown striking failures of hippocampus-dependent spatial memory tasks, that neatly parallel the effects on LTP (e.g. Reisel et al., 2002). If we can combine these new techniques in molecular biology with sophisticated behavioural analysis, we will have ways to study the relation between brain and cognition at a finer level of detail than has ever been possible before. So we have seen that the adult nervous system is highly modifiable: our brains change in accordance with our experiences. Is the development of our brains modified by the environment too?

LTP and LTD

LTP and LTD
Although there are many ways to modulate neuronal function, they are not usually linked to specific psychological functions. But neuromodulators neurochemicals that indirectly affect neuronal activity, usually by modifying response to other chemical neurotransmitters paracrine non-classical effects of neurotransmitters that may not be released at the synapse, and/or whose receptors are not located at the synapse neurocrine classical neurochemical action of transmitters that are released at the axon terminal to affect specialized receptor sites across the synaptic cleft there is one form of synaptic modifiability that has led researchers to make striking and specific claims, presenting it as a possible neural basis for some forms of learning or the ability to lay down new memories. The first crucial observation was made by electrophysiologists studying the responses of cells in the hippocampus, a structure that is crucial in memory processing. They found that the size of the neuronal response to a single pulse of electrical stimulation at a given intensity could be increased, in a long-lasting way, by giving a relatively brief burst of high frequency stimulation. By comparing the size of the response to a single pulse before and after this high frequency series of pulses, researchers showed beyond doubt that neuronal responsiveness had increased. This change is called longterm potentiation, or LTP. It is now clear that LTP can be seen in many structures in the brain, and not only those thought to be associated with memory. It is highly likely, though, that LTP always reflects experience-dependent changes in neuronal functioning, whether in the sensitization produced by painful stimuli, or in perceptual development in the visual cortex, or in the laying down of memory traces in the brain. It has also become clear that there is a complementary process – long-term depression, or LTD – which describes a decrease in neuronal response. The ability either to increase or to decrease synaptic connectivity as appropriate offers maximum flexibility for adjusting neuronal function.


[Tim Bliss (1940– ) and Terje Lømo (1935– ) first reported the phenomenon of long-term potentiation. The plausibility that a strengthening of synapses might underlie memory storage increased tremendously when the phenomenon of long-term potentiation (LTP) was discovered by these two researchers. In the 1970s, Bliss and Lømo noticed that if they applied a few seconds of high frequency electrical stimulation to certain neurons in the rabbit hippocampus, synaptic transmissions to those neurons would increase in amplitude. More surprisingly, this enhancement seemed to be long-lasting, sometimes persisting for weeks (Bliss & Lømo, 1973). This phenomenon has since been termed long-term potentiation, or LTP. In the twenty years since its discovery, a great debate has raged among neuroscientists about whether this LTP might be the crucial mechanism underpinning learning and memory.]

Neuromodulators and hormones

Neuromodulators and hormones
A still further level of complexity is provided by non-classical neurotransmitter substances. Some of these are released by neurons like the conventional neurotransmitters already described, but they can have longer-lasting actions and (like paracrine neurotransmitters) act at greater distances from their release sites. Some may have no directly measurable effects on their targets, but they may change the target neuron’s responsiveness to its other classical neurotransmitter inputs. There is a more or less indefinable boundary between these substances – often called neuromodulators – and hormones. For example, cholecystokinin (CCK) is a peptide that is released as a hormone by the duodenum (part of the digestive tract), but is also released like a neurotransmitter from dopaminergic neurons in the brain, where it modifies the responses of dopamine autoreceptors. So the same molecule can operate as a neuromodulator in the brain and as a hormone in the gut. Hormones are molecules that are released into the bloodstream from specialized endocrine glands (such as the pituitary gland or the adrenal gland) and can therefore, in principle, act anywhere in the body. For hormones, specificity of action results from the presence of chemically specific receptors on the target structures that are bathed by the bloodstream. The hormonal receptors are activated when the hormones pass by in the blood. Hormones can affect neuronal function in a similar way to neuromodulators, changing sensitivity to other inputs and altering the release of neurotransmitters.

Paracrine effects

Paracrine effects
Target neurons can also have receptors located outside the specialized synaptic region. They are presumably activated either by neurotransmitter that escapes from the synaptic cleft, or perhaps by a transmitter that is itself released outside specialized synaptic regions.

These extra-synaptic routes for chemical communication are sometimes called paracrine systems, as opposed to the more classical neurocrine routes. Their existence adds yet further subtleties to neuronal activity. It is possible that, under certain neuronal conditions, overflow from the synaptic cleft becomes more likely. This overflow could then differentially activate these non-classical paracrine communication routes, potentially producing qualitatively different actions on the target structures.

Autoreceptors

Autoreceptors
We have so far described transmitter being released by the pre-synaptic neuron, crossing the synaptic cleft, binding to postsynaptic receptors, and shortly afterwards being deactivated. In fact, by no means all receptor sites are located on the post-synaptic membrane. Surprisingly, some axons have receptors for their own neurotransmitter – autoreceptors. For example, there are dopaminergic cells with dopamine autoreceptors on their axons. These can be activated by the dopamine that their own cells release, to provide a local, negative feedback loop, which can inhibit the cell from further firing. So input neurons can modify their own activity while activating their post-synaptic targets.

Autoreceptors

Autoreceptors
We have so far described transmitter being released by the pre-synaptic neuron, crossing the synaptic cleft, binding to postsynaptic receptors, and shortly afterwards being deactivated. In fact, by no means all receptor sites are located on the post-synaptic membrane. Surprisingly, some axons have receptors for their own neurotransmitter – autoreceptors. For example, there are dopaminergic cells with dopamine autoreceptors on their axons. These can be activated by the dopamine that their own cells release, to provide a local, negative feedback loop, which can inhibit the cell from further firing. So input neurons can modify their own activity while activating their post-synaptic targets.