Neurogenesis (the birth of new neurons) continues postnatally and into adulthood in the brains of many animal species, including humans. This is particularly prominent in the dentate gyrus of the hippocampal formation. One of the factors that potently suppresses adult neurogenesis is stress, probably due to increased glucocorticoid release. Complementing this, we have recently found that increasing brain levels of serotonin enhance the basal rate of dentate gyrus neurogenesis. These and other data have led us to propose the following theory regarding clinical depression. Stress-induced decreases in dentate gyrus neurogenesis are an important causal factor in precipitating episodes of depression. Reciprocally, therapeutic interventions for depression that increase serotonergic neurotransmission act at least in part by augmenting dentate gyrus neurogenesis and thereby promoting recovery from depression. Thus, we hypothesize that the waning and waxing of neurogenesis in the hippocampal formation are important causal factors, respectively, in the precipitation of, and recovery from, episodes of clinical depression.
Recent developments in cell and molecular biology hold the promise of more effective treatments for the major forms of human psychopathology. By targeting specific molecular sites within brain cells, these approaches provide a precise, powerful and effective means for influencing brain function.
Until now, the treatments of choice for the major psychoses almost universally have been pharmacotherapy. There has been a special focus on those drugs whose primary action is on the chemical communication between neurons (neurotransmitters). Although effective, these treatments nonetheless invariably produce unwanted side effects and typically fall short of the desired level of therapeutic efficacy (either in the individual patient or in the proportion of the patient population).
New approaches are emerging that provide the opportunity, not only for altering chemical communication between neurons, but also of directly changing the basic structure and morphology of the brain. Controlling brain neurogenesis, in the adult, is one of these novel technologies that have the potential to revolutionize treatment of mental illness. Heretofore, the adult brain has been considered largely refractory to regeneration and new neuronal birth (neurogenesis). However, recent studies have clearly revealed that the adult hippocampus, specifically the dentate gyrus (DG), continues to give rise to new neurons throughout life in all mammalian species, including humans. Currently, this approach is limited to those brain areas where neurogenesis occurs normally in adults (ie olfactory bulb and hippocampus), but in the future it may be possible to induce neurogenesis more generally throughout the brain and spinal cord. This would provide a powerful mode of intervention in cases where neurons have been lost to trauma, neurodegenerative disease, or aging.
We begin this review with an examination of the current state of knowledge regarding adult brain neurogenesis. Then we focus specifically on the role of neurogenesis in one form of psychopathology, that of chronic clinical depression. Finally, we look to the future and the opportunities provided by employing this technology for the treatment (and possibly prevention) of other forms of neuro- and psychopathology. We focus this review of neurogenesis on the hippocampus because it is presumed to be more directly relevant to psychiatry than the olfactory bulb.
A primer for adult brain neurogenesis
In attempting to understand the cellular basis of this unique plasticity, it is worthwhile reviewing some terminology that is often used. Stem cells are primitive cells that can divide indefinitely, giving rise to a copy of themselves (self-renewal) and one of a variety of other differentiated cells. There are likely stem cells distinct for different tissues, brain, liver, blood etc, but the exact molecular characteristics of these cells are not known. Progenitor cells are the intermediate daughter cells of the cells that can divide for a limited number of times, and can give rise to a more limited variety of cell types. Thus a cell shown to divide in a culture dish can be either a stem cell or a progenitor, and only by more detailed analysis can one conclude that it is one or the other. For the purposes of this review we will call the dividing cells in the adult brain progenitor cells, with the understanding that they can divide and give rise to neurons and glia, depending on their location and modification in the local environment.
Most neurons in the mammalian brain and spinal cord are generated during the pre- and perinatal periods of development. However, recent evidence has shown that, at least in two brain areas, olfactory bulb and dentate gyrus of the hippocampus, neurons continue to be born throughout life. This has now been found in a variety of species, including humans.1, 2, 3, 4
These new neurons are derived from progenitor cells that reside in the brain's subventricular zone, which lines the ventricles, or in a layer of the hippocampus called the subgranular zone (SGZ). The existing neurons in the adult brain have terminally differentiated and are therefore incapable of further division. However, some progenitor cells appear to remain outside the cell cycle in a somewhat dormant-like stage called (GO). Through a process that is as yet not well understood, a signal(s) induces progenitor cells to enter the cell cycle and eventually go through mitosis (cell division) to produce either two daughter neurons or through asymmetric division to produce one glial cell or neuron and one progenitor cell capable of further division. In adults, it is this process which is suppressed in all other brain areas besides the olfactory bulb and DG. Thus, understanding the mechanisms involved in this process would provide the opportunity for disinhibiting progenitor cells throughout the central nervous system to allow them to produce new neurons.
Progenitor cells in the SGZ, which lies in the DG at the border of the granule cell layer (GCL) toward the hilus, produce progeny which then migrate into the granule cell layer and differentiate into neurons. In this way, these new granule cells are added to the population of existing neurons in this brain region. At this site, these newly born cells begin to mature and send their dendrites into the molecular layer and their cell processes (neurites) into the hilus and along the mossy fiber tract to other structures within the hippocampus such as the CA3 cell fields. These new neurons are thus integrated into the basic circuitry of this part of the brain, thereby mediating information processing from the entorhinal cortex, through the granule cell layer of the DG, to the various cell fields of the hippocampus.
Another critical link in the process is cell survival. It appears that many more cells are produced in the adult DG than the number that ultimately survives. One of the factors involved in determining survival is whether the new cells are sufficiently activated by incoming neural signals, a process that may be termed ‘use it or lose it’. Some of the behavioral, environmental, and biological factors that are known to influence cell proliferation and survival are discussed in the following section.
The number of new neurons generated in the DG seems small (1000–3000 per day in mice and rats) in comparison with the number of existing neurons in the GCL (˜1–2 million). However, when considered over the lifespan, this can represent a substantial proportion, of perhaps 10–20%, of the total population. Furthermore, it is possible that these newly born neurons may serve a preferentially more important role in current information processing than do the extant population of granule cells.
Factors that affect neurogenesis
Behavior and environment
The mechanisms by which new neurons are generated in the DG of the adult hippocampus are not well understood. Recent studies in mice indicate that exposure to an enriched environment (larger housing, varied stimuli, and greater opportunity for social stimulation, physical activity and learning),5 resulted in a significant increase in neurogenesis above control conditions.6 Subsequent studies determined that the most important components of this enrichment were most likely the increased physical activity and learning experience. Similar to enrichment, voluntary exercise in a running wheel increases neurogenesis.7 In addition, running enhances cell proliferation in the DG of the mouse.7 Others have reported that hippocampal-dependent tasks, such as spatial learning, increase the number of surviving BrdU-positive cells in rats.8 Apart from these environmental and behavioral manipulations, several pathological events are known to increase granule cell number: damage to the hippocampus by seizures;9, 10 ischemia;11 and mechanical lesions.12
Stress or increased levels of glucocorticoid hormones inhibit proliferative activity in the DG.13, 14, 15 For example, administration of high levels of corticosterone diminishes cell division in the adult rat hippocampus.16 Exposure of marmoset monkeys to a resident intruder is stressful and results in a decrease in cell proliferation.17 In contrast, adrenalectomy which leads to a reduction in serum glucocorticoid levels, elicits cell division in the DG.14
Drugs and biochemicals
Trophic factors that have been shown to regulate progenitor cell proliferation and survival in vitro, such as FGF and BDNF,18, 19, 20, 21, 22 may also affect these cells in vivo.23 Intracerebro-ventricular infusion of EGF and FGF-2 in rats increased proliferation in the subventricular zone.24 In the SGZ of the DG the mitogenic effect of EGF was more pronounced than that of FGF-2. With regard to differentiation, EGF promoted glial differentiation, whereas FGF-2 did not influence phenotype distribution.24 In canaries, seasonal regulation of adult neurogenesis depends on testosterone levels which mediate their effect through BDNF.25 In addition, both FGF and BDNF are elevated in rodents by voluntary exercise in a running wheel,26, 27, 28 an activity that increases cell proliferation and neurogenesis.7
Effects of hormones on cell proliferation and neurogenesis have been studied by several investigators. As mentioned above, glucocorticoid hormones have been found to have a profound effect on hippocampal cell proliferation.14 Steroid hormones, such as testosterone, enhance neurogenesis in birds whereas estrogen results in a transient increase in proliferation in rats.29 Thyroid hormone can affect neuronal differentiation of hippocampal progenitor cells in vitro.30 In vivo, hypothyroidism interferes with cell migration,31 but does not affect postnatal cell proliferation.32
Neurotransmitters have also been shown to play a role in adult neurogenesis. Systemic injection of glutamate analogs inhibits birth of new cells, whereas glutamate antagonists, such as MK801, enhance cell division.33 Recently, another neurotransmitter has been suggested to play a role in this process. Reduction or inhibition of serotonin reduces stem cell proliferation in the DG,34 possibly through the 5-HT1A receptor.35
Kempermann et al36 found that strains of mice differ with respect to rate of cell division and amount of cell survival and neurogenesis. Comparisons were made between C57BL/6, BalB/c, CD1 and 129/SVJ strains. Proliferation was found to be highest in C57BL/6 mice. However, net neurogenesis was highest in the CD1 strain, whereas 129/SVJ produced relatively more astrocytes and fewer neurons than other strains. The degree to which environmental, behavioral and biochemical factors can affect cell proliferation and neurogenesis may also differ depending on the species or strain of animal involved. Indeed, exposure to an enriched environment had different effects on two of these strains of mice. In C57BL/6 mice, enrichment promoted the survival of progenitor cells but did not affect proliferation, whereas the net increase in neurogenesis in 129/SVJ mice was accompanied by a two-fold increase in proliferation.37
A theory of neurogenesis and depression
Statement of hypothesis
As mentioned above, in a variety of species, stress is one of the important controlling influences on neurogenesis. Stress is also believed to be the most significant casual agent in the etiology of depression (with the possible exception of genetic predisposition) (see Kendler et al for data on this issue and for a brief review of the topic).38 In addition, the hippocampal formation is only one of two brain regions where robust neurogenesis continues into adulthood, and nerve cells in the hippocampal formation are among the most sensitive to the deleterious effects of stress. Considering this evidence together, it is reasonable to propose that a stress-induced decrease in neurogenesis may be an important factor in precipitating episodes of depression.
Reciprocally, increases in serotonergic neurotransmission have been shown to augment hippocampal neurogenesis. Since increases in brain serotonergic neurotransmission are the most effective treatment for depression, it is also reasonable to propose that serotonin-induced increases in neurogenesis may be an important factor in promoting the recovery from depression.
Thus, we suggest that the waning and waxing of neurogenesis in the hippocampal formation are important causal factors, respectively, in the precipitation of, and recovery from, episodes of clinical depression, in the following manner. Exposure to acute stressors produces short-term changes in mood from which a person recovers once the stressor in removed. However, chronic stress exposure and/or endogenous (genetic) factors not only precipitate profound shifts in affect, but concurrently suppress neurogenesis in the DG of the hippocampal formation. This is brought about by increased glucocorticoids and/or decreases in neurotransmitters such as serotonin, which in turn may exert their effects via various neurotrophic factors (described above). Loss of DG neurogenesis disenables throughput in brain circuitry critical for the formation of new cognitions and memories: neocortex → entorhinal cortex → DG → hippocampus. Thus, patients cannot ‘escape’ the psychological impact of the initial precipitating events and remain mired in a chronic depressive state. Decreased DG neurogenesis may also be directly involved in mediating depressive affect because of anatomical connections between the DG and limbic structures such as the amygdala. Recovery from depression necessitates a restoration of the original basal rate of neurogenesis. This may occur spontaneously (due to endogenous changes) or in response to antidepressant therapies such as administration of SSRIs or ECT which augment neurogenesis by increasing synaptic levels of serotonin and other neurotransmitters. In the remainder of this section we provide an overview of the evidence that forms this hypothesis.
Preclinical studies of stress and neurogenesis
As mentioned above, in a series of studies, Gould and colleagues reported that stress suppresses DG neurogenesis and further, that it does so in large part through increases in plasma glucocorticoids. In turn, these glucocorticoids appear to exert their effects via a downstream action on NMDA glutamate receptors.39 (The precise cellular sites of these effects have not been worked out.) One of the important consequences of these changes in the hippocampus is a dysregulation of hippocampal control over the hypothalamic-pituitary-adrenal system. This breakdown of normal negative feedback regulation sets the stage for a pernicious cycle of events involving the hippocampus.
The initial study in this series reported that adrenalectomy increased neurogenesis in the adult rat DG and that this effect could be reversed by corticosterone replacement.14 Thus, it appears that under normal conditions the circulating level of glucocorticoids suppressed the birth of neurons in the DG. In an extension of these results it was found that systemic administration of corticosterone to intact animals suppressed DG neurogenesis.16 In the first direct examination of the role of stress it was found that exposure of a rat to the odor of a natural predator (fox) suppressed cell proliferation in the adult rat DG.40 This was followed by an experiment demonstrating that psychosocial stress in adult tree shrews (exposure to same sex conspecifics) significantly reduced DG neurogenesis.41 In the most recent study in this series, it was reported that the normal rate of neurogenesis, which continues in adulthood in the marmoset monkey, is suppressed by acute stress (placement in a cage occupied by a ‘resident’ marmoset).17 In sum, the evidence clearly shows that stress suppresses the rate of DG neurogenesis in adults of a number of species, and further, that it most likely does so via increases in brain glucocorticoids acting in turn to increase brain glutamate levels.
An additional, but older, literature is also relevant here. Over the past 15 years, work by Sapolsky and McEwen and others has shown that in a number of species stress and/or glucocorticoids also causes widespread morphological changes and even cell death in the hippocampus, eg in the CA3 subfields (for a review, see Sapolsky).42 This region of the hippocampus is the main target of the output of neurons in the DG. Whether this hippocampal damage is at least in part dependent on the suppression of neurogenesis in the DG is not known.
Clinical evidence linking hippocampus to depression
There are a number of different pieces of evidence linking clinical depression to changes in the hippocampus. We emphasize, however, that this is not to suggest that this is the only change in brain associated with depression, nor do we suggest that alterations in the hippocampus underlie all of the phenomenological aspects of depression.
Utilizing the brain imaging technique of MRI, Sheline et al43 reported loss of hippocampal volume in a group of older women with recurrent major depression. The subjects were currently in remission and were screened for comorbidity. When compared to carefully selected controls, depressed subjects had smaller left and right hippocampal volumes in the absence of differences in total cerebral volumes. (This latter finding excludes any of a number of spurious correlations as accounting for these results.) There was also a significant negative correlation between total days of depression and the volume of the left hippocampal gray matter. The authors speculate that this hippocampal loss may be due to glucocorticoid-induced neurotoxicity associated with recurrent episodes of depression. In a more recent study, this same group has confirmed their original report and, further, found that the decrease in hippocampal volume is correlated with total lifetime duration of depression and not with age.44 Another recent MRI study reported similar results in chronically depressed patients, but found no decrease in hippocampal volume in recovered depressed patients.45 Finally, Bremner46 also describes his own unpublished data showing hippocampal atrophy in a group of depressed patients.
Data on Cushing's Syndrome patients may be related to the aforementioned data on depression and hippocampal atrophy because of the elevation of plasma levels of cortisol in both cases.47 Of special relevance for this review is the fact that the incidence of depression in some groups of Cushing's patients may be as high as 50%48 (Sonino and Fava review the various psychosomatic aspects of Cushing's Disease).49 Similarly, when patients with any of a variety of non-psychiatric disorders are treated with glucocorticoids, such as prednisone, this often precipitates psychiatric side-effects, especially in those instances when high drug doses are employed. The mostly commonly noted disturbances in these cases are affective disorders, especially depression (reviewed in Lewis and Smith).50
Temporal lobe epilepsy (TLE) provides another clinical example regarding the issue of hippocampal damage and depression. Depression is the most common psychiatric complication in patients with epilepsy. More specifically, patients with TLE have a higher incidence of depression than patients with other forms of epilepsy or compared to patients with other equally chronically debilitating diseases (see Perini et al for a brief overview of these data).51 It is, of course, well known that TLE is characterized by massive cell loss in various structures in and around the hippocampus (reviewed in Houser).52 Obviously a limitation of these data is the fact that they are both retrospective and correlative (if there is a causal relationship between TLE and depression, there is some evidence to indicate that it may be bidirectional). In addition, since the neuropathology in TLE is widespread throughout the temporal lobe, no definitive conclusion can be drawn regarding the site of specific damage that might underlie the psychopathology. Paradoxically, it has recently been demonstrated that a significant increase in neurogenesis is observed in the dentate gyrus of rats following experimentally induced seizures,9 though there is no clear causal link between this increase in neurogenesis and the persistence of the seizures. This does however raise the possibility that the increases in neurogenesis may sometimes have deleterious effects.
Serotonin and neurogenesis
Serotonin has long been known to be an important mitogenic factor in a variety of peripheral tissues (see Fanburg and Lee for a review).53 In addition, serotonin has been shown to exert a neurogenic effect in the CNS during development.54, 55 Furthermore, in the adult CNS, serotonin plays an important role in neuronal and synaptic plasticity, and its action at the serotonin 5-HT1A receptor is particularly significant in this regard.56
We recently reported that systemic administration of the drug d,l-fenfluramine (a releaser of serotonin throughout the CNS) exerts powerful mitogenic effects in the granule cell layer of the adult DG.35 (We found 2–3 fold increases above the baseline level of mitogenesis.) Importantly, we also found that this fenfluramine effect could be completely blocked by systemic pretreatment with a specific 5-HT1A antagonist (the antagonist also produced a significant decrease in the rate of basal or spontaneous mitogenesis). These studies underline the critical importance of the 5-HT1A receptor in this process. Finally, in an experiment examining longer survival times following administration of 8-OH-DPAT, a 5-HT1A agonist, we found that much of this mitogenic effect was indeed a neurogenic effect, and that activation of the 5-HT1A receptor may augment survival of newly born cells. A recent experiment has confirmed and extended this line of investigation.34 These investigators reported that either depletion of brain serotonin or neurotoxic destruction of brain serotonergic neurons resulted in a significant reduction in granule cell neurogenesis in the adult rat DG.
Most recently, we have completed an experiment that has the most direct relevance to the present theme. It is well-known that most drugs effective in the treatment of clinical depression act by means of increasing brain serotonergic neurotransmission. Therefore, we decided to examine the effects of systemic, chronic administration (3 weeks) of fluoxetine on DG mitogenesis in the adult rat. In support of our general hypothesis, we found an approximately 70% increase in the number of cells produced.57 The next studies in this series will determine whether chronic antidepressant treatment can block and/or reverse a stress-induced suppression in the rate of DG neurogenesis. Initial studies from Duman's laboratory confirm and extend our result.58 They report that not only fluoxetine, but antidepressants acting preferentially on norepinephrine, and chronic electroconvulsive shock also increase cell proliferation in the rat DG.
These studies demonstrate not only that serotonin can dramatically augment neurogenesis, but that it does so at least in part by an action at the 5-HT1A receptor. In this context, it is also interesting to note that the hippocampus, and especially the DG, is the site of an extremely dense concentration of these receptors.59
If this hypothesis is correct, then there should be evidence directly relating the 5-HT1A receptor to depression and/or its treatment. On the preclinical side, it is noteworthy that a number of studies have reported an inhibitory effect of glucocorticoids on the number of 5-HT1A binding sites (reviewed in Meijer et al).60 It sounds contrary to our hypothesis that 5-HT1A activation increases plasma glucocorticoid levels (reviewed in Sibug et al),61 however our data indicate that despite this, an increase in neurogenesis still occurs. It appears that the suppressive effects of glucocorticoids are more than surmounted by the enhancing effects of 5-HT1A stimulation on neurogenesis.
The major impediment to testing whether 5-HT1A agonist drugs would be effective therapeutic agents for depression is the failure, thus far, to develop a potent and specific 5-HT1A receptor agonist drug for human use. Nonetheless, there is evidence that 5-HT1A partial agonists have anxiolytic properties (anxiety is often considered an important antecedent to depression) and some antidepressant efficacy.62, 63 Unfortunately, there has been little direct examination of altered brain 5-HT1A function in depressed patients. One study reported a decreased number of 5-HT1 binding sites in the hippocampus of depressed suicide victims, but specific 5-HT1A binding in the hippocampus was not examined.64 More recently, Lopez et al65 reported that a group of depressed suicide victims had a decrease in the expression of 5-HT1A mRNA in their hippocampi.
Although we are proposing that alterations in hippocampal neurogenesis play a crucial role in the etiology and recovery from depression, we do not exclude other changes as being important. For example, besides suppressing neurogenesis, increased glucocorticoids may mediate additional direct neuronal effects in the cerebral cortex, hippocampus, and other subcortical areas such as the amygdala. Similarly, increases and decreases in serotonin neurotransmission, besides affecting hippocampal neurogenesis, may exert additional direct effects in the brain stem, subcortical sites, and in the cortex. All of these changes, acting in concert, give rise to the complex syndrome of depression.
While the present hypothesis focuses on the augmentation of DG neurogenesis by serotonin, there are additional means of increasing neurogenesis that may also have clinical relevance. For example, it is well-known that exercise, especially running, has an antidepressant action. As described above, we have recently found that wheel-running in mice for 4–10 days can induce a significant increase in DG mitogenesis and neurogenesis7 (Jacobs et al, unpublished results). Also, as noted above, increases in norepinephrine appear to exert a mitogenic effect in the DG.
A final feature of this hypothesis is that it provides a conceptually simple explanation for the ‘therapeutic lag’ (antipressant treatments, both drugs and ECT, typically require 3–6 weeks to become effective). We suggest that this is due to the time it takes for newly born DG neurons to extend their neurites and become fully functionally integrated into the existing brain circuitry.
Our theory does not supplant, nor does it contradict, a number of previous theories regarding the role of glucocorticoids and/or the hippocampus in depression. Rather, it complements and extends these previous ideas by pointing to a particular neural event, the waning and waxing of dentate gyrus neurogenesis, as a nexus in the development of, and recovery from, depression.
Blier and de Montigny66 hypothesize that the hippocampus plays a critical role in depression because of its involvement in anxiety (often a precursor to depression) and memory, and also because of its role in regulating the pituitary-adrenal axis. They suggest that antidepressant therapies act in the hippocampus by increasing neurotransmission at the serotonin 5-HT1A receptor and by decreasing it at the β adrenergic receptor. Watson and colleagues65 also emphasize the importance of the hippocampus in depression because of its role in regulating the pituitary-adrenal axis and because of its strategic site in the neural circuitry involved in affect and memory. Furthermore, they emphasize the importance of glucocorticoid (stress)-induced down regulation of the 5-HT1A receptors in the hippocampus of experimental animals. As mentioned above, this position is bolstered by their report that a group of depressed suicide victims had decreased expression of 5-HT1A mRNA in their hippocampi. Similar views, focussing on the pituitary-adrenal system and/or the hippocampus have also been expressed by Nemeroff67 and McEwen.68
Duman, Heninger, and Nestler69 have taken a somewhat similar position regarding glucocorticoids, the hippocampus and depression; however, they have also emphasized the therapeutic roles of serotonin and norepinephrine, and have moved the issue further into the molecular and cellular domain. They review studies indicating that long-term antidepressant treatments activate the cAMP pathway in specific brain regions and, further that this augmentation includes that of the transcription factor cAMP response element-binding protein (CREB). The activation of this system regulates specific target genes, including those for neurotrophic factors such as BDNF. Importantly, stress can decrease the expression of BDNF and lead to altered neuronal morphology or even neuronal death. Thus, their view is that depression may be especially dependent upon variation in neurotrophic factors, caused by stress and/or genetic factors, which can either impair or enhance the survival and function of particular neurons, especially those in the hippocampus.
Neurogenesis in the next millennium
Until recently neurogenesis was considered a process reserved for early development. Once adulthood was reached, it was thought, no new neurons could be added, only lost. We now know that in at least two regions of the adult mammalian brain neurogenesis continues through life. However, we do not know the mechanisms that control this process in the adult, or the functions subserved by these late-born neurons. Importantly, these primitive cells in the adult hippocampus can be isolated from the brain and expanded in culture where they can be studied. Considerable effort is underway to investigate the cellular and molecular mechanisms that control and regulate the proliferation, migration, and differentiation of these cells in vitro. In parallel, considerable effort is being exerted to define the factors and conditions that control the proliferation, migration and differentiation of these cells in situ. A more difficult, but no less important task is to understand the selective value to the organism for permitting this process to continue in the adult, and alternatively why neurogenesis is restricted to these two brain regions. Understanding the normal mechanisms for neurogenesis in the adult will lay the foundation for understanding the consequences of disruptions of this process. In addition, integrating the knowledge of the mechanism for neurogenesis into our present knowledge of the neurobiology of cognitive processes will expand our concept of, and tools to investigate, what we describe as ‘neural plasticity’. No longer are the structural correlates of neural plasticity limited to synaptic reorganization of changes in axon terminals and dendritic arborization, but now include the addition of new neurons to important circuits.
When considering the therapeutic value of this new information, two separate strategies are being weighed. One is where the primitive cells (stem cells) are harvested from the adult brain and expanded in culture, induced to differentiation towards a specific lineage and then transplanted in the corresponding brain region where they will take up functional residence to replace or augment endogenous cells. The second strategy is where the host endogenous cells are activated by either pharmacological or environmental stimulation and induced to proliferate, and migrate to a damaged or diseased brain locus where they take up residence in areas to replace or augment lost function. On a speculative note, uncommitted progenitors may exist throughout the adult brain and spinal cord, some dividing slowly, while others are quiescent for most of their life. These cells may represent the vestige of an earlier period when robust cellular replacement was common, but incompatible with survival. Nevertheless, the availability of this distributed replacement population may represent a virtually untapped reservoir of cells that can be induced to augment existing brain structures, with significant influence on behavior. While progress is being made on both of these fronts, much additional work is needed to make these repair strategies routine. In the meantime, a more complete understanding of adult neurogenesis and how it integrates into our understanding of adult brain function and neural plasticity will be important and will likely be accompanied by some surprises.
Note added in proof
A recent study (Gould E, Reeves AJ, Graziano MSA, Gross CG. Neurogenesis in the neocortex of adult primates. Science 1999; 286: 548–552) reports evidence of neurogenesis in the cortex of adult monkeys. This finding does not detract from the present hypothesis, but merely expands the brain areas in addition to the DG where the waning and waxing of neurogenesis may impact upon clinical depression.
This work was supported by Princeton University and a grant from the NIMH (BLJ). We would also like to thank S Forbes, L Moore, B Miller and L Kitabayashi for their excellent technical assistance. Our special thanks to ML Gage for critical reading of this manuscript. We are grateful for continued support of the Hollfelder Foundation, Robert J and Claire Pasarow Foundation, and a grant and contract from the National Institutes of Health (HV and FHG).