Pharmacology of Adult Neurogenesis: Compensatory and Regenerative Processes

Neurogenesis occurs in the adult brain of mammals and is modulated by a broad range of stimuli and conditions. Drugs used in the treatment of Alzheimer’ s disease (AD) and depression stimulate neurogenesis in the adult hippocampus. Neurogenesis is enhanced in the brain of patients with neurological diseases and disorders and in the brain of animal models of neurological diseases and disorders. Despite controversies surrounding such studies and the need to confirm these data, adult neurogenesis and neural stem cells (NSCs) may be the target of drugs used in the treatment of neurological diseases and disorders, particularly AD and depression, and adult neurogenesis and the hippocampus may contribute to the pathol­ogy of neurological diseases and disorders.

Enhanced neurogenesis in neurological diseases and disorders may represent regenerative attempts by the nervous system. Drug treatments may contribute to compensatory mechanisms in the adult hippocampus. It points to a broader involvement of the adult NSCs and the hippocampus in neurological diseases and drug therapy. Hence, adult NSCs represent not only a promising model for cel­lular therapy, but may also contribute to the physiopathology of the ner­vous system and its pharmacology, particularly for AD and depression, the understanding of which will lead to a better understanding of the ner­vous system, and the development of novel and more effective treatments and cures for neurological diseases and disorders.

In the mammalian brain, neurogenesis occurs throughout adulthood pri­marily in two regions, the dentate gyrus (DG) of the hippocampus and the anterior part of the subventricular zone (SVZ) in various species, includ­ing humans (Eriksson et al. 1998; Curtis et al. 2007; Taupin 2008, “Adult neural”). In the DG, newly generated neuronal cells in the subgranular zone (SGZ) migrate to the granule cell layer, where they differentiate into granule-like cells and extend axonal projections the to CA3 region (Cam­eron et al. 1993; Toni et al. 2007; Taupin 2009). In the SVZ, newly generated neuronal cells migrate to the olfactory bulb, through the rostro-migratory stream, where they differentiate into interneurons (Lois and Alvarez-Buylla 1994; Belluzzi et al. 2003; Curtis et al. 2007). About 0.1% of the granule cell population or 9,000 new neuronal cells are generated per day in the DG of young adult rodents and about 0.004% of the granule cell population is generated per day in the DG of adult macaque monkeys (Kornack and Rakic 1999; Cameron and McKay 2001) .

Though newly generated neu­ronal cells in the adult brain undergo programmed cell death rather than achieving maturity, the ones that survive survive for an extended period of time, at least two years in humans (Eriksson et al. 1998; Cameron and McKay 2001). It is postulated that newly generated neuronal cells in the adult brain originate from NSCs. NSCs are the self-renewing multipotent cells that generate the main phenotypes of the nervous system. Because of their potential to generate the main phenotypes of the nervous system, NSCs represent a promising model for cellular therapy for the treatment of a broad range of neurological diseases and injuries, particularly neurode- generative diseases, cerebral strokes and spinal cord injuries. The stimula­tion of endogenous neural progenitor or stem cells and the transplantation of adult-derived neural progenitor and stem cells are proposed to restore and repair the degenerated or injured nerve pathways.

Neurogenesis in the adult hippocampus and SVZ is modulated by a broad range of stimuli and conditions, including environmental enrich­ment, physiological processes, pathological conditions, trophic factors/ cytokines, and drugs (Taupin 2007). Neurogenesis in the adult hippocam­pus is modulated by drugs used in the treatments of AD and depression, and is modulated in the brain of patients with neurological diseases and disorders and in the brain of animal models of neurological diseases and disorders, like AD, epilepsy, and Huntington’s disease (HD) (Parent et al. 1997; Malberg et al. 2000; Curtis et al. 2003; Jin, Galvan et al. 2004; Jin, Peel et al. 2004; Jin et al. 2006). Do adult neurogenesis and newly generated neuronal cells of the adult brain contribute to the pharmacology of drugs used in the treatments of neurological diseases and disorders? Do adult neurogenesis and the hippocampus contribute to the pathology of neuro­logical diseases and disorders? In the following sections, we will review and discuss the potential involvement of adult neurogenesis and newly generated neuronal cells of the adult brain in the pharmacology of AD and depression, and in the pathology of neurological diseases and disorders.

Alzheimer’s Disease and Drug Therapy

Alzheimer’ s disease is a fatal neurodegenerative disease for which there is no cure. It is the most common dementia among elderly, affect­ing more than 26 million patients worldwide (Ferri et al. 2006). It starts with mild memory problems and ends with severe brain damages. AD is associated with loss of nerve cells, particularly in areas of the brain that are vital to memory and other mental abilities, like the hippocampus. The disease is characterized in the brain by amyloid or senile plaque deposits and neurofibrillary tangles (Caselli et al. 2006' . There are two forms of the disease: the late-onset form (LOAD), diagnosed after age 65, and the early-onset form (EOAD), diagnosed at younger age. Most of the cases of LOAD are sporadic forms of the disease, whereas most cases of EOAD are inherited or familial forms of AD (FAD). Risks factors for LOAD include genetic, acquired, and environmental risk factors, like the presence of certain alleles in the genetic makeup of the individual (e.g., the apolipo- protein E varepsilon 4 allele), hypertension, diabetes, and oxidative stress (Raber, Huang, and Ashford 2004). Genetic mutations in the beta-amyloid precursor protein gene (APP), the presenilin-1 gene (PSEN1) and the pre- senilin-2 gene (PSEN2) have been identified as causative factors for FAD (Schellenberg 1995; St. George-Hyslop and Petit 2005). LOAD represents most cases of AD, with over 93% of all cases of the disease.

Actual treatments for AD consist in drug and occupational therapies (Scarpini, Scheltens, and Feldman 2003). Three types of drugs are currently used in the treatment of AD: blockers of the formation of amyloid plaques like alzhemed (Aisen 2005); inhibitors of acetylcholine esterase like galan- tamine, rivastigmine and tacrine (Wilkinson et al. 2004); and N-methyl-D- aspartate glutamate receptor antagonists like memantine (Creeley et al. 2006)     . Inhibitors of acetylcholine esterase improve cognitive functions by enhancing cholinergic neurotransmission that are important for learning and memory and that are affected in brain regions of patients with AD.

N-methyl-D-aspartate glutamate receptor antagonists confer protection against excitotoxic neurodegeneration. These drugs produce improve­ments in cognitive and behavioral symptoms of AD. Other treatments that are considered and are being developed involve drugs for lowering cholesterol levels, anti-inflammatory drugs and protein beta-amyloid vac­cination ( Estrada and Soto 2007 ; Solomon 2007).

Adult Neurogenesis and Alzheimer’s Disease Drugs

The drugs used in the treatment of AD have been assessed for their effects on adult neurogenesis in rodents. Galantamine and memantine increase neurogenesis in the DG and SVZ of adult rodents by 26-45%, as revealed by bromodeoxyuridine (BrdU) labeling (Jin et al. 2006) (see Table 1). This suggests that adult neurogenesis may contribute to the activities of these drugs in the treatment of AD.

Depression and Antidepressants

Depression is a major public health issue. It affects an estimated 19 million Americans. Twenty-five percent of adults will have a major depressive epi­sode sometime in their life (Kessler et al. 1994). It is proposed that an imbal­ance in the 5-hydroxytryptamine (serotonin or 5-HT) and noradrenaline pathways underlies the pathogenesis of depressive disorders (Hindmarch 2001; Owens 2004). Stress and neuroinflammation are causal factors precipi­tating episodes of depression in humans (Minghetti 2005; Miura et al. 2008).

Actual treatments for depression consist in drug therapy, and psycho­logical support and therapy. Five types of drugs are used in the treatment of depression: selective serotonin reuptake inhibitors like fluoxetine, mono­amine oxidase inhibitors like tranylcypromine, selective norepinephrine reuptake inhibitors like reboxetine, tricyclic antidepressants like imipramine and desipramine, and phosphodiesterase-IV inhibitors, like rolipram (Wong and Licinio 2001; Brunello et al. 2002). The efficiency and therapeutic ben­efits of some of the antidepressants currently prescribed for the treatment of depression have been questioned in recent publications (Kirsch et al. 2008), mandating the development of new drugs for treating depression.

Adult Neurogenesis and Antidepressants

The activity of antidepressants has been assessed for their effects on adult neurogenesis in rodents and nonhuman primates (see Table 1). The chronic administration of fluoxetine and of agomelatine, the melatonergic agonist and serotoninergic antagonist, increases neurogenesis in the DG, but not SVZ of adult rats and nonhuman primates (Malberg et al. 2000' Banasr et al. 2006; Perera et al. 2007). The X-irradiation of the hippocampal region inhibits neurogenesis in the DG and prevents the behavioral effect of fluoxetine in adult mice in vivo (Santarelli et al. 2003). In this report, the activity of fluoxetine was reported to be mediated by 5-HT1A receptor in 129SvEvTac mice. In other strains of mice, BALB/cJ mice, fluoxetine activ­ity was reported not to be mediated by 5-HT1A receptor and produces its antidepressant activity independently of neurogenesis (Holick et al. 2008). Autopsy studies report that neurogenesis is not altered in the hippocam­pus of adult patients who were on antidepressant medication (Reif et al. 2006)       . The anxiolytic/antidepressant N-[3-(1-{[4-(3,4-difluorophenoxy)- phenyl]methyl}(4-piperidyl))-4-methylphenyl]-2-methylpropanamide (SNAP 94847) stimulates the proliferation of progenitor cells in the DG, but its activity is unaltered in mice in which neurogenesis was suppressed by X-irradiation (David et al. 2007).

Galantamine and memantine increase neurogen­esis in the DG of adult rodents by 26-45%

Chronic administration of fluoxetine increases neu­rogenesis in the DG of adult rodents and nonhu­man primatesa

Neurogenesis is enhanced in the hippocampus of AD patients’5 Neurogenesis is decreased in the DG of adult mice deficient for APP and/or PSEN1 Neurogenesis is decreased in the DG of adult PDAPP transgenic mice Neurogenesis is increased in the DG of adult trans­genic mice that express the Swedish and Indiana APP mutations

Post-mortem studies do not reveal any increase in neurogenesis in the hippocampus of patients with depressionc

Neurogenesis is increased in the DG of animal mod­els of epilepsy, like after pilocarpine treatment

Neurogenesis is decreased in the DG of adult R6/1 transgenic mice Neurogenesis is increased in the SVZ of adult rodents after quinolinic acid striatal lesioning

Notes: Study and quantification of adult neurogenesis was performed primarily by immuno- histology for markers of the cell cycle and for the thymidine analog BrdU. a In other strains of mice, fluoxetine and other antidepressants were reported to produce their activities independently of adult neurogenesis.

b Autopsies in this study were performed most likely on patients with the sporadic form of AD.

c Autopsies in this study were performed on patients that were on antidepressant medica­tion. Adult neurogenesis may mediate the activities of drugs used in the treatment of AD and depression. The modulation of neurogenesis in the adult hippocampus would repre­sent a phenomenon of plasticity or compensatory mechanisms of recovery. Neurogenesis is enhanced in the brain of patients with and of animal models of neurological diseases and disorders, particularly AD, epilepsy and HD. Enhanced neurogenesis in the DG of the brain with neurological diseases and disorders, particularly neurodegenerative diseases, may contribute to regenerative attempts, to compensate for the neuronal losses. Due to the limi­tations and pitfalls over the methodologies and paradigms used to study and quantify neu­rogenesis, these data remain to be validated and confirmed.

This shows that adult neurogenesis may mediate the activities of anti­depressants, particularly selective serotonin reuptake inhibitors like flu­oxetine, but that fluoxetine and other antidepressants may also produce their activity via distinct mechanisms, some independently of adult neu­rogenesis. The mechanisms underlying the activity of antidepressants on adult neurogenesis remain to be determined. It may be mediated by glu­cocorticoids, stress-related hormones, interleukin-6, a cytokine involved in neuroinflammation, and brain-derived neurotrophic factor, a trophic factor that has antidepressant effects (Siuciak et al. 1997; Cameron, Tana- pat, and Gould 1998; Vallieres et al. 2002; Scharfman et al. 2005).

Alzheimer’s Disease: Increased Neurogenesis and Possible Disease Mechanisms Related to Neurogenesis

Amyloid plaques and neurofibrillary tangles are the hallmarks of AD. Amyloid plaques are extracellular deposits of proteins surrounded by degenerating nerve cells, in the brain of patients with AD (Anderson et al. 2004) . They are composed of amyloid fibrils. Amyloid fibrils are aggregates of protein beta-amyloid. Protein beta-amyloid is a 40 amino acid beta-peptide. It is synthesized and secreted by nerve cells, by post- transcriptional maturation of APP (Kang et al. 1987). APP is processed by alpha-, beta- and gamma-secretase enzymes.

Protein beta-amyloid is an amyloidogenic protein. These proteins are soluble in their physiological state. Under pathological conditions, they form insoluble extracellular aggregates or deposits of amyloid fibrils (Ser- pell, Sunde, and Blake 1997). In physiological conditions, APP is cleaved by the alpha- and gamma-secretase enzymes into a 40 amino acid beta- peptide. Certain pathological conditions, like the presence or expression of amyloid-promoting factors or certain gene mutations, including in APP, cause excessive cleavage of APP by the beta- and gamma-secretase enzymes, resulting in an increase production of a 42 amino acid beta- amyloid peptide. This latter form of protein beta-amyloid aggregates into insoluble amyloid deposits, particularly in the brain, forming aggregates and deposits of amyloid fibrils.

Amyloid plaques are thought to be the first histological change that occurs in the brain of patients with AD (St. George-Hyslop 2000). The den­sity of amyloid plaques increases as the disease advances. They are distrib­uted throughout the brain of those patients, particularly in the region of degeneration, like the entorhinal cortex, hippocampus, temporal, frontal, and inferior parietal lobes. The role and contribution of amyloid plaques in the pathology of AD remain unclear and the source of controversies. On the one hand, it is proposed that deposits of protein beta-amyloid may be a causative factor of AD. According to this hypothesis, referred as amyloid hypothesis, as the amyloid deposits in the brain, brain cells start dying, and the signs and symptoms of the disease begin (Hardy and Sel- koe 2002; Meyer-Luehmann et al. 2006). On the other hand, the correlation between the density of amyloid plaques and the severity of the dementia is not clearly established (Terry 1996). The deposit of protein beta-amyloid would be a consequence rather than a cause of AD.

Neurofibrillary tangles are deposits of proteins present inside neuronal cells in the brain of patients with AD. They are composed of hyperphos- phorylated tau proteins (Fukutani et al. 1995). Tau protein is a microtubule- associated phosphoprotein. It is involved in the formation of microtubules (Kim, Jensen, and Rebhun 1986). The hyperphosphorylation of tau pro­teins result in their aggregation and in the breakdown of microtubules (Iqbal et al. 1998). This leads to the formation of neurofibrillary tangles and cell death (Alonso et al. 2001). As the disease advances, the regions of the brain affected expand, leading to severe incapacity and death (Brun and Gustafson 1976).

There are two forms of the diseases, sporadic and inherited. Most cases of LOAD are sporadic forms of the disease and are diagnosed after the age of 65. EOAD is diagnosed at younger age than 65 and most cases of EOAD are inherited forms of AD. LOAD is believed to be caused by genetic, acquired, and environmental factors, among them the presence of certain alleles in the genetic makeup of the individuals, hypertension and diabe­tes, and neuroinflammation and oxidative stress (Cankurtaran et al. 2008). The presence of the apolipoprotein E varepsilon 4 allele (ApoE4) is the best established genetic risk factor for LOAD.

ApoE is a plasma protein; it participates in the transport of cholesterol and other lipids in the blood (Mahley 1988). There are four major isoforms of the gene coding for ApoE encoded by different alleles in humans, ApoE, ApoE2, ApoE3 and ApoE4. ApoE accounts for the vast majority of causes and risks to develop LOAD: up to 50% of people who have AD have at least one ApoE4 allele. Neuronal sortilin-related receptor (SORL1) belongs to a family of proteins termed retromer (Raber, Huang, and Ashford 2004). Retromers are involved in intracellular trafficking. Reduced expression of the gene coding for SORL1 (SORL1) is associated with an increase in density of amyloid plaques in the brain and increased risk for LOAD. The variants of SORL1 may pro­mote AD by suppressing the activity of the gene. This may affect the pro­cessing of APP and increase its production (Rogaeva et al. 2007). Other genes have been linked with the occurence of LOAD, among them vari­ants for the genes coding for alpha2-macroglobulin, monoamine oxidase A, myeloperoxidase and cystatin C (CST3) (Finckh et al. 2000). These risk factors increase the probability of developing the disease.

So far, three genes have been identified as carrying genetic mutations underlying the development of EOAD. These genes are also known as FAD genes. These genes are the APP gene, the presenilin-1 gene (PSEN1) and the presenilin-2 gene (PSEN2) (Schellenberg 1995). APP is a 695-770 amino acid protein coding for the protein beta-amyloid. The PSEN proteins are components of the gamma-secretase complex. These enzymes play a role in the maturation of APP into the 42 protein beta-amyloid (Nishimura, Yu, and St. George-Hyslop 1999). Mutations in PSEN1 and PSEN2 lead to excessive cleavage by gamma-secretase enzyme, resulting in increased production and aggregation of protein beta-amyloid (Newman, Mus- grave, and Lardelli 2007). Mutations in these genes almost always result in the individual developing the disease (Hardy 2001).

Aneuploidy is an abnormal number of chromosomes in the cells of the body. It is a common cause of genetic disorders. Several studies report that cells of patients with AD elicit aneuploidy, particularly for chromosome 21, 13, and 18. Lymphocytes of patients with LOAD present an elevation in aneuploidy for chromosomes 13 and 21 (Migliore et al. 1999). Prepara­tions of lymphocytes of patients with sporadic and inherited forms of AD elicit a two-fold increase in the incidence of aneuploidy for chromosomes 18 and 21 (Geller and Potter 1999). In regions of degeneration 4-10% of neurons, like the hippocampus, are aneuploids and express proteins of
the cell cycle in the brain of patients with AD (Busser, Geldmacher, and Herrup 1998; Yang, Geldmacher, and Herrup 2001). The adult brain con­tains a substantial number of cells that are aneuploids; estimated at 5-7% of the cells in the brain of adult mice (Rehen et al. 2005) . The genetic imbalance in aneuploid cells signifies that they are fated to die (Herrup et al. 2004). The relatively high percentage of aneuploid cells in regions of degeneration in AD brains suggests that they undergo a slow death process. These cells may live in this state for months, possibly up to 1 year ( Herrup and Arendt 2002 ; Yang, Mufson, and Herrup 2003 ). This sup­ports their involvement in the slow and progressive neurodegenerative process of AD. Cyclin B, the marker of the phase G2 of the cell cycle, is also expressed in neurons in regions of degeneration, particularly the hip­pocampus, in patients with AD (Vincent, Rosado, and Davies 1996).

In the adult brain, most nerve cells are post-mitotic cells. The charac­terization of aneuploidy and cyclin B in nerve cells in the region of degen­eration reveal that cell cycle re-entry and DNA duplication, without cell division, precedes neuronal death in the brain of patients with AD. The deregulation and/or re-expression of proteins of the cell cycle in nerve cells triggering cycle re-entry, with blockage in phase G2, and aneuploidy would underlie the neurodegenerative process and pathogenesis of AD.

The expression of markers of immature neuronal cells, like doublecor- tin and polysialylated nerve cell adhesion molecule, is enhanced in the hippocampus, particularly the DG, in the brain of AD patients, most likely with LOAD (Jin, Peel, et al. 2004). In animal models, neurogenesis is decreased in the DG of adult mice deficient for PSEN1 and/or APP, in the DG of adult transgenic mice over expressing variants of APP or PSEN1, and in the DG of adult PDAPP transgenic mice, a mouse model of AD with age-dependent accumulation of protein beta-amyloid (Wen et al. 2002; Donovan et al. 2006; Verret et al. 2007; Zhang et al. 2007; Rodriguez et al. 2008). It is increased in the DG of adult transgenic mice that express the Swedish and Indiana APP mutations (Jin, Galvan, et al. 2004). Mice deficient for or over expressing variants of APP or PSEN1, and transgenic mice that express the Swedish and Indiana APP mutations, a mutant form of human APP, are transgenic mice that express variants of FAD genes. Transgenic mice deficient for APP and PSEN1 provide information on the activities and functions of the proteins involved in EOAD. They are not representative of complex diseases, like LOAD. They do not repre­sent the disease. The aggregation of protein beta-amyloid affects adult neurogenesis (Heo et al. 2007). It may have adverse effects on neurogen­esis during development in transgenic mice for APP, affecting the adult phenotype. In all, the discrepancies of the data observed on adult neu­rogenesis in autopsies and animal models of AD may originate from the validity of the animal models used in those studies, as representative of AD and to study adult phenotypes (German and Eisch 2004).

The discrepancies of the data observed on adult neurogenesis may also originate from the validity of the protocols used as a paradigm to study adult neurogenesis, like the immunohistochemistry for markers of the cell cycle and for the thymidine analog bromodeoxyuridine (BrdU). Most of the studies conducted in autopsies and animal models of neurological dis­eases and disorders use either immunoshistochemistry for markers of the cell cycle or the BrdU labeling paradigm, to study and quantify adult neu­rogenesis in situ. Proteins of the cell cycle, like cyclin B—the marker of the phase G2—are expressed in neurons in regions where neurodegeneration occurs. Some at-risk neurons in regions of degeneration are aneuploids in the brain of patients with AD (Busser, Geldmacher, and Herrup 1998; Yang, Geldmacher, and Herrup 2001). Cell cycle re-entry and DNA duplication, without cell division, precedes neuronal death in degenerating regions of the CNS. This suggests that when using immunohistochemistry for pro­teins of the cell cycle, to study adult neurogenesis, this paradigm does not allow discrimination between cells undergoing DNA duplication, without cell division, as part of their pathological fate and newly generated neu­ronal cells (Taupin 2007). BrdU is used for birth dating and monitoring cell proliferation (Miller and Nowakowski 1988) .

There are pitfalls and limitations over the use of thymidine analogs, and particularly BrdU, for studying neurogenesis (Nowakowski and Hayes 2000; Gould and Gross 2002). BrdU is a thymidine analog. It is not a marker of cell proliferation; it is a marker for DNA synthesis. Studying and quantifying neurogen­esis with BrdU require distinguishing cell proliferation and neurogenesis from other events involving DNA synthesis, like DNA repair, abortive cell cycle re-entry and cell cycle re-entry and gene duplication, without cell division, leading to aneuploidy (Taupin 2007). In addition, BrdU has a number of side effects. It is a toxic and mutagenic substance. It alters DNA stability and lengthens the cell cycle. BrdU has mitogenic, transcrip­tional, and translational effects on cells that incorporate it. It triggers cell death and the formation of teratomes. Hence, data involving the use of immunohistochemistry for proteins of the cell cycle and BrdU labeling, as paradigms for studying adult neurogenesis in neurological diseases and disorders, and particularly in AD, must be carefully assessed, analyzed, and discussed.

In all, AD is a neurodegenerative disease that affect mostly individu­als over 65 years of age. There are two forms of the disease, sporadic and inherited. It is characterized by widespread neurodegeneration, amyloid deposits and neurofibrillary tangles, aneuploidy and enhanced neurogen­esis, though this latter observation remains to be fully established. It is proposed that enhanced neurogenesis may be a result, rather than a cause, of the illness (Taupin 2008, “Adult neurogenesis pharmacology”) Taupin

2008,   “Adult neurogenesis and drug therapy”). Enhanced neurogenesis in the DG of the brain with neurological diseases and disorders, particularly neurodegenerative diseases, may contribute to a regenerative attempt, to compensate for the neuronal loss.