David A. Greenberg, MD, PhD, Professor, Buck Institute for Age Research, 8001 Redwood Boulevard, Novato, CA 94945, USA. Tel.: 415 209 2087; fax: 415 209 2230; e-mail: email@example.com
Neurogenesis, which may contribute to the ability of the adult brain to function normally and adapt to disease, nevertheless declines with advancing age. Adult neurogenesis can be enhanced by administration of growth factors, but whether the aged brain remains responsive to these factors is unknown. We compared the effects of intracerebroventricular fibroblast growth factor (FGF)-2 and heparin-binding epidermal growth factor-like growth factor (HB-EGF) on neurogenesis in the hippocampal dentate subgranular zone (SGZ) and the subventricular zone (SVZ) of young adult (3-month) and aged (20-month) mice. Neurogenesis, measured by labelling with bromodeoxyuridine (BrdU) and by expression of doublecortin, was reduced by ∼90% in SGZ and by ∼50% in SVZ of aged mice. HB-EGF increased BrdU labelling in SGZ at 3 months by ∼60% and at 20 months by ∼450%, which increased the number of BrdU-labelled cells in SGZ of aged mice to ∼25% of that in young adults. FGF-2 also stimulated BrdU labelling in SGZ, by ∼25% at 3 months and by ∼250% at 20 months, increasing the number of newborn neurones in older mice to ∼20% of that in younger mice. In SVZ, HB-EGF and FGF-2 increased BrdU incorporation by ∼140% at 3 months and ∼170% at 20 months, so the number of BrdU-labelled cells was comparable in untreated 3-month-old and growth factor-treated 20-month-old mice. These results demonstrate that the aged brain retains the capacity to respond to exogenous growth factors with increased neurogenesis, which may have implications for the therapeutic potential of neurogenesis enhancement in age-associated neurological disorders.
Neurogenesis continues throughout life in the subgranular zone (SGZ) of the hippocampal dentate gyrus (DG) and in the subventricular zone (SVZ). Although neurogenesis persists in the aged brain, its rate declines with age in rats (Kuhn et al., 1996), mice (Kempermann et al., 1998), monkeys (Gould et al., 1999) and humans (Cameron & McKay, 1999; Kukekov et al., 1999). Decreased hippocampal neurogenesis may be involved in age-related cognitive deficits because of its proposed role in learning and memory function (Shors et al., 2001). Hippocampal neurogenesis is increased in aged rats after adrenalectomy (Cameron & McKay, 1999) and in aged mice living in an enriched environment (Kempermann et al., 2002). In the latter case, enhanced neurogenesis is accompanied by improved learning, exploratory behaviour and locomotor activity. Therefore, restoring hippocampal neurogenesis may be a strategy for reversing age-related cerebral dysfunction.
Acute and chronic neurodegenerative diseases in humans occur with increased frequency during aging and contribute to age-related impairment of brain function, and some such disorders may be associated with alterations in neurogenesis. Transgenic mice expressing a presenilin-1 mutation found in some cases of familial Alzheimer's disease have impaired hippocampal neurogenesis compared to mice that overexpress wild-type presenilin-1 (Wen et al., 2002). Neuronal progenitor cells in the substantia nigra may have the capacity to replenish neurones lost in Parkinson's disease (Lie et al., 2002). Neurogenesis is increased in animal models of stroke (Liu et al., 1998; Jin et al., 2001), and the new neurones that result can migrate to areas of ischaemic brain injury, where they may replace cells damaged by ischaemia (Arvidsson et al., 2002; Nakatomi et al., 2002). Thus, reduction of neurogenesis in aging may not only contribute to physiological decline in brain function, but may also compromise the brain's response to disease.
Growth factors stimulate proliferation of neuronal precursors from adult brain in vitro and in vivo. For example, fibroblast growth factor (FGF)-2 is expressed in adult rodent SVZ and SGZ (Yazaki et al., 1994; Goldman, 1998) and chronic intracerebroventricular infusion of FGF-2 in adult rats increases neurogenesis in SVZ and neuronal migration to OB (Kuhn et al., 1997). Insulin-like growth factor-I (IGF-1) also stimulates neurogenesis (O’Kusky et al., 2000) and, while IGF-1 levels decline in aged rats, their restoration increases neurogenesis in aged rat brain (Lichtenwalner et al., 2001). These observations suggest that through their ability to promote adult neurogenesis, growth factors might have a role in the treatment of aged-related neurodegenerative disorders.
We found recently that FGF-2 and heparin-binding epidermal growth factor-like growth factor (HB-EGF) increased neurogenesis in DG and SVZ of the adult rat brain (Jin et al., 2002b). In this study, we investigated whether FGF-2 and HB-EGF have the capacity to restore neurogenesis in the aged mouse brain. We report that neurogenesis decreased in the aged mouse brain, and that both FGF-2 and HB-EGF enhanced neurogenesis towards levels seen in young adults. These observations could have therapeutic implications for neurological disorders associated with aging.
Neurogenesis is decreased in DG and SVZ of aged mice
To confirm that neurogenesis decreased in DG and SVZ of our aged mice, young adult (3 months old) and aged (20 months old) mice were given BrdU for three consecutive days and killed 1 week later, and the number of BrdU-labelled cells in the DG and SVZ between brain was compared by immunohistochemistry. BrdU-labelled cells were detected in DG (Fig. 1A) and SVZ (Fig. 2A) in both young adult and aged mouse brain, but the number of BrdU-labelled cells decreased with age in both regions. To quantify these changes, BrdU-labelled nuclei were counted and found to decrease by ∼90% in DG (Fig. 1B) and by ∼50% in SVZ (Fig. 2B) of aged compared to young adult mouse brain. Immunocytochemistry with an antibody against DCX, a microtubule-associated protein found in the soma and processes of migrating immature neurones (Francis et al., 1999; Gleeson et al., 1999; Magavi et al., 2000), also showed a decrease in immunoreactive cells in both DG (Fig. 1C) and SVZ (Fig. 2C) of aged mice, which is further consistent with an age-related decline in neurogenesis.
To determine whether the BrdU-labelled cells that we observed expressed phenotypic neuronal features, we double-labelled brain sections with antibodies against BrdU and against the cell-type marker proteins NeuroD, an immature neuronal nucleoprotein (Lee et al., 1995); nestin, a protein expressed by neuroepithelial progenitor cells (Kaneko et al., 2000); glial fibrillary acid protein (GFAP), an astrocytic marker (Debus et al., 1983). BrdU labelling colocalized with NeuroD and nestin but not with GFAP expression in DG and SVZ, indicating that BrdU labelled primarily cells of neuronal lineage (Fig. 3).
HB-EGF and FGF-2 stimulate neurogenesis in DG and SVZ of aged mice
Growth factors are the best characterized exogenous influences on neurogenesis (Cameron et al., 1998). Growth factors such as FGF-2 and HB-EGF stimulate neurogenesis in young adult rodent brain (Kuhn et al., 1997; Yoshimura et al., 2001; Jin et al., 2002b), but whether they retain this capacity in the aged brain is unknown. Because the acute and chronic neurodegenerative disorders in which therapeutic neurogenesis might have application occur primarily in the elderly, this is an important uncertainty to resolve. To investigate whether FGF-2 and HB-EGF can promote neurogenesis in aged brain, FGF-2 and HB-EGF were infused into the right lateral ventricle and BrdU was injected intraperitoneally for 3 days in young adult (3-month) and aged (20-month) mice, and animals were killed 1 week later for BrdU immunohistochemistry.
In the absence of growth factor treatment, the number of BrdU-labelled cells in DG of 20-month-old mice was only ∼10% of that in 3-month-old mice (Figs 4 and 5). HB-EGF stimulated BrdU labelling in DG at 3 months by ∼60% and at 20 months by ∼450%, which increased the number of BrdU-labelled cells in DG of aged mice to ∼25% of that in young adults. FGF-2 also stimulated BrdU labelling in DG, by ∼25% at 3 months and by ∼250% at 20 months, increasing cell number in the older mice to ∼20% of that in younger mice. There were no significant differences between BrdU-positive cell counts in the DG ipsilateral compared to that contralateral to growth factor infusion. Thus, growth factor treatment had proportionally greater effects on BrdU incorporation in DG of aged than of young adult mice, although the number of BrdU-labelled cells in DG of growth factor-treated 20-month mice was still lower than that in untreated 3-month mice.
In SVZ, the age-related reduction in BrdU labelling was less pronounced than in DG, resulting in BrdU-positive cells counts at 20 months that were ∼40% of those at 3 months (Figs 6 and 7). In SVZ, as in DG, HB-EGF and FGF-2 stimulated BrdU incorporation to a greater extent at 20 than at 3 months, although the discrepancy was small (∼170% increases at 20 months and ∼140% increases at 3 months for both growth factors). However, in SVZ as opposed to DG, growth factor treatment resulted in levels of BrdU labelling at 20 months that were comparable to levels in untreated 3-month mice on the side contralateral to the infusion, and somewhat higher on the ipsilateral side. Therefore, the aged SVZ, like the aged DG, retained responsiveness to exogenous growth factors.
To identify the cell lineage that proliferates in response to FGF-2 and HB-EGF in the mouse brains we studied, we stained brain sections with antibodies against BrdU and against cell-type markers. In DG and SVZ of both young adult and aged mouse brain, most BrdU-positive cells expressed the immature neuronal marker DCX (Fig. 8). Occasional cells coexpressed BrdU and the neuronal nuclear antigen NeuN (which is expressed at a later stage of neuronal development than DCX), the astroglial marker GFAP, the microglial marker OX-42, or the endothelial cell marker von Willebrand factor (not shown).
The major finding we report is that despite a reduction in basal neurogenesis in DG and SVZ, the aged mouse brain retains the capacity to respond to the neurogenesis-stimulating effects of growth factors, specifically FGF-2 and HB-EGF. The ages chosen for study – 3 months (young adult) and 20 months (aged) – are typical for studies of brain aging in mice (Gower & Lamberty, 1993). Neurogenesis was measured by counting BrdU-labelled cells in DG and SVZ, most of which expressed the immature neuronal marker protein, DCX. We have shown in previous studies that when BrdU is administered in vivo as described in this report, the cells it labels in DG and SVZ are primarily ones that also (a) express the cell-proliferation marker proliferating cell nuclear antigen, (b) show no evidence of injury by conventional histological staining or labelling of DNA strand breaks with the Klenow fragment of DNA polymerase I, and (c) express immature (DCX, Hu, NeuroD) and later more mature (NeuN) neuronal marker proteins (Jin et al., 2001, 2002a, 2002b). In the present study, the loss of basal BrdU labelling in DG and SVZ of aged brains was associated with a decrease in the number of DCX-immunopositive cells in these regions, consistent with reduced production of nascent neurones.
Although basal BrdU labelling in aged mice was only ∼10% of that in young adult mice in DG, and ∼50% in SVZ, HB-EGF and FGF-2 produced proportionally greater increments in labelling in the older mice. This was most striking in DG, where the growth factor-induced increment was ∼8–10-fold higher in aged mice. Because of the enhanced response of the aged mice to growth factors, BrdU labelling in growth factor-treated aged mice reached ∼20% (in DG) and ∼110% (in SVZ) of levels in untreated young adults. The magnitude of stimulation of BrdU incorporation by FGF and by HB-EGF was comparable, which resembles findings from in vitro studies (Jin et al., 2002b). BrdU labelling was similar ipsilateral and contralateral to the site of intracerebroventricular injection of growth factors, except in SVZ of 20-month rats, which showed greater ipsilateral effects with both growth factors. We observed similar laterality in a previous study of HB-EGF-induced (Jin et al., 2002d), but not stem cell factor-induced (Jin et al., 2002a), neurogenesis in young adult rats. The explanation is unclear, but could relate to the ipsilateral injury caused by intracerebroventricular injection, because there was a trend towards increased BrdU labelling on the side ipsilateral to artificial cerebrospinal fluid (aCSF) injection in control mice as well. Alternatively, or in addition, the transiently higher levels of growth factor in the ipsilateral lateral ventricle immediately after injection might lead to enhanced stimulation of neurogenesis in the nearby SVZ.
Kuhn and colleagues (Kuhn et al., 1996) reported decreased neurogenesis, detected by BrdU labelling, in DG but not SVZ of aged (21-month) compared to young adult (6-month) rats. This was accompanied by a reduction in the number of cells in DG that expressed the immature neuronal marker, embryonic (polysialylated) nerve cell adhesion molecule (ENCAM), and the magnitude of reduction in BrdU labelling was equivalent to what we observed. However, our results differ in that we also found reduced BrdU labelling in SVZ, although its magnitude was less than in DG. Aside from the possible effect of species differences, there is no obvious explanation for this discrepancy. To investigate the basis for the age-related decrease in neurogenesis in DG, Kuhn and colleagues measured BrdU labelling within hours after BrdU administration, and before possible age-related differences in cell death would be manifested. Under these conditions, the difference between BrdU labelling in DG of young adult and aged rats persisted, consistent with decreased neuroproliferative activity in the aged DG. However, no conclusion could be drawn as to whether aging reduced neurogenesis in this model via the loss of extrinsic signals that stimulate neuronal precursors to proliferate, or through reduced intrinsic responsiveness of precursor cells to normal signals.
Subsequent studies have shown that, under appropriate conditions, the ability of neuronal precursors to proliferate can be restored in DG of aged rodents. Because corticosteroids inhibit neurogenesis, Cameron and McKay studied BrdU labelling in DG of 5-month and 26-month rats after sham surgery or adrenalectomy (Cameron & McKay, 1999). They found that although the number of BrdU-positive cells in sham-operated aged rats was reduced to about ∼30% of young adult levels, adrenalectomy stimulated BrdU incorporation to achieve similar levels of BrdU labelling irrespective of age. In another study, Kempermann and colleagues found that long-term environmental enrichment increased the number of BrdU- and neuronal nuclear antigen (NeuN)-immunopositive cells in DG of 20-month mice approximately five-fold, indicating that neuronal precursors in DG of aged mice also remain responsive to this well-documented stimulus to neurogenesis (Kempermann et al., 2002). In addition to responding to physiological stimuli such as hormones and experience, neuronal progenitors in aged brain may also be capable of responding to pathological processes, because focal cerebral ischaemia, which stimulates neurogenesis in young adult rats (Jin et al., 2001), enhanced ENCAM immunoreactivity in SVZ of 55–70-week-old rats (Sato et al., 2001). As with our present results, this finding indicates that the capacity to override whatever endogenous factors suppress neurogenesis in the aged brain is not restricted to DG.
The possibility that the brain can be stimulated to produce new cells with the ability to replace neurones that are damaged or lost in neurological diseases could have important therapeutic implications. However, stroke and most chronic neurodegenerative disorders usually occur on a backdrop of advanced age, and the ability of various tissues to generate stem and other precursor cells may be compromised with aging. In addition, as has been suggested for the haematopoietic system, aging could diminish the ability of newly proliferating cells to respond to pathological conditions, despite normal basal levels of cytogenesis (Globerson, 1999). The finding that growth factors can stimulate neurogenesis in the aged mouse brain is significant in that, like other recent studies discussed above (Kuhn et al., 1996; Cameron & McKay, 1999; Sato et al., 2001; Kempermann et al., 2002), it demonstrates that the therapeutic possibilities of stimulating endogenous cerebral neurogenesis are not precluded by age alone. Moreover, because some growth factors can be administered systemically to stimulate neurogenesis (Wagner et al., 1999; Aberg et al., 2000), it suggests a relatively non-invasive means for exploring this approach to treatment.
Growth factors and BrdU administration in vivo
Young adult (3 months old, 25–35 g) and aged (20 months old, 40–50 g) male CD1 mice (Charles River Laboratories) were housed in standard cages and maintained on a 12-h light/12-h dark cycle. Experiments were conducted according to policies on the use of animals of the Society for Neuroscience and were approved by local committee review. Mice were anaesthetized with 2% isoflurane in 70% N2O/30% O2 and implanted with an osmotic minipump (Alzet 1007D; Durect Corporation, Cupertino, CA, USA). The cannula was placed in the right lateral ventricle 3 mm deep to the pial surface, −0.6 mm anteroposterior relative to bregma, and 1.3 mm lateral to the midline. Each mouse was treated for 3 days with 0.5 µL h−1 of recombinant human HB-EGF (10 µg mL−1, R & D Systems, Minneapolis, MN, USA) in aCSF (128 mm NaCl, 2.5 mm KCl, 0.95 mm CaCl2, 1.99 mm MgCl2) (n = 4), FGF-2 (10 µg mL−1, R & D Systems) in aCSF (n = 4), or aCSF alone (n = 6). BrdU (50 mg kg−1, Sigma) was dissolved in saline and given intraperitoneally, twice daily at 8-h intervals, for the same three consecutive days, and mice were killed 1 week later.
Brains (five per condition) were removed after perfusion with saline and 4% paraformaldehyde in PBS. Adjacent 50-µm sections, corresponding to coronal coordinates interaural 4.98–3.82 mm, bregma 1.18 to bregma 0.02 mm (SVZ) and interaural 2.46 to 1.34 mm, bregma −1.34 to bregma −2.46 (DG), were cut with a cryostat and stored at −80 °C. Sections were pretreated with 50% formamide, 280 mm NaCl and 30 mm sodium citrate at 65 °C for 2 h, incubated in 2 m HCl at 37 °C for 30 min, and rinsed in 0.1 m boric acid (pH 8.5) at room temperature for 10 min. Sections were incubated in 1% H2O2 in PBS for 15 min, in blocking solution (2% goat serum, 0.3% Triton X-100 and 0.1% bovine serum albumin in PBS) for 2 h at room temperature, and with mouse monoclonal anti-BrdU antibody (Roche, Indianapolis, IN, USA; 2 µg mL−1) at 4 °C overnight. Sections were washed with PBS, incubated with biotinylated goat-anti-mouse secondary antibody (Vector, 1 : 200) for 2 h at 25 °C, washed, and placed in avidin-peroxidase conjugate (Vector) solution for 1 h. The horseradish peroxidase reaction was detected with 0.05% diaminobenzidine (DAB) and 0.03% H2O2. Processing was stopped with H2O and sections were dehydrated through graded alcohols, cleared in xylene and coverslipped in permanent mounting medium (Vector). Sections were examined with a Nikon E800 epifluorescence microscope.
Sections were fixed with 4% paraformaldehyde in PBS for 1 h at room temperature, washed twice with PBS, and incubated in 2 m HCl at 37 °C for 1 h. After washing again, sections were incubated with blocking solution, then with primary antibodies at 4 °C overnight, and with secondary antibodies in blocking solution at room temperature for 2 h. The primary antibodies used were: mouse monoclonal anti-BrdU (Roche; 2 µg mL−1), sheep polyclonal anti-BrdU (Biodesign, Saco, ME, USA; 25 µg mL−1), affinity-purified goat polyclonal anti-Neuro D and anti-DCX (Santa Cruz Biotechnology; 1 : 100), mouse monoclonal anti-GFAP (Sigma, St. Louis, MO, USA; 1 : 200) and mouse monoclonal anti-rat nestin (BD PharMingen, San Diego, CA, USA; 1 : 400). The secondary antibodies were: rhodamine-conjugated rat-absorbed donkey anti-mouse and anti-sheep IgG (Jackson ImmunoResearch, West Grove, PA, USA; 1 : 200), and fluorescein isothiocyanate (FITC)-conjugated pig anti-goat and donkey anti-mouse IgG (Jackson ImmunoResearch; 1 : 200). Fluorescence signals were detected with a Nikon PCM-2000 laser-scanning confocal microscope and results were recorded with Simple PCI Imaging software (Compix, Cranberry Township, PA, USA). Controls included omitting or preabsorbing the primary and omitting the secondary antibody.
Quantification of BrdU-incorporated cells
BrdU-positive cells in SGZ and SVZ were counted blindly in 5–7 DAB-stained, 50-µm coronal sections per animal, spaced 200 µm apart. Cells were counted under high-power on a Nikon E800 microscope with Magnifire digital camera, and the image was displayed on a computer monitor. Results were expressed as the average number of BrdU-positive cells per section.
Quantitative data were expressed as mean ± SEM from at least three experiments. anova and Student's t-test were used for statistical analysis, with P < 0.05 considered significant.