SEARCH

SEARCH BY CITATION

Keywords:

  • Adult neurogenesis;
  • Doublecortin;
  • Neural stem/progenitor cells;
  • Stem/progenitor cell transplantation;
  • Stem cell aging Stem cell differentiation

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

Neurogenesis in the dentate gyrus (DG) declines severely by middle age, potentially because of age-related changes in the DG microenvironment. We hypothesize that providing fresh glial restricted progenitors (GRPs) or neural stem cells (NSCs) to the aging hippocampus via grafting enriches the DG microenvironment and thereby stimulates the production of new granule cells from endogenous NSCs. The GRPs isolated from the spinal cords of embryonic day 13.5 transgenic F344 rats expressing human alkaline phosphatase gene and NSCs isolated from embryonic day 9 caudal neural tubes of Sox-2:EGFP transgenic mice were expanded in vitro and grafted into the hippocampi of middle-aged (12 months old) F344 rats. Both types of grafts survived well, and grafted NSCs in addition migrated to all layers of the hippocampus. Phenotypic characterization revealed that both GRPs and NSCs differentiated predominantly into astrocytes and oligodendrocytic progenitors. Neuronal differentiation of graft-derived cells was mostly absent except in the dentate subgranular zone (SGZ), where some of the migrated NSCs but not GRPs differentiated into neurons. Analyses of the numbers of newly born neurons in the DG using 5′-bromodeoxyuridine and/or doublecortin assays, however, demonstrated considerably increased dentate neurogenesis in animals receiving grafts of GRPs or NSCs in comparison with both naïve controls and animals receiving sham-grafting surgery. Thus, both GRPs and NSCs survive well, differentiate predominantly into glia, and stimulate the endogenous NSCs in the SGZ to produce more new dentate granule cells following grafting into the aging hippocampus. Grafting of GRPs or NSCs therefore provides an attractive approach for improving neurogenesis in the aging hippocampus.

Disclosure of potential conflicts of interest is found at the end of this article.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

Dentate gyrus (DG) is one of the few brain regions that show plasticity and neurogenesis in the adult. This is evidenced by the recruitment of new granule cells to the hippocampal circuitry all through life [1, [2], [3], [4], [5]6]. However, the extent of new granule cell addition dwindles dramatically by middle age [7, [8], [9]10]. Because the persistent insertion of new neurons to the circuitry is an important aspect of hippocampal-dependent learning and memory functions, an enormous waning of neurogenesis may be a major impediment for sustaining hippocampal functions during aging. Although the linkage between diminished neurogenesis and impairments in hippocampal-dependent functions during old age is still unresolved [11, [12], [13], [14]15], it is deemed that reduced neurogenesis contributes to age-related cognitive deficits. Consequently, there is immense attention to both understanding the mechanisms of decreased neurogenesis and developing strategies that enhance neurogenesis during aging [15, [16], [17], [18], [19], [20]21].

Investigation of discrete regulatory steps of DG neurogenesis in rats insinuates that aging does not lower either the extents of neuronal differentiation and survival of newly born cells [10] or the number of putative neural stem/progenitor cells (NSCs) in the subgranular zone (SGZ) of the DG [22]. However, the number of proliferating NSCs in the SGZ decreases greatly at middle age [22, 23]. This is considered to be due to age-related changes in the DG milieu, because studies imply that various factors that are known to stimulate the proliferation of NSCs exhibit considerable decline at middle age. These comprise fibroblast growth factor-2 (FGF-2), insulin-like growth factor-1 (IGF-1), vascular-endothelial growth factor (VEGF), brain-derived neurotrophic factor (BDNF), neuropeptide Y, and phosphorylated cyclic-AMP response-element-binding protein [24, [25]26]. Moreover, the concentration of glucocorticoids, one of the negative regulators of neurogenesis, intensifies at middle age [27], suggesting that the microenvironment of the DG becomes nonconducive for greater levels of neurogenesis at middle age. Hence, strategies that alter the DG milieu and promote enhanced proliferation of NSCs may increase neurogenesis in the aging hippocampus.

We hypothesize that grafting of fresh glial-restricted progenitors (GRPs) or NSCs into the aging hippocampus may enrich the DG microenvironment and thereby stimulate an increased production of new granule cells from endogenous NSCs. We characterized DG neurogenesis in the hippocampi of middle-aged F344 rats after grafting of GRPs from embryonic day 13.5 transgenic F344 rats expressing human alkaline phosphatase (AP) gene or NSCs from the embryonic day 9 caudal neural tubes of Sox-2:EGFP transgenic mice. At 3 weeks postgrafting, the survival and differentiation of grafted cells were analyzed. Additionally, new cells/neurons born in the SGZ after grafting were rigorously characterized. Dentate neurogenesis was similarly quantified from middle-aged rats receiving sham-grafting surgery and naive middle-aged rats for comparison. We chose middle-aged hippocampus for studying the effects of GRP and NSC grafting because the window of major decline in neurogenesis during aging occurs between young age and middle age [23]. Our results confirm the hypothesis and suggest that, although NSCs can generate neurons in the hippocampus, their predominant effect is to mobilize endogenous stem cells.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

Animals and Antibodies

Middle-aged (12 months old) male Fischer 344 rats were obtained from the National Institute of Aging colony at Harlan Sprague Dawley (Indianapolis, http://www.harlan.com). Duke University's Institutional Animal Care and Use Committee and the animal studies subcommittee of the Durham Veterans Affairs Medical Center approved all experiments performed in this study. Immunohistochemical studies utilized both monoclonal and polyclonal primary antibodies. The monoclonal primary antibodies comprised anti-5′-bromodeoxyuridine (BrdU) (mouse 1:20, Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com; rat 1:200, Serotec Ltd., Oxford, U.K., http://www.serotec.com), anti-neuron-specific nuclear antigen (anti-NeuN, mouse, 1:1,000; Chemicon, Temecula, CA, http://www.chemicon.com), anti-AP (mouse, 1:200; Accurate Chemical, Westbury, NY, http://www.accuratechemical.com), anti-glial fibrillary acidic protein (anti-GFAP, mouse, 1:1,000; Chemicon), anti-rip (mouse, 1:200; Developmental Studies Hybridoma Bank, Iowa City, IA, http://www.uiowa.edu/∼dshbwww), anti-NG2 (1:500; Chemicon), anti-FGF-2 (1:200, Chemicon), anti-VEGF (1:200, Chemicon), anti-IGF (1:200, Chemicon), and anti-green florescent protein (anti-GFP, 1:200; Chemicon). The polyclonal primary antibodies included anti-BDNF (1:200; Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com), anti-Olig-2 (rabbit, 1:500; Chemicon), anti-S-100β (rabbit, 1:1,000; Chemicon), anti-β-III tublin-1 (anti-TuJ-1, rabbit, 1:1,000; Covance, Princeton, NJ, http://www.covance.com), anti-GFAP (rabbit, 1:1,000; Dako, Glostrup, Denmark, http://www.dako.com), and anti-doublecortin (anti-DCX, 1:250; Santa Cruz Biotechnology). The secondary antibodies included biotinylated anti-mouse, anti-rat, and anti-rabbit IgG from Vector Laboratories (1:100; Burlingame, CA, http://www.vectorlabs.com); anti-mouse, anti-rat, and anti-rabbit IgG conjugated to Alexa Fluor (488 and 594); and streptavidin fluorescein/Texas Red from Molecular Probes (Eugene, OR, http://www.probes.invitrogen.com).

Preparation of AP+ GRPs and Sox-2:EGFP Neural Stem Cells for Grafting

The donor GRPs isolated from the spinal cords of embryonic day 13.5 (E13.5) transgenic F344 rats that express the marker gene human placental AP [28, 29] and the NSCs isolated from E9 caudal neural tubes of Sox-2:EGFP transgenic mice brains [30, [31]32] were expanded in vitro using standard methods [29, 30, 32]. The cultured GRPs were dissociated, washed, and resuspended in a fresh culture medium. The NSCs were labeled with BrdU in vitro, dissociated, washed, and resuspended in a fresh medium. The labeling procedure mainly comprised overnight incubation of NSCs in a medium containing 0.5 μM BrdU. Prior to grafting, the cells were rinsed thoroughly to remove excess BrdU. For preparation of both donor cells for grafting, the dissociated cells were centrifuged and cells were counted, and the viability of cells was assessed using trypan blue exclusion test in a hemocytometer. The live and dead cells were counted and the percentage of live cells relative to the total number of cells was calculated. The final volume was adjusted to a cell density of 0.55 × 105 viable cells per microliter of the culture medium for GRPs and 1.3 × 105 viable cells per microliter for NSCs, aliquoted into centrifuge tubes and stored on ice. A few aliquots were cultured on poly-l-lysine (0.05 mg/ml) coated 35-mm tissue culture plates for up to 6 days to determine their differentiation potential at the time of grafting. Cultures were maintained in a differentiation medium comprising Neurobasal (96.4 ml/100 ml; Invitrogen, Carlsbad, CA, http://www.invitrogen.com), B-27 nutrient medium with retinoic acid (2 ml/100 ml; Invitrogen), l-glutamine (1.25 ml/100 ml; Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com), penicillin (40 U/ml), streptomycin (0.1 μg/ml), and Fungizone (0.004%). The presence of neurons, astrocytes, and oligodendrocytes in GRP+ and NSC cultures was ascertained using immunofluorescence for TuJ-1, GFAP, and rip [33]. The BrdU labeling index of NSC suspension was ascertained using BrdU immunostaining of one-hour cultures of NSCs [34]. This revealed that 98% of cells (mean ± SEM = 98.0 ± 1.5, n = 4) in the suspension were immunopositive for BrdU. We chose Sox-2:EGFP mice for collecting NSCs because Sox-2:EGFP expression in NSCs allowed collection of a purified population of NSCs as donor cells in this study. Additionally, because Sox-2:EGFP expression disappears once these cells differentiate into neuronal or glial phenotypes, the use of Sox-2:EGFP cells labeled with BrdU was helpful for determining whether most NSCs differentiate or remain undifferentiated following grafting into the host brain.

Analyses of Neurotrophic Factors Within AP+ GRPs and Sox-2:EGFP NSCs In Vitro

To determine whether AP+ GRPs and Sox-2:EGFP NSCs synthesize neurotrophic factors that are considered positive regulators of dentate neurogenesis, we performed extensive immunocytochemical analyses of these cells using antibodies specific for various neurotrophic factors. We plated both GRPs and Sox-2:EGFP cells on poly-l-lysine coated dishes using Dulbecco's modified Eagle's medium (DMEM) alone. Cultures were terminated using 2% paraformaldehyde solution either at 1 hour or at 24 hours after plating and processed for immunostaining. The cultures were washed in phosphate-buffered saline (PBS), blocked using appropriate serum (10%), and incubated overnight in primary antibody solutions of BDNF, IGF-1, FGF-2, or VEGF (1:200). Following this, cultures were rinsed thrice in PBS, incubated in either goat anti-mouse IgG or anti-rabbit IgG conjugated to Alexa Fluor 488/594 for 1 hour, washed thoroughly in PBS, and counterstained using 4,6-diamidino-2-phenylindole. Negative control cultures (where primary antibody step was omitted) were processed in parallel to rule out any nonspecific immunoreaction. The cultures were washed again and covered with a slow-fade mounting medium (Molecular Probes) and observed under a florescent microscope for the expression of various neurotrophic factors.

Transplantation Procedure

The grafting surgery was done in four sessions (n = 3 rats per session; total = 12 rats; six for grafting GRPs and six for grafting NSCs). The host middle-aged rats were anesthetized and fixed into a stereotaxic apparatus. The grafting procedure is described in earlier reports [35, 36]. One microliter of the cell suspension (containing 55,000 GRPs or 130,000 Sox-2:EGFP NSCs) was injected into each of the three sites in the hippocampus bilaterally. We chose to transplant differing amounts of GRPs and NSCs per graft because our preliminary studies suggested that GRPs survive grafting into the adult brain very well, whereas NSCs are more vulnerable to grafting-related trauma. The following stereotaxic coordinates were employed for implantation of the grafts into the hippocampus: (a) anteroposterior = 3.1 mm from bregma, lateral (L) = 1.5 mm from midline, ventral (V) = 3.5 mm from the surface of brain; (b) anteroposterior = 4.1 mm, L = 3.0 mm, V = 3.5 mm; and (c) anteroposterior = 4.8 mm, L = 4.0 mm, V = 4.0 mm. Grafting was performed within 3–4 hours after the preparation of cell suspension. Rats receiving mouse NSC grafts were immunosuppressed using daily subcutaneous injections of the cyclosporine A (12 mg/kg body weight). To determine the effects of grafting surgery on neurogenesis, a group of middle-aged rats (n = 6) received sham-grafting surgery, which involved 1.0-μl injections of the tissue culture medium at the above stereotaxic coordinates. In addition, to ascertain the changes in the extent of neurogenesis in grafted and sham-grafted animals, age-matched naive control rats (n = 6) were also included in this study.

Characterization of Transplant Survival and Differentiation

Three weeks after grafting or sham-grafting surgery, rats were perfused transcardially with 4% paraformaldehyde. Brains were cryoprotected in sucrose and sliced coronally (30-μm-thick sections) through the hippocampus using a cryocut and collected serially in phosphate buffer. Every 10th section was stained for Nissl, and the sections were scanned to identify the location of transplants in relation to the different hippocampal cell layers. In all animals, transplants were located either predominantly inside the hippocampus or partly inside the hippocampus. When transplants were located only partly inside the hippocampus, they projected into the corpus callosum above and/or into the lateral ventricle below. For further analyses of transplants, we chose transplants that were predominantly intrahippocampal. In animals receiving GRP cell grafts, every 10th section through each chosen transplant was processed for AP immunostaining, whereas, in animals receiving grafts of NSCs, every 10th section through each chosen transplant was processed for BrdU immunostaining using the avidin-biotin-complex (ABC) method, as detailed in our recent reports [10, 37]. These BrdU immunostained sections were later used for quantification of the yield of surviving cells per graft using the optical fractionator method.

To identify the presence of neurons, astrocytes, and oligodendrocytes among surviving grafted cells in rats receiving GRP cell transplants, representative sections through transplants were processed for dual immunofluorescence for AP and markers of neurons (NeuN, Tuj-1), astrocytes (GFAP, S-100β), and oligodendrocytes (NG2, olig-2, rip). In addition, to ascertain proliferating grafted cells among cells derived from transplanted GRP, BrdU, and AP, dual immunostaining was performed. In animals receiving grafts of BrdU-labeled Sox-2:EGFP NSCs, dual immunofluorescence staining was performed for BrdU and the above-mentioned markers of neurons and glia. In addition, to identify undifferentiated Sox-2:EGFP NSCs in the transplant, we stained representative sections using an antibody to enhanced green fluorescent protein (EGFP). For visualizing AP and markers of neurons or glia using dual immunofluorescence, free-floating sections containing grafts were washed in PBS, blocked in 3% normal horse serum, incubated overnight in a monoclonal antibody against AP, washed in PBS, and treated with goat anti-mouse IgG conjugated to Alexa Fluor 594. Following this, sections were rinsed in PBS, incubated overnight in antibodies against markers of neurons or glia, washed with PBS, treated with a biotinylated secondary antibody, washed in PBS, and incubated in streptavidin fluorescein solution, rinsed again in PBS, and mounted on clean slides using a slow-fade mounting medium (Molecular Probes). For demonstrating proliferating cells among grafted GRPs, sections were first processed for AP staining as mentioned above and then proceeded for BrdU pretreatment protocols [37] and incubated overnight in rat anti-BrdU and developed using goat anti-rat IgG conjugated to Alexa Fluor 488. To demonstrate BrdU and markers of neurons or glia in cells derived from NSC grafts, free-floating sections containing grafts were first washed in PBS, subjected to BrdU pretreatment protocols [37], rinsed in PBS, blocked in 3% normal goat serum, and incubated overnight in rat anti-BrdU solution. The sections were then washed in PBS, treated with goat anti-rat IgG conjugated to Alexa Fluor 594, and washed again in PBS. Following this, sections were blocked in normal serum, incubated overnight in antibodies against neuronal or glial markers (1:100–1:1,000), washed in PBS, and treated with an appropriate secondary antibody (biotinylated goat anti-mouse IgG/goat anti-rabbit IgG). The sections were then washed in PBS, incubated in streptavidin fluorescein solution, rinsed in PBS, and mounted on clean slides using a slow-fade mounting medium (Molecular Probes). Dual-labeled cells in all sections were analyzed via Z-section analyses in a confocal laser-scanning microscope (LSM-410; Carl Zeiss, Jena, Germany, http://www.zeiss.com; [10]). To identify undifferentiated Sox-2:EGFP cells in NSC grafts, representative sections were incubated overnight in mouse anti-GFP (1:200) and then treated with goat anti-mouse Alexa Fluor 488 (1:200) for 60 minutes.

Analyses of Newly Born Cells Using BrdU in Animals Receiving GRP Cell Grafts

In order to quantify newly born cells and neurons following GRP cell grafting, one intraperitoneal injection of BrdU (Sigma) at a dose of 100 mg/kg of body weight was given daily to animals in this group (n = 6) for 12 consecutive days, commencing from the day of grafting. For comparison, animals receiving sham-grafting surgery (n = 6) and naïve control animals (n = 6) were also given similar BrdU injections. Animals in all groups were perfused with 4% paraformaldehyde at 10 days after the last BrdU injection. In animals receiving GRP cell grafts and sham-grafting surgery, this time point is equivalent to 3 weeks after grafting surgery, as described above for characterization of transplant survival and differentiation. Every 15th 30-μm-thick section through the entire hippocampus was selected in each of the animals belonging to the above three groups and processed for single BrdU immunostaining using a monoclonal antibody to BrdU. The BrdU immunostaining was performed using the ABC method (Elite ABC kit; Vector) with diaminobenzidine as the chromogen. The detailed methods for visualization of BrdU in newly born cells are described in our earlier report [37]. The immunostained sections were mounted on gelatin-coated slides, air-dried, counter-stained with hematoxylin, dehydrated, cleared, and coverslipped. The BrdU immunostained sections were used for quantification of the total number of newly born cells added to the SGZ/granule cell layer (GCL) over a period of 12 days following grafting of GRP cells or sham-grafting surgery using the optical fractionator method. For quantification of the effects of GRP cell grafts on the number of newly born cells added to the SGZ/GCL, we chose hippocampi where GRP cell grafts were predominantly intrahippocampal. Additional series of sections (6–10 per animal) in all groups were processed for BrdU and NeuN dual immunofluorescence, and dual-labeled cells (i.e., cells positive for both BrdU and NeuN) in the SGZ/GCL were analyzed via Z-section analyses in a confocal laser scanning microscope (LSM-410; [10]).

Assessment of Neurogenesis Using DCX Immunostaining

In all animals belonging to different groups (animals receiving GRP cell grafts, NSC grafts or sham-grafting surgery, and naïve control animals), we analyzed neurogenesis using DCX immunostaining of serial (every 15th) sections through the entire hippocampus. The choice of DCX as a marker of new neurons in the DG is based on our earlier finding that neurons visualized with DCX immunostaining in the adult rat DG are new neurons that are predominantly born during the 12 days prior to euthanasia [37]. Sections were processed for DCX immunostaining using a polyclonal antibody to DCX and the ABC method as described in an earlier report [38]. The peroxidase reaction was visualized using vector gray as the chromogen, and the sections were counterstained using nuclear red staining. These sections were employed for assessment of the status of dentate neurogenesis via quantification of the total number of DCX+ neurons in the SGZ and GCL using the optical fractionator method.

Quantification of BrdU+ Cells and DCX+ Neurons Using the Optical Fractionator Method

The BrdU and DCX positive cells in the SGZ (two-cell-thick region from the inner margin of the dentate GCL) and GCL were counted in the following groups: hippocampus receiving GRP cell grafts, hippocampus receiving sham-grafting surgery, and naïve control hippocampus (n = 5 per group). In hippocampus receiving NSC grafts (n = 5), DCX+ new neurons in the SGZ and GCL and BrdU+ grafted cells dispersed in the entire hippocampus were measured. All cell counts were performed using the StereoInvestigator system (MicroBrightField Inc., Williston, VT, http://www.mbfbioscience.com). These measurements utilized every 15th section through the entire septotemporal extent of the hippocampus for counting cells in the SGZ and GCL and every 10th section through the entire extent of transplants for counting the dispersed BrdU+ cells from NSC grafts. The StereoInvestigator system consisted of a color digital video camera (Optronics Inc., Muskogee, OK, http://www.optronicsinc.com) interfaced with a Nikon E600 microscope. In each animal, cells were counted from 50–200 randomly and systematically selected frames (each measuring 40 × 40 μm, 0.0016 mm2 area) in every selected section using the 100× oil immersion objective lens. The numbers and densities of frames were determined by entering the parameter grid size in the optical fractionator component of the StereoInvestigator system [37].

We cut 30-μm-thick sections through the hippocampus using a cryostat. Confirmation of the section thickness using the StereoInvestigator system incorporating XYZ stage controller equipped with z-axis position control (Ludl Electronic Products Ltd, Hawthorne, NY, http://www.ludl.com) suggested that the variability among sections is minimal (i.e., ± 1 μm). However, following immunostaining, sections showed significant shrinkage along the z-axis. Measurement of the thickness of sections in different regions of the DG using the StereoInvestigator system described above revealed that, in sections processed for BrdU or DCX immunostaining, the average thickness was reduced to 53% of the initial thickness. Hence, at the time of data collection, the thickness was 16 μm for both BrdU and DCX immunostained sections. For cell counting, in every section, the contour of GCL/SGZ area or transplant area was first delineated using the tracing function. The optical fractionator component was then activated, and the number and location of counting frames and the counting depth for that section were determined by entering parameters such as the grid size, the thickness of top guard zone (4 μm), and the optical dissector height (i.e., 8 μm). The grid size was 75 × 75 μm for all cell counts except for the measurement of BrdU+ cells derived from NSC grafts, where 200 × 200 μm was chosen. A computer-driven motorized stage then allowed the section to be analyzed at each of the counting frame locations. In every counting frame location, the top of the section was set, after which the plane of the focus was moved 4 μm deeper through the section (guard zone) to get rid of the problem of uneven section surface. This plane served as the first point of the counting process. All cells that came into focus in the next 8-μm section thickness were counted if they were entirely within the counting frame or touching the upper or right side of the counting frame. Based on the above parameters and cell counts, the StereoInvestigator program calculated the total number of BrdU/DCX positive cells per SGZ and GCL utilizing the optical fractionator formula [37].

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

Expansion, Maintenance, and Characterization of AP+ GRPs and Sox-2:EGFP Neural Stem Cells

Expansion, maintenance, and characterization of both AP+ GRPs and Sox-2:EGFP NSCs have been described in our earlier studies [29, 30]. The GRP cultures were maintained and expanded on plates coated sequentially with laminin (20 μm/ml) and fibronectin (250 μm/ml) and in a medium comprising DMEM-F12 supplemented with FGF-2 (20 ng/ml), platelet-derived growth factor (20 ng/ml), and neurotrophin-3 (20 ng/ml) as previously described in Kalyani et al., [32]. The Sox-2:EGFP NSCs were expanded in vitro using adherent cultures that included dishes coated with fibronectin (250 μM/ml) and a culture medium comprising DMEM-F12, FGF-2 (20 ng/ml), and 1% serum [29, 30]. In vitro characterization revealed that GRPs comprise cells that are immunopositive for A2B5 (a marker of glial progenitors) and nestin (a marker of undifferentiated cells) but do not include cells that are immunoreactive for GFAP (putative marker of stem cells and astrocytes) and polysialic acid-neural cell adhesion molecule (PSA-NCAM; a marker of immature neurons) [29]. The Sox-2:EGFP NSCs, on the other hand, did not contain cells that are immunopositive for A2B5, MAP-2, or PSA-NCAM (markers of neurons) but included cells that are immunopositive for putative NSC markers such as Sox-1, Sox-2, nestin, GFAP, and NG2 [30].

Expression of Neurotrophic Factors in AP+ GRPs and Sox-2:EGFP NSCs

We analyzed GRPs and Sox-2:EGFP NSCs for the expression of various neurotrophic factors in vitro. We found that cells derived from both GRPs and NSCs (at 60 minutes after plating on poly-l-lysine coated dishes containing just DMEM) were robustly immunoreactive for many of the neurotrophic factors that are considered positive regulators of dentate neurogenesis. These include BDNF, FGF-2, VEGF, and IGF-1 (Fig. 1A1–1D3, 1F1–1I3). As negative control cultures (with omission of primary antibody during immunocytochemistry) displayed no such immunoreactivity (Fig. 1E1–1E3, 1J1–1J3), these observations imply that both GRPs and NSCs have the ability to secrete these neurotrophic factors following grafting. Interestingly, all of these neurotrophic factors were also found in the cytoplasm of cells derived from GRPs and NSCs at 24 hours after plating (data not shown).

thumbnail image

Figure Figure 1.. Analyses of neurotrophic factor expression and phenotypic differentiation of glial restricted progenitors and Sox-2:EGFP neural stem cells in vitro. Top panel: Expression of various neurotrophic factors by the glial restricted progenitor (GRP) and Sox-2:EGFP NSCs in vitro after 60 minutes of incubation on a poly-l-lysine coated dish containing Dulbecco's modified Eagle's medium. The left panel shows neurotrophic factors expressed by alkaline phosphatase positive glial restricted progenitors (A1–D3), whereas the right panel shows neurotrophic factors expressed by Sox-2:EGFP neural stem cells (F1–I3). The neurotrophic factors include BDNF (A1–A3, F1–F3), FGF-2 (B1–B3, G1–G3), IGF-1 (C1–C3, H1–H3), and VEGF (D1–D3, I1–I3). Note the absence of fluorescence in negative control samples where the primary antibody incubation step was omitted (E1–E3, J1–J3). Lower panels show in vitro differentiation of GRPs (K1–K3) and NSCs (L1–M3) after 6 days of incubation in a differentiation medium. Note that GRPs predominantly differentiate into GFAP+ astrocytes (K1). However, some of these cells also differentiate into rip+ oligodendrocytes (K2). In contrast, the Sox-2+ NSCs differentiate into GFAP+ astrocytes (L1, M1), TuJ-1+ neurons (L2), and rip+ oligodendrocytes (M2). Scale bar, 50 μm. Abbreviations: BDNF, brain-derived neurotrophic factor; DAPI, 4,6-diamidino-2-phenylindole; FGF-2, fibroblast growth factor-2; GFAP, glial fibrillary acidic protein; IGF-1, insulin-like growth factor-1; TuJ-1, β-III tublin-1; VEGF, vascular-endothelial growth factor.

Download figure to PowerPoint

Differentiation Potential of AP+ GRPs and Sox-2:EGFP NSCs In Vitro

Analyses of fractions of AP+ GRPs harvested for grafting studies in vitro using a differentiation medium (comprising DMEM-F12, B-27 nutrient medium with retinoic acid, and l-glutamine) revealed that these cells predominantly differentiate into GFAP+ astrocytes in vitro (Fig. 1K1, 1K3). Some of these cells also differentiated into rip+ oligodendrocytes (Fig. 1K1, 1K3). However, none of the cells derived from GRPs differentiated into TuJ-1+ neurons. On the other hand, in vitro characterization of samples of Sox-2:EGFP NSCs harvested for transplantation experiments demonstrated the ability of these cells to differentiate into all three CNS phenotypes (Fig. 1L1–1M3). However, the fraction of TuJ-1+ neurons was smaller (mean ± SEM = 4.7% ± 0.5%, n = 4 cultures) than glial fractions in these cultures. The glial population comprised GFAP+ astrocytes (67.0% ± 1.8%) and rip+ oligodendrocytes (28.0% ± 2.2%). Thus, although a vast majority of glial cells derived from NSCs differentiate into GFAP+ cells (presumably astrocytes based on characteristic morphology), we found ∼28% of cells differentiating into rip+ oligodendrocytes. This high percentage of oligodendroglial differentiation likely reflects the culture condition employed in this study. From the above, it is also clear that the two donor cell types chosen in this study are quite distinct; one (i.e., GRPs) is capable of differentiating mostly into astrocytes and oligodendrocytes, whereas the other one (i.e., NSCs) is capable of differentiating into all three CNS phenotypes.

Survival of AP+ GRPs and Sox-2:EGFP NSCs in the Aging Hippocampus

We analyzed the survival of grafted GRPs through AP immunostaining of serial brain sections from animals receiving GRP cell grafts. This revealed the location and distribution of GRP cell grafts. In all animals, grafts were located mostly inside the hippocampus with extensions into the DG (Fig. 2A1). However, portions of grafts also projected either into the corpus callosum above (likely a backflow along the needle track) or into the lateral ventricle below (Fig. 2A1). Characteristically, a vast majority of grafted GRPs remained close to the grafted site as a core with migration of only a few cells away from the core (Fig. 2A2). However, processes emanating from the core invaded the surrounding hippocampal tissue (Fig. 2A3). These processes presumably represent glial processes, as the subsequent phenotypic analyses revealed a large number of cells morphologically resembling astrocytes and expressing immunoreactivity for S-100β and GFAP were seen in the core of the graft (Fig. 3A1–3B15). As Sox-2+ NSCs were prelabeled with BrdU prior to grafting, we analyzed the survival of these cells through BrdU immunostaining of serial sections through grafts and optical fractionator cell counting. In contrast to GRPs, grafted NSCs migrated profusely into all layers of the DG and hippocampal CA1 and CA3 subfields (Fig. 2B1–2B3). In the DG, fractions of cells were engrafted into the SGZ (Fig. 2B4). When grafts were placed close to the end of the upper blade of GCL, fractions of cells also migrated into the stratum lucidum of the CA3 region (Fig. 2B2). Quantification revealed that the yield of surviving cells from NSC grafts is equivalent to 205% of injected cells (mean ± SEM = 205% ± 29.0%, n = 5), suggesting significant cell division after grafting. Thus, grafts of both GRPs and NSCs exhibit robust survival following grafting into the intact aging hippocampus. Additionally, the GRP cells tend to stay in clusters following grafting into the aging hippocampus, whereas NSCs exhibit widespread migration into all layers of the aging hippocampus.

thumbnail image

Figure Figure 2.. Distribution of cells derived from grafted glial restricted progenitor cells and Sox-2:EGFP neural stem cells. (A1–A3): Location and distribution of alkaline phosphatase positive (AP+) GRPs following grafting into the aging hippocampus, visualized through AP immunostaining (A1). Note that the graft is located inside the hippocampus with extensions into the dentate gyrus. Additionally, a vast majority of grafted GRPs remained close to the grafted site. However, glial processes emanating from the graft core invaded the surrounding hippocampal tissue (A2, A3). Scale bar: (A1) = 500 μm; (A2) = 200 μm; (A3) = 100 μm. (B1–B4): 5′-bromodeoxyuridine immunostaining showing the robust survival and profuse migration of grafted neural stem cells in the aging dentate gyrus and hippocampal CA1 and CA3 subfields (B1). (B2) and (B3) are enlarged views of boxed regions from (B1). (B2) illustrates migration of grafted cells into the strata oriens, pyramidale, lucidum, and radiatum of the CA3 region, whereas (B3) shows migration of grafted NSCs into all regions of the dentate gyrus. (B4) is a magnified view of a region of dentate SGZ and GCL from (B3). Note that grafted NSCs engraft into both SGZ and GCL. Scale bar: (B1) = 500 μm; (B2, B3) = 200 μm; (B4) = 50 μm. Abbreviations: DH, dentate hilus; GCL, granule cell layer; GRPs, glial restricted progenitors; ML, molecular layer; SGZ, subgranular zone; SL, stratum lucidum; SO, stratum oriens; SP, stratum pyramidale; SR, stratum radiatum.

Download figure to PowerPoint

thumbnail image

Figure Figure 3.. Phenotypic differentiation of glial restricted progenitors and Sox-2:EGFP neural stem cells following grafting into the middle-aged hippocampus. The left panel (A1–E15) illustrates differentiation of AP+ glial restricted progenitors (GRPs) in the aging hippocampus at 3 weeks postgrafting, visualized through dual immunofluorescence for AP and progenitor/glial antigens. Cells derived from grafted GRPs express GFAP immunoreactivity (A1–A15) and differentiate into S-100β+ mature astrocytes (B1–B15). Additionally, some of the grafted cells differentiate into Olig-2+ cells (C1–C15) and others into NG2+ cells (D1–D15). Arrows indicate dual-labeled cells. (E1–E15) illustrates examples of proliferating cells (indicated by arrows) within the core of GRP transplants, as determined by the incorporation of BrdU following daily injections of BrdU into the host for 12 days after grafting. Scale bar, (A1–E15) = 50 μm. The right panel (F1–K3) illustrates phenotypic differentiation of BrdU-labeled NSCs in different regions of the aging hippocampus at 3 weeks postgrafting, visualized through dual immunofluorescence for BrdU and different neural antigens. (F1–F3) illustrates neuronal differentiation of NSCs that migrated into the dentate subgranular zone (SGZ). Arrows indicate cells that express the neuron-specific nuclear antigen (NeuN), whereas arrowheads indicate cells that lack NeuN (presumably glia). (G1–H3) illustrates differentiation of NSCs into S-100β+ mature astrocytes (arrows) in the dentate SGZ (G1–G3) and the dentate molecular layer (H1–H3). Arrowheads denote S-100β+ mature host astrocytes. (I1–J3) show differentiation of NSCs into Olig-2+ cells (arrows) in the dentate SGZ (I1–I3) and the CA1 region (J1–J3). (K1–K3) show an example of NG-2 positive cells derived from grafted cells (arrows). (L1) and (L2) are examples of undifferentiated Sox-2:EGFP cells derived from Sox-2:EGFP NSC grafts (arrows) in the dentate hilus (L1) and the posterior subventricular zone (L2). Scale bar, (F1–L2) = 50 μm. Abbreviations: AP, alkaline phosphatase; BrdU, 5′-bromodeoxyuridine; GFAP, glial fibrillary acidic protein; LV, lateral ventricle; pSVZ, posterior subventricular zone.

Download figure to PowerPoint

Differentiation of GRPs and Sox-2:EGFP NSCs After Grafting into the Aging Hippocampus

Grafted GRPs and cells derived from them were identified using the AP transgene expression. As we used AP transgenic rats where expression of AP is stable over multiple generations as a source of GRPs, this method was adequate for tracking cells derived from grafted GRPs in the host brain. Our previous study showing that AP expression is not downregulated over time and expression persists for up to a year and a half after transplantation further supports the above protocol [39]. None of the GRPs differentiated into neurons following grafting into the aging hippocampus, based on both AP and TuJ-1 and AP and NeuN dual immunofluorescence assays. However, a large number of these cells differentiated into GFAP+ and S-100β+ cells, as evidenced by AP and GFAP and AP and S-100β dual immunofluorescence analyses (Fig. 3A1–3B15). A smaller fraction of these cells also differentiated into Olig-2 expressing cells (presumably oligodendrocytes; Fig. 3C1–3C15). The GFAP+ cells were conspicuous in the core as well as the periphery of grafts; however, morphologically, they did not resemble typical astrocytes found in the adult hippocampus. These cells exhibited very thick and short processes rather than the long tapering processes characteristic of mature astrocytes (Fig. 3A1–3A15), suggesting that the GFAP+ cells derived from GRP cell grafts are either relatively immature astrocytes or astrocytic progenitors. The S-100β+ cells were observed in both core and periphery of grafts, and phenotypical features of these cells were closer to mature astrocytes (Fig. 3D1–3D15) found in the adult hippocampus. On the other hand, the olig-2+ cells (presumably oligodendrocytes) were frequently located in the periphery of the transplant mass. As olig-2 is not specific for oligodendrocytes, we also characterized AP+ cells that are immunopositive for rip (a marker of mature oligodendrocytes) and NG2 (a marker of oligodendrocyte progenitors). We found that many AP+ cells express NG2 (Fig. 3D1–3D15) but none display rip immunoreactivity, suggesting that mature oligodendrocyte differentiation of grafted GRPs does not occur at the postgrafting time point examined in this study. However, the presence of NG2+ cells among AP+ cells implies that some of the grafted GRPs will eventually differentiate into mature oligodendrocytes.

We analyzed cells derived from grafts of NSCs in different regions of the hippocampus using BrdU and NeuN, BrdU and S-100β, BrdU and olig-2, and BrdU and NG2 dual immunofluorescence methods (Fig. 3F1–3K3). Analyses of NSC-derived cells engrafted into the SGZ/GCL revealed that 4% (mean ± SEM = 4.3% ± 0.6%) of migrated cells differentiated into NeuN+ mature neurons (Fig. 3F1–3F3), 37% (37% ± 0.2.6%) differentiated into S-100β+ mature astrocytes (Fig. 3G1–3G3), and 36% (36% ± 3.3%) differentiated into olig-2+ cells (presumably oligodendrocytes; Fig. 3I1–3I3). Thus, ∼23% of cells that migrated into the dentate SGZ exhibited immature characteristics. These cells likely remain as NSCs in the SGZ. In other regions of the hippocampus, none of the cells derived from NSC grafts differentiated into NeuN positive neurons. However, a significant fraction (39.0% ± 9.6%) of NSC-derived cells expressed S-100β and morphologically resembled astrocytes (Fig. 3H1–3H3). In addition, a fraction of cells (41% ± 7.8%) expressed olig-2 (Fig. 3J1–3J3). We also examined cells derived from NSCs (i.e., BrdU+ cells) for the expression of rip and NG2. Many BrdU+ cells displayed NG2 expression (Fig. 3K1–3K3) but none exhibited rip immunoreactivity, suggesting that differentiation of grafted NSCs into mature oligodendrocytes does not happen at the postgrafting time point examined in this study. Nevertheless, the occurrence of NG2+ cells among BrdU+ cells suggests that some of the grafted NSCs will eventually give rise to mature oligodendrocytes.

Thus, in the aging hippocampus, NSC-derived cells exhibit some neuronal differentiation only when migrated into the neurogenic zone. In non-neurogenic regions of the aging hippocampus, NSC-derived cells differentiate into glia. As most of the grafted NSCs differentiated, EGFP expression was lost in a great majority of BrdU+ grafted cells. However, a small fraction of undifferentiated cells (Sox-2:EGFP positive) was seen in the host hippocampus (Fig. 3L1), and some of the undifferentiated cells migrated into the posterior subventricular zone (Fig. 3L2). These undifferentiated cells retained robust Sox-2:EGFP expression, providing a clear demonstration that stem cells mostly differentiate and do not persist as undifferentiated cells after transplantation except in stem cell niches.

Effects of GRP Cell Grafting on Dentate Neurogenesis in the Aging Hippocampus

We first quantified newly born cells and neurons in middle-aged hippocampus following GRP cell grafting via BrdU immunostaining and BrdU and NeuN dual immunofluorescence assays, and the data were compared with results from naïve control hippocampus and hippocampus receiving sham-grafting surgery (n = 5 per group). In all groups, BrdU immunostaining visualized newly generated cells in the GCL and the SGZ. However, the density of BrdU + cells clearly appeared greater in middle-aged hippocampus receiving sham-grafting surgery or GRP cell grafts in comparison with naïve middle-aged hippocampus (Fig. 4A1–4C2). Quantification of the total number of new cells (i.e., BrdU+ cells) generated over a period of 12 days in the GCL and SGZ using the optical fractionator method revealed that addition of new cells to GCL is increased in hippocampus receiving GRP cell grafts in comparison with both naïve age-matched hippocampus (128% increase; p < .001) and hippocampus receiving sham grafting surgery (23% increase; p < .05; Fig. 4D1).

thumbnail image

Figure Figure 4.. Changes in the production of new cells in the dentate gyrus following grafting of glial restricted progenitors. Newly generated cells in the dentate subgranular zone and the granule cell layer of a naïve middle-aged rat (A1), a middle-aged rat receiving sham-grafting surgery (B1), and a middle-aged rat receiving GRP cell grafts (C1), visualized through 5′-bromodeoxyuridine (BrdU) immunostaining. Note that the density of newly born cells (i.e., BrdU+ cells) is greater in the middle-aged animals receiving either sham-grafting surgery or GRP cell grafts in comparison with the naïve control animal. (A2, B2) and (C2) are magnified views of regions from (A1, B1) and (C1). Scale bar: (A1, B1, C1) = 200 μm; (A2, B2, C2) = 100 μm. Bar chart (D1) illustrates the total number of new cells (i.e., BrdU+ cells) added to the SGZ and GCL over a period of 12 days. Note that addition of new cells is increased in rats receiving GRP cell grafts in comparison with both age-matched naïve rats (128% increase) and rats undergoing sham grafting surgery (23% increase). Abbreviations: DG, dentate gyrus; GCL, granule cell layer; GRP, glial restricted progenitor; ML, molecular layer; SGZ, subgranular zone.

Download figure to PowerPoint

In the naïve control hippocampus, 5,402 new cells were added to the SGZ/GCL over a period of 12 days, which was increased to 10,019 cells in animals receiving sham surgery and 12,340 in animals receiving GRP cell grafts. Further analyses of BrdU+ cells through BrdU and NeuN dual immunofluorescence using confocal microscopy (Fig. 5A1–5B3) revealed that the neuronal differentiation of newly born cells was 49% in hippocampus receiving GRP cell grafts, 59% neurons in naïve control hippocampus, and only 16% in hippocampus receiving sham-grafting surgery (Fig. 5C). Extrapolation of BrdU cell counts with neuronal differentiation data suggested that a far greater number of new neurons are added to the GCL following GRP cell grafting (Fig. 5D). The overall increase was 93% in comparison with naïve control hippocampus and 277% in comparison with hippocampus receiving sham-grafting surgery (Fig. 5D).

thumbnail image

Figure Figure 5.. Mature neurons among newly born cells in different groups. Neuronal differentiation of newly generated cells within the dentate subgranular zone and granule cell layer of a naïve control animal (A1–A3) and an animal receiving glial restricted progenitor cell grafts (B1–B3), analyzed by BrdU and NeuN dual immunofluorescence and confocal microscopy. Arrows indicate newly generated NeuN+ neurons. Scale bar, (A1–B3) = 25 μm. Bar chart (C) shows the percentages of NeuN+ neurons among newly born cells in different groups of animals. Note that neuronal differentiation of newly born cells is minimal in animals receiving sham-grafting surgery in comparison with both naïve animals and animals receiving GRP cell grafts. Bar chart (D) illustrates net neurogenesis based on the extrapolation of BrdU cell counts with neuronal differentiation data. Note that a far greater number of new neurons are added to the SGZ and GCL following GRP cell grafting. The overall increase was 93% in comparison with naïve control rats and 277% in comparison with rats receiving sham-grafting surgery. Abbreviations: BrdU, 5′-bromodeoxyuridine; DG, dentate gyrus; GCL, granule cell layer; GRP, glial restricted progenitor; SGZ, subgranular zone.

Download figure to PowerPoint

Thus, the effects of GRP cell grafting on dentate neurogenesis in the aging hippocampus are considerable. It is interesting that, although animals in the sham surgery group exhibited an increased number of BrdU-labeled cells (i.e., newly born cells), the overall numbers of newly born cells are less than animals receiving GRP cell grafts. Furthermore, fractions of newly born cells that differentiate into neurons are considerably less than that observed in naive control animals. Reduced differentiation of BrdU-labeled cells into neurons in the sham-surgery group (in comparison with the naive group) likely suggests that sham surgery leads to proliferation of mostly glial cells rather than NSCs in the hippocampus. This possibility is corroborated by the observation that comparable fractions of BrdU-labeled cells differentiate into neurons in animals receiving GRP cell grafts and naive control animals, implying that GRP grafting mostly leads to proliferation of NSCs. However, we cannot completely rule out the possibility that some of the increase observed in net neurogenesis following GRP cell grafting may be linked to promotion of neuronal cell fate of newly born cells by grafted GRPs. Additionally, because injections of BrdU after grafting would also label proliferating grafted cells, we wondered whether some of the BrdU-labeled cells quantified in the SGZ and GCL (for estimation of endogenous neurogenesis) were derived from the grafted GRPs. However, examination of BrdU+ cells via dual immunofluorescence for BrdU and AP (the transgene expressed in grafted GRPs) did not reveal any BrdU+ cell displaying AP immunoreactivity in the SGZ and GCL. This clearly suggests that none of the BrdU+ cells in the SGZ and GCL quantified in this study is derived from AP+ grafted GRPs. However, in the core of GRP transplants, we did see GRP-derived cells expressing BrdU, suggesting that some proliferation of GRPs occurs after grafting (Fig. 3E1–3E15).

We next analyzed newly born neurons in the entire GCL and SGZ through measurement of DCX+ neurons. In all groups, DCX immunostaining visualized newly formed neurons in the SGZ and GCL (Fig. 6A1–6C3). The cell bodies of DCX+ neurons were located in the SGZ and the inner third of the GCL. In naive hippocampus and hippocampus receiving sham-grafting surgery, DCX+ neurons were far fewer in number and morphologically appeared immature with horizontally oriented or basal dendrites (Fig. 6A1–6B3). Additionally, vertically oriented dendrites going through the GCL were infrequent and, when present, they were conspicuously shorter. In contrast, in middle-aged hippocampus receiving GRP cell grafts (Fig. 6C1–6C3), a significant fraction of DCX+ neurons exhibited relatively more mature phenotypes with vertically oriented dendrites extending into the outer two thirds of the dentate molecular layer [40]. The overall dendritic branching also appeared greater in these neurons in comparison with their counterparts in naive control hippocampus and hippocampus receiving sham-grafting surgery. Quantification of the number of DCX+ neurons in the entire DG revealed that addition of new neurons to the GCL is increased in rats receiving GRP cell grafts in comparison with both naïve control hippocampus (87% increase; p < .001) and hippocampus receiving sham-grafting surgery (237% increase; p < .001; Fig. 6D1). Because DCX+ cells in F344 rats represent neurons that are mostly added during the 2 weeks prior to euthanasia [10, 37] and animals were killed at 3 weeks postgrafting in this study, the results obtained with DCX analyses reflect neurogenesis that occurs at 2–3 weeks postgrafting. Thus, both BrdU and DCX assays suggest that neurogenesis in the aging hippocampus can be considerably enhanced through grafting of GRP cells. Additionally, it is clear that the increase in overall neurogenesis is not due to injury inflicted at the time of grafting surgery, as hippocampi receiving sham-grafting surgery do not upregulate neurogenesis.

thumbnail image

Figure Figure 6.. Changes in the production of doublecortin immunopositive neurons in the dentate gyrus following grafting of glial restricted progenitors. Newly generated neurons in the dentate subgranular zone and the granule cell layer of a naïve middle-aged rat (A1), a middle-aged rat receiving sham-grafting surgery (B1), and a middle-aged rat receiving glial restricted progenitor cell grafts (C1), visualized through doublecortin immunostaining. Note that the density of newly born neurons (i.e., DCX+ cells) is greater in the middle-aged animal receiving GRP cell grafts in comparison with the animal receiving either sham-grafting surgery or the naïve control animal. The morphology of DCX+ neurons appears immature with horizontally oriented or basal dendrites in the naive animal (A3) and the animal receiving sham-grafting surgery (B3) in comparison with the animal receiving GRP cell grafts (C3), whereas a significant fraction of DCX+ neurons exhibited relatively more mature phenotypes with vertically oriented dendrites. Scale bar: (A1, A3, B1, B3, C1, C3) = 100 μm; (A2, B2, C2) = 200 μm. Bar chart (D1) shows the absolute number of DCX+ neurons within the entire SGZ and GCL of different groups of animals. Note that addition of new neurons to the GCL is increased in rats receiving GRP cell grafts in comparison with both naïve rats (87% increase) and rats receiving sham-grafting surgery (237% increase). Abbreviations: DCX, doublecortin; DG, dentate gyrus; GCL, granule cell layer; GRP, glial restricted progenitor; SGZ, subgranular zone.

Download figure to PowerPoint

Effects of Grafting of Sox-2:EGFP NSCs on Dentate Neurogenesis in the Aging Hippocampus

We analyzed the effects of NSC grafts on DG neurogenesis through DCX immunostaining of serial sections through the entire hippocampus and optical fractionator cell counting. Similar to hippocampi receiving GRP cell grafts described above, in hippocampi receiving NSC grafts, the density of DCX + neurons appeared much greater than age-matched naive hippocampus and hippocampus receiving sham-grafting surgery (Fig. 7A1–7B3). Furthermore, a considerable fraction of DCX+ neurons in hippocampus receiving NSC grafts exhibited relatively more mature phenotypes with vertically oriented dendrites extending into the outer two thirds of the dentate molecular layer (Fig. 7B1–7B3). The total dendritic growth also appeared more extensive in these neurons than newly born neurons in naive control hippocampus and hippocampus receiving sham-grafting surgery. Quantification of the total number of DCX+ neurons demonstrated that the addition of newly born neurons to the DG increases considerably in hippocampus receiving grafts of NSCs (Fig. 7C1). The overall increase was 291% in comparison with the sham-surgery group and 118% in comparison with the naïve control group (Fig. 7C1). Thus, cells derived from grafted NSCs appear to influence the endogenous NSCs in the SGZ, as greater numbers of new dentate granule cells are produced in the aging hippocampus after NSC grafting. Interestingly, grafting of differing amounts of GRPs (55,000 live cells per graft) and NSCs (130,000 live cells per graft) did not result in statistically significant differences pertaining to their effects on dentate neurogenesis. A direct comparison of DCX+ newly born neurons between these two groups revealed that animals receiving NSC grafts exhibited only a 16% greater level of neurogenesis than animals receiving GRP grafts. Somewhat greater effects of grafted NSCs on host dentate neurogenesis from endogenous NSCs may be due to extensive migration of these cells into virtually all regions of the host hippocampus. The GRPs, on the other hand, were mostly found clustered at the site of grafting with minimal migration away from the transplant core.

thumbnail image

Figure Figure 7.. Changes in the production of doublecortin immunopositive neurons in the dentate gyrus following grafting of Sox-2:EGFP neural stem cells. Dentate neurogenesis revealed by doublecortin immunostaining in a rat receiving sham-grafting surgery (A1–A3) and a rat receiving Sox-2+ NSC grafts (B1–B3). Note that the density of DCX+ neurons in the rat receiving Sox-2+ NSC grafts is much greater, and, morphologically, the DCX+ neurons appear more mature in comparison with their counterparts in the rat receiving sham-grafting surgery. Scale bar: (A1, B1) = 200 μm; (A2, B2) = 100 μm; (A3, B3) = 50 μm. Bar chart (C1) illustrates the absolute number of DCX+ neurons within the entire SGZ and GCL of different groups of animals. Note that the overall addition of newly born neurons to the GCL is increased in rats receiving grafts of NSCs in comparison with both naïve rats (118% increase) and rats receiving sham-grafting surgery (291% increase). Abbreviations: DCX, doublecortin; DG, dentate gyrus; DH, dentate hilus; GCL, granule cell layer; SGZ, subgranular zone.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

This study provides the first evidence for the competence of grafts of both GRPs and NSCs to kindle the endogenous NSCs in the SGZ to create an enhanced number of new dentate granule cells in the aging hippocampus. Increased dentate neurogenesis in the aging hippocampus following grafting likely reflects a more conducive microenvironment for proliferation of endogenous NSCs in the presence of cells derived from grafts of GRPs and NSCs, as animals receiving only sham-grafting surgery did not exhibit any increase in dentate neurogenesis. The positive change in the milieu following grafting is likely mediated by the invasion of fresh astrocytes and oligodendrocytic progenitors, as most of the cells derived from grafts of GRPs and NSCs differentiated into these two cell types. Additionally, as in vitro characterization of GRPs and NSCs suggested synthesis of multiple neurotrophic factors such as BDNF, FGF-2, IGF-1, and VEGF by these cells, it is likely that some of the increase in neurogenesis observed after grafting is mediated by secretion of the above neurotrophic factors by grafted GRPs and NSCs.

Pattern of Integration of Grafts of GRPs and NSCs into the Aging Hippocampus

We employed two diverse populations of cells (GRPs and NSCs) to investigate their behavior and fate following grafting into the aging hippocampus. The GRPs were obtained from transgenic animals where all cells expressed AP under a ubiquitous promoter [29], which facilitated visualization of GRP-derived cells after grafting through simple AP immunostaining. The NSCs, in contrast, were harvested from mice where GFP was knocked in to the Sox-2 locus to obtain NSCs that are GFP+ [30, 31]. As these NSCs lose their GFP expression following differentiation into neurons or glia, we used BrdU to track NSCs following grafting. Transplanted GRPs differentiated into astrocytes and oligodendrocytic progenitors. Nonetheless, it appeared that some grafted GRPs remained as progenitors, as these cells expressed GFAP but lacked S-100β and rip/olig-2. Interestingly, cells derived from GRPs were mostly located in and around the graft sites, in sharp disparity to the widespread migration of GRPs detected along white matter tracts after grafting into the adult spinal cord [39, 41]. The incongruity reflects differences in both age of the host and the site of grafting between the two studies. Nevertheless, the current observations insinuate that the migration of GRPs is restricted in the aging brain, which is likely a consequence of swift differentiation of most GRPs following grafting. The likelihood of decreased cues for migration of GRPs in the aging hippocampus is improbable, since NSCs grafted into the age-matched hippocampus demonstrated extensive migration in this study. Additionally, grafted GRPs did not contribute neurons because no AP+ neurons could be found in the hippocampus of animals receiving GRP cell grafts, suggesting that both transdifferentiation of GRP-derived astrocytes into neurons [42] and dedifferentiation of GRPs into NSCs do not occur after grafting. This is consistent with our earlier studies showing that GRPs harvested from Sox-2:EGFP mice fail to dedifferentiate into Sox-2:EGFP NSCs (M.S. Rao, unpublished observations). Thus, transplanted GRPs display limited migration but readily differentiate into astrocytes and oligodendrocytes following grafting into the aging hippocampus.

Grafted NSCs exhibited pervasive migration, as they invaded the entire hippocampus including the dentate GCL and the hippocampal CA1 and CA3 cell layers. Some of the grafted NSCs also appeared to persist in the stem cell niches, as few EGFP+ cells could be visualized in the dentate gyrus and posterior subventricular zone at 3 weeks after transplantation. However, a great majority of cells from NSC grafts differentiated into astrocytes and oligodendrocytic progenitors and so lost their GFP expression. Some cells derived from NSC grafts also differentiated into neurons when they migrated into the SGZ/GCL but not when they engrafted into the white matter or the non-neurogenic regions of the hippocampus. This is in difference to results obtained earlier with both transplants of neuronal restricted precursors, which differentiate into neurons even when grafted into the white matter [39, 41], and committed postmitotic neurons from the fetal hippocampus, which differentiate into mature neurons even when grafted into the senescent host brain [36, 43]. Collectively, these observations indicate that, although NSCs have the proficiency to differentiate into neurons, the induction of neuronal phenotype from NSCs critically requires exogenous signals that seem to be present in only the neurogenic regions of the brain. Although the specific nature of these signals remains to be identified, studies suggest that the possible candidates that regulate neuronal fate-choice decision of NSCs include bone morphogenetic proteins [44], wnt proteins [45, 46], sonic hedgehog [47, [48], [49]50], and cytokines such as IL-1β and IL-6 released by astrocytes [51]. However, transplanted NSCs that migrated into the neurogenic SGZ did not differentiate exclusively into neurons, as the SGZ of the host hippocampus also contained glia derived from NSC grafts. This suggests that the behavior of grafted NSCs is analogous to the behavior of endogenous NSCs in the SGZ, where they generate neurons and glia throughout life. Thus, grafted NSCs exhibit robust survival, migrate extensively, and differentiate mostly into glia in all regions of the aging hippocampus. Furthermore, grafted NSCs that migrate into the neurogenic region of the aging hippocampus produce both neurons and glia. However, it remains to be seen whether NSCs that persist in the dentate gyrus would continue to produce new neurons and glia for prolonged periods in the aging hippocampus. Although a recent study suggests that in vitro expanded NSCs are incapable of self-renewal after grafting [52], careful long-term analyses of grafted NSCs are needed in future to resolve this issue.

Potential Mechanisms and Extent of Enhanced Neurogenesis After Grafting of GRPs and NSCs

The precise mechanisms by which grafts of NSCs or GRPs increase neurogenesis in the aging hippocampus are unknown. Nevertheless, increased neurogenesis after grafting is not due to injury or nonspecific release of neurotrophic factors associated with grafting surgery because animals receiving sham-grafting surgery exhibited decreased neurogenesis compared with naïve control animals. Increased neurogenesis might reflect induction of a more conducive microenvironment for proliferation of endogenous NSCs in the presence of cells derived from GRP and NSC grafts. This is because recent studies have shown that the major reason underlying declined dentate neurogenesis during aging is reduced proliferation of NSCs [10, 23, 53], which is likely due to multiple age-related alterations in the hippocampus. These include reduced FGF-2 activity and FGF-2 receptors in astrocytes, decreased concentrations of multiple NSC proliferation factors, and an increased quiescence of NSCs [22, 24, 25, 54]. The positive changes in the milieu following grafting are likely mediated by fresh astrocytes and oligodendrocytic progenitors permeating the aging hippocampus, as most of the cells derived from both GRP and NSC grafts differentiated into these two cell types. It is plausible that addition of these fresh glia increases the concentration of NSC proliferation factors, which in turn stimulates the proliferation of quiescent endogenous NSCs in the aging hippocampus. Indeed, our in vitro analyses of GRPs and NSCs reveal that these cells do synthesize a variety of neurotrophic factors that are considered positive regulators of dentate neurogenesis. These include BDNF, FGF-2, IGF-1, and VEGF. Thus, the effects of GRP and NSC grafts on dentate neurogenesis are likely mediated by the cell types derived from these grafts.

The overall extent of enhancement in neurogenesis was slightly (16%) greater with grafts of NSCs than with grafts of GRPs, although the differences were not significant statistically. There may be many possibilities for this difference. First, although both GRPs and NSCs seem to influence the proliferation of endogenous NSCs by altering their milieu via contribution of new glia, the GRPs did not give rise to new neurons even when grafted directly into the dentate SGZ and GCL. In contrast, NSCs migrating into the SGZ contributed new neurons to the GCL. Second, the timing and duration of growth factor delivery may be different between the two types of grafts. The GRPs, being lineage committed, likely differentiate rapidly into astrocytes and oligodendrocytic progenitors following grafting, leading to swift effects on the proliferation of endogenous NSCs. However, these early effects may not be adequate for increased proliferation of NSCs for prolonged periods, whereas grafts of NSCs, being noncommitted at the time of grafting, likely proliferate for prolonged periods and produce cohorts of fresh astrocytes and oligodendrocytic progenitors at different time points after grafting, leading to a sustained effect on the proliferation of endogenous NSCs. Third, the overall positive effects on microenvironment are likely greater with NSC grafts than with GRP cell grafts because GRPs exhibit very limited migration, whereas NSCs display pervasive migration throughout the hippocampus. Fourth, it is possible that NSCs provide a different set of trophic molecules than those derived from grafted GRPs, which is consistent with our comparative gene expression studies showing significant differences in growth factor profiles of NSCs and GRPs ([55] and M.S. Rao, unpublished observations). Overall, although both GRPs and NSCs are efficient for enhancing neurogenesis in the aging hippocampus, NSC grafts may be slightly better suited than GRP cell grafts. However, additional long-term studies are needed to dissociate differences in mediating neurogenesis between these two groups. Furthermore, unlike NSCs, GRPs do not differentiate into neurons [29], which may be an advantage for using GRPs as donor cells for enhancing neurogenesis in the aging hippocampus. This is because lack of neuronal differentiation may mitigate some of the potential side effects of neural cell grafting in the intact aging hippocampus, such as the development of inappropriate synaptogenesis or the formation of ectopic foci of neurons. Preventing the formation of ectopic foci of neurons is important, as such foci observed in the DG following status epilepticus or infusions of BDNF have been shown to contribute to epilepsy development [56, [57], [58], [59]60].

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

Grafting of GRPs or NSCs offers a novel approach for improving neurogenesis in the aging hippocampus. Continual incorporation of new neurons to the hippocampal circuitry is considered an important aspect of hippocampal-dependent learning and memory functions [15]. Furthermore, it is supposed that age-related waning of neurogenesis would diminish the threshold for neurodegenerative disorders such as depression and Alzheimer disease and the aged brain's capacity for coping and adapting to novel stimuli [61, 62]. Considering this, increased neurogenesis observed in the aging hippocampus after grafting of GRPs and NSCs appears beneficial for improving learning and memory function in the aged. However, rigorous studies at multiple time points after grafting of GRPs and NSCs are needed in future to ascertain the beneficial effects of these approaches for sustained increases in neurogenesis and lasting improvements in learning and memory function.

Disclosure of Potential Conflicts of Interest

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

The authors indicate no potential conflicts of interest.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

This research was supported by Grants from the National Institute for Aging (R01 AG 20924 to A.K.S.), the National Institute of Neurological Disorders and Stroke (R01 NS054780 and R01 NS043507 to A.K.S.), and the Department of Veterans Affairs (VA Merit Award to A.K.S.). M.S.R. and A.K.S. are sharing senior authorship.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References