SEARCH

SEARCH BY CITATION

Keywords:

  • adult neurogenesis;
  • aging;
  • common marmoset;
  • neural stem cells

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contribution
  9. References
  10. Supporting Information

Adult neurogenesis within the subgranular zone (SGZ) of the hippocampal dentate gyrus and the subventricular zone (SVZ) of the lateral ventricle (LV) has been most intensely studied within the brains of rodents such as mice and rats. However, little is known about the cell types and processes involved in adult neurogenesis within primates such as the common marmoset (Callithrix jacchus). Moreover, substantial differences seem to exist between the neurogenic niche of the LV between rodents and humans. Here, we set out to use immunohistochemical and autogradiographic analysis to characterize the anatomy of the neurogenic niches and the expression of cell type–specific markers in those niches in the adult common marmoset brain. Moreover, we demonstrate significant differences in the activity of neurogenesis in the adult marmoset brain compared to the adult mouse brain. Finally, we provide evidence for ongoing proliferation of neuroblasts within both the SGZ and SVZ of the adult brain and further show that the age-dependent decline of neurogenesis in the hippocampus is associated with a decrease in neuroblast cells.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contribution
  9. References
  10. Supporting Information

Adult neurogenesis has been well characterized in at least two specific areas of the brain, the subgranular zone (SGZ) of the hippocampal dentate gyrus (DG) and the subventricular zone (SVZ) of the lateral ventricle (LV). Newly generated cells of the DG were shown to mature into granule neurons of the granular cell layer (GCL) while cells generated in the SVZ seem to migrate longer distances into the olfactory bulb (OB) where they give rise to interneurons (Lois & Alvarez-Buylla, 1994). The way of migration into the OB seems to be substantially different between rodents and humans. In rodents transiently amplifying progenitor cells migrate along the rostral migratory stream (RMS) in a pattern of chain migration. The RMS is ensheathed by a tube-like structure built of slowly dividing radial glial cells, which may also exhibit precursor cell function in the RMS (Gritti et al., 2002). In humans, however, it was suggested that the RMS is organized as a lateral ventricular extension at which neuroblasts migrate toward the OB (Curtis et al., 2007).

Interestingly, adult neurogenesis was shown to be influenced by various different types of conditions. Factors like enriched environment (Kempermann et al., 1997) and physical activity (van Praag et al., 1999; Kronenberg et al., 2006) positively regulate hippocampal neurogenesis. In contrast, aging (Kempermann et al., 1998) and stress (Cameron & Gould, 1994) were shown to act in a negative manner on the generation of new neurons. Additionally, pathophysiological factors like global and focal brain ischemia (Jin et al., 2001), epileptic seizures (Parent et al., 1997) and traumatic brain injury (Rice et al., 2003) stimulate neurogenesis.

The processes involved in neurogenesis were most intensely studied within the brains of rodents such as mice and rats but little is known about adult neurogenesis within our closer relations; primates such as the common marmoset (Callithrix jacchus) (Gould et al., 1998; Leuner et al., 2007; Zhao et al., 2008).

The common marmoset is a new world monkey, which reaches sexual maturity at 1.5 years of age while the onset of senescence was determined at around 8–10 years (Rose et al., 1993; Harada et al., 1999; Berkovitz & Pacy, 2000; Geula et al., 2002). Initial work on neurogenesis in the common marmoset by Gould et al. (1998) showed that new neurons are added to the DG of adult marmosets and that this process could be diminished by stress paradigms. Additionally, Leuner et al. (2007) demonstrated an age-dependent decline of newly generated neurons that are added to the DG. However, no data are yet available on animals that have passed the age of senescence, an age associated with the onset of neurodegenerative diseases such as Alzheimer’s and Parkinson’s disease. Furthermore, a detailed identification of the cells that are involved in the adult neurogenesis in primates is still missing.

Here, we analyze the anatomy of the neurogenic niches and the expression of cell type–specific markers in those niches in the adult common marmoset brain. Moreover, we demonstrate significant differences in the activity of neurogenesis in the adult marmoset brain compared to the adult mouse brain. Finally, we demonstrate an exponential age-dependent decline of neurogenesis into old age that is accompanied by a reduction in the amount of neuroblast cells.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contribution
  9. References
  10. Supporting Information

The anatomic structure of the SVZ, RMS, and OB system does not exhibit a lateral ventricular extension in the common marmoset brain

To first address whether the LV of the common marmoset brain exhibits a lateral ventricular extension in which neuroblasts that are generated within the SVZ migrate as observed in humans (Curtis et al., 2007), we performed autoradiography using radiolabeled DTPA as a tracer. High-resolution digital autoradiographic images revealed that while the tracer was equally distributed within the LV structure no anterior elongation toward the OBs was observed (Fig. 1A,A′). Analysis was performed over the entire width of the ventricular structure by either high-resolution digital autoradiography (Fig. S1 in Supporting Information) or film autoradiography (Fig. S2 in Supporting Information) and did not proof the existence of such structure. Counterstaining of the brain sections with H/E equally proofed no expansion of the LV but rather showed cell clusters along the RMS similar to the chain-like migration pattern observed in rodents (Fig. 1B,C,C′). Additionally, brain sections were stained for the neuroblast marker doublecortin (DCX), which is commonly used to detect migrating cells along the RMS in rodents (Fig. 2A–C′). Indeed, we found DCX-positive cells within this structure of the common marmoset (Fig. 2A–C). Interestingly, when we compared the relative number of DCX-positive cells within the RMS of the marmoset to the mouse RMS we found significantly less DCX-positive neuroblasts in the marmoset RMS (Fig. 2C). We conclude that the neurogenic niche containing SVZ, RMS, and OB in the common marmoset brain is anatomically organized similarly to the rodent rather than to the human brain. However, the level of migrating neuroblasts reaching the OB seems to be reduced compared to the mouse.

image

Figure 1.  The anatomic structure of the subventricular zone, rostral migratory stream (RMS), and olfactory bulb (OB) system does not exhibit a lateral ventricular extension in the common marmoset brain. Radiolabeled DTPA was injected into the lateral ventricle, and autoradiographic analysis was performed on sagittal brain sections (A, autoradiography overlayed on a H/E stained section, A′, autoradiography alone). Radiographs showed equal distribution of the tracer within the ventricle structure but no tube-like extension from the lateral ventricle (LV) toward the OB was observed. H/E staining similarly did not prove an anterior elongation of the LV (B,C), but cell clusters along the RMS (C′). For further analysis, the hippocampal structure was included, shown in H/E staining (C″). Scale bars: 5 mm (A,A′), 5 mm (B), 200 μm (C–C″).

Download figure to PowerPoint

image

Figure 2.  Cell migration in the common marmoset rostral migratory stream (RMS) is associated with DCX+ neuroblasts, but the amount of neuroblasts is notably lower in the marmoset when compared to the mouse RMS (A–D). Cell numbers for doublecortin (DCX) and neuronal nuclei (NeuN)-positive cells within the hippocampal dentate gyrus (DG), subventricular zone (SVZ), RMS, and olfactory bulb (OB) reveal differences in the level of ongoing neurogenesis between the common marmoset and the mouse (E–K). Sagittal brain sections were stained for DCX (green, A–C′) and NeuN (red, A,A′,B), and the expression pattern was analyzed along the RMS. Additionally, DCX- and/or NeuN-positive cells were quantified and cell numbers were expressed as percentage from the total cells per region (D–G′). Total cell numbers within those regions were quantified and expressed as cell number per mm3 (H–K). DCX+ cell populations were detected within the RMS of marmosets (A,A′,B). The white arrowhead in A points to a cell magnified in A′. Figure B illustrates a 3D reconstruction of A. Comparison to the mouse RMS revealed significantly lower numbers of DCX+ cells in the marmoset RMS (C, marmoset; C′, mouse; maximal projections of stack images between 19 (mouse) and 42 μm (marmoset) and D for quantification). Within the OB mice had significantly more NeuN-positive cells when compared to the marmoset (E). No difference in the number of NeuN-positive cells within the DG and SVZ comparing marmoset and mouse was detected (F,G). Cell counts in DG & SVZ revealed significantly less DCX-positive cells in the marmoset brain (F′,G′). Total cell counts demonstrated that while there were less cells in the mouse DG compared to the marmoset (H), the SVZ and RMS were more densely populated in the mouse than in the marmoset (I,J). There was no difference in the total cell number within the OB comparing both species (K). Scale bars: 50 μm (A,B), 10 μm (A′), 30 μm (C,C′). Statistical analysis was performed using t-testing. * 0.06, ** 0.05 and *** 0.001.

Download figure to PowerPoint

The expression patterns of cellular markers for adult neurogenesis are similar in the common marmoset and rodent brain

To verify whether cellular markers commonly used in rodents to analyze different cell stages during the process of neuronal generation also apply for the common marmoset, immunohistochemistry was performed on sagittal brain sections. Robust staining was observed for proteins marking all transient stages of neuronal differentiation during adult neurogenesis in both the SGZ and SVZ-OB system [glial fibrillary acidic protein (GFAP), radial glia-like stem cells; Sox2, neural stem cells; DCX, transiently amplifying progenitor cells/neuroblasts; Tuj1, immature neurons; neuronal nuclei (NeuN), mature neurons] (Fig. 3). The expression patterns were similar to those observed in rodents suggesting that expression profiles during neuronal differentiation in the adult brain are conserved between the species. In the hippocampus, GFAP-positive cells were found in the SGZ as well as in the hilus most likely representing radial glia-like stem cells and astrocytes, respectively. Additionally, GFAP expression was detected in cells lining along the SVZ of the LV similar to GFAP-positive cells in the rodent SVZ (Fig. 3A). In the hippocampus, DCX-positive cells were primarily located in the SGZ extending their dendritic arbor into the GCL. Additionally, DCX-expressing cells were also found within the GCL, most likely representing more mature newly generated cells, which had migrated into the GCL. Moreover, DCX expression was also observed within the SVZ; however, it seemed to appear to a somewhat lesser extent as seen in rodents (Fig. 3B and see also below Fig. S4). As expected, mature granule neurons of the DG were found to be positive for the neuronal marker NeuN. Interestingly, similar to rodents, no NeuN expression was observed in the 2–3 cell layer wide SGZ indicating that this area contains a population of neuronal progenitor cells, which do not express NeuN (Fig. 3C). Likewise, no NeuN expression was found in the SVZ of the LV, the second well-characterized neurogenic niche (data not shown), whereas mature neurons expressed NeuN in the OB (Fig. 3C). Tuj1-positive cells, a marker for young, immature neurons, could be observed in the GCL as well as in the SGZ of the DG. However, the expression was somewhat faint when compared to the rodent hippocampus and only few cells were detected within any given brain section, suggesting that the antibody might not have been suitable for recognizing the marmoset protein (Fig. 3D). Finally, Sox2-expressing cells were found along the SVZ of the LV indicating the existence of neural stem cells in that area (Fig. 3E). However, some proteins could not be detected suggesting that those antibodies did not recognize common marmoset proteins (Table S1 in Supporting Information). Additionally, immunostaining for DCX and NeuN was analyzed in the OB (Fig. S3 in Supporting Information). As expected not only NeuN but also DCX was detected, suggesting that DCX-positive cells migrate into the OB (Fig. S3A,B). Moreover, also NeuN/DCX-double-positive cells could be observed (Fig. S3C) indicating that DCX-positive cells in the OB are in a transitional stage to become mature neurons.

image

Figure 3.  The expression pattern of cellular markers for adult neurogenesis in the common marmoset resembles that seen in rodents. Sagittal brain sections were stained for known markers of different cell stages during neuronal differentiation, and expression patterns were analyzed in hippocampal dentate gyrus (DG), subventricular zone and olfactory bulb (OB). Glial fibrillary acidic protein (A), doublecortin (B), neuronal nuclei (C), Tuj1 (D) and Sox2 (E). Upper panel (DG) and lower panel (lateral ventricle) in (A,B) show maximal projections of stack images between 11 and 24 μm. Scale bars: 50 μm (A–E), 10 μm (C; lower panel, OB).

Download figure to PowerPoint

Proliferation takes place in both neurogenic niches of the adult brain

To further analyze whether cell proliferation takes place in the neurogenic niches, brain sections were stained for Ki67, a marker for cells in active cell cycle. Ki67-positive cells could be observed in SVZ and SGZ suggesting that proliferation indeed occurs in the marmoset brain (Fig. 4). Additionally, animals received 5-bromo-2-deoxyuridine (BrdU), a thymidine analog that is incorporated into DNA of diving cells, 3 h prior to sacrifice, and double-labeling of BrdU and DCX was performed on brain sections (Fig. 5A,B). As expected, three hours following BrdU application, BrdU-incorporated cells were also positive for DCX suggesting that neurogenesis in the common marmoset was associated with proliferation of DCX-expressing transiently amplifying progenitor cells in both neurogenic regions. Moreover, DCX-positive cells in the OB were also found to express Ki67 clearly indicating that neuronal progenitors proliferate in the OB (Fig. 5C).

image

Figure 4.  Proliferation takes place in both neurogenic niches the subventricular zone (SVZ) and the subgranular zone (SGZ). Sagittal brain sections were stained for Ki67. Ki67-expressing cells were found in the SVZ (A, upper panel) and SGZ (A, lower panel). Higher magnification images of (A) are shown in (B). Scale bars: 50 μm (A), 20 μm (B).

Download figure to PowerPoint

image

Figure 5.  Proliferation in subventricular zone (SVZ), subgranular zone (SGZ), and olfactory bulb (OB) is associated with doublecortin (DCX)-positive, transient amplifying progenitor cells. Animals received 5-bromo-2-deoxyuridine (BrdU) 3 h prior to sacrifice, and sagittal brain sections were stained for DCX (green) and BrdU or Ki67 (red). BrdU-incorporated cells were found to express DCX in the SGZ (A) and SVZ (B). Cells within active cell cycle as shown by Ki67 staining in the OB also expressed DCX. Scale bars: 20 μm (A,B), 10 μm (C).

Download figure to PowerPoint

Comparison of DCX cell numbers in the neurogenic regions of the common marmoset and the mouse reveals differences in the level of ongoing neurogenesis

To further analyze whether the relative abundance of ongoing neurogenesis in the common marmoset is comparable to other species such as mice, we quantified the total cell number as well as the amount of DCX- and/or NeuN-positive cells within the DG, SVZ, RMS, and OB of marmoset and mouse brains (Fig. 2D–K). Interestingly, we detected that while NeuN cell numbers in the DG and SVZ were similar between marmoset and mouse (Fig. 2F–G), the same regions had significantly less DCX-positive cells in the marmoset brain (Fig. 2F′–G′, DG: marmoset 35.9% ± 6.4, mouse 177.4% ± 28.2, SVZ: marmoset 50.9%± 18.9, mouse 604.2% ± 163.5,  0.05) suggesting that neurogenesis occurs to a somewhat lower level compared to mice. Further analysis in the RMS revealed that also in this area DCX-positive cells were sparse in the marmoset compared to the mouse RMS (Fig. 2D, RMS: marmoset 39.6% ± 9.1, mouse 823.8% ± 181.6,  0.001). Furthermore, NeuN-positive cell numbers in the OB were significantly higher in the mouse rather than in the marmoset indicating that in marmosets a higher amount of cells other than neurons reside in the OB (Fig. 2E, OB: marmoset 76.3% ± 3.4, mouse 90.9% ± 2.0,  0.05). Additionally, the total cell number was quantified in the described regions and compared between marmoset and mouse (Fig. 2H–K). We observed a significant higher level of cells within the marmoset DG (Fig. 2H, DG: marmoset 1844.0 ± 110.9 cells, mouse 1526.4 ± 73.3 cells,  0.05), while cell numbers were lower in SVZ and RMS when compared to the same structures in mice (Fig. 2I,J, SVZ: marmoset 1286.5 ± 229.5 cells, mouse 2217.2 ± 360.2 cell,  0.05, RMS: marmoset 1102.4 ± 242.8 cells, mouse 2417.9 ± 562.9 cells,  0.06). No differences were found within the OB between marmoset and mouse (Fig. 2K). These results demonstrate that the number of neuroblasts produced in the mouse neurogenic niches is substantially higher when compared to the common marmoset.

However, even though we analyzed animals of comparable state of maturity analyzing two different species might be limited for example by the fact that in rodents the level of neurogenesis seems to decrease with aging of the animals Kuhn et al. (1996). Therefore, we were interested to see whether this age-dependent decline in neurogenesis could also be observed in the marmoset brain.

Age-dependent decrease in neurogenesis is associated with the loss of DCX-expressing cells in the SGZ

A study by Leuner et al. (2007) demonstrated in marmosets that with aging the number of BrdU/Tuj1-positive cells decreased linearly when young and middle-aged animals were analyzed. Here, we set out to verify whether this decline in neuronal generation further continues into old age when senescence has already commenced using the progenitor cell marker DCX. Animals were grouped according to their age into a young group (under 3 years), a middle-aged group (4–7 years) and an old group (8–15 years). Brain sections were stained for DCX, and positive cells were counted within the DG (Fig. 6). As shown in Fig. 6A,A′, young animals contained numerous DCX-positive cells within the SGZ of the hippocampal DG (Fig. 6A,A′, upper panel) while DCX-positive cells in old animals were sparse (Fig. 6A,A′, lower panel). Quantification revealed that there was a significant decrease in DCX-positive cell numbers comparing the three age groups (Fig. 6B, young vs. middle-aged approximately 64% decline with = 0.002, young vs. old approximately 90% decline with = 0.001 and middle-aged vs. old approximately 70% decline with = 0.005) suggesting that the age-dependent decline of neurogenesis in the hippocampus continues into old age and is associated with a decrease in DCX-positive transiently amplifying progenitors cells. Additionally, we set out to analyze whether a correlation between the number of DCX-positive cells and the age of the animals was of exponential or linear nature as suggested by Leuner et al. (2007). Interestingly, while there was a strong negative exponential correlation (Fig. 6C, y = 75.884e−0.004x, r2 = −0.8) only a much weaker correlation was found using a linear equation (Fig. S4, y = −0.0903x + 57.634, r2 = −0.5) suggesting that the decrease in neurogenesis occurs exponentially over time in the common marmoset brain.

image

Figure 6.  Neurogenesis decreases exponentially with age and is associated with the loss of doublecortin (DCX)-expressing cells in the subgranular zone (SGZ). Sagittal brain sections from animals of different age groups were stained for DCX, and the expression was analyzed in the SGZ. The SGZ of young animals was densely packed with DCX-expressing cells (A,A′, upper panel). Old animals had only few DCX-positive cells within the SGZ of the hippocampal dentate gyrus (A,A′, lower panel). Higher magnification images of (A) are shown in (A′). Quantification revealed a significant reduction in the number of DCX-positive cells between the age groups (B; young vs. middle-aged approximately 64% decline, = 0.002; young vs. old approximately 90% decline, = 0.001; middle-aged vs. old approximately 70% decline, = 0.005). Correlation analysis between the number of DCX-positive cells and age of the animals demonstrated a strong exponential correlation (r2 = −0.8) C. Statistical analysis was performed using Mann–Whitney U ranking followed by Bonferroni correction. *< 0.01. Scale bars: 30 μm (A), 10 μm (B).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contribution
  9. References
  10. Supporting Information

The SVZ, RMS, and OB system have been intensely studied within the brains of rodents such as mice and rats and it is well established that newly generated neuroblast migrate from the SVZ into the OB via chain-like migration along the RMS (for details see review Zhao et al. 2008). However, little is known about these processes within higher species. Interestingly, it was suggested by Curtis et al. (2007) that the lateral ventricles of humans’ extent a tube-like structure into the OB filled with cerebrospinal fluid, which was termed the ventriculo-olfactory neurogenic system (VONS). Similar observations have also been made in other species such as sheep (Rae, 1994). However, others did not find a continuous tube-like structure extending from the LV into the OB within the adult human brain but rather demonstrated that this human olfactory ventricle closes during fetal development (Sanai et al., 2007). In favor of the latter, we similarly did not observe a tube-like extension of the LV within the adult common marmoset brain using radioactive labeling and H/E staining. However, by using BrdU/DCX or Ki67/DCX co-staining methods, our data provide clear evidence that neurogenesis takes place not only in the hippocampal DG but also in the SVZ of the LV and OB system in marmosets suggesting that neuroblasts must have reached the OB via migration from the lateral ventricle. In agreement with this we could demonstrate that DCX-positive neuroblasts are located within the RMS of the common marmoset brain. However, not only DCX-positive cells but also the total cell number appears to be significantly smaller in the marmoset RMS compared to the same structure in the mouse suggesting that neurogenesis and cell migration occurs at a substantially lower level. Studies in macaque monkeys and humans have suggested that local proliferation of resident precursor cells in the OB could contribute to the addition of new neurons to the olfactory system (Kornack & Rakic, 2001; Bedard & Parent, 2004). Indeed, we found DCX-positive cells expressing the active cell cycle marker Ki67 indicating that proliferation of neuronal progenitors does occur within the OB. However, we observed only up to ten Ki67/DCX-positive cells per OB section but numerous DCX-positive, Ki67-negative cells (data not shown) and it therefore seems unlikely that only local proliferation would account for the generation of new neurons within the OB of the common marmoset. The facts that we did not observe a ventricular extension of the lateral ventricles but found proliferation within the SVZ, cell clusters of DCX-positive neuroblasts along the RMS and numerous neuroblasts in the OB all point to a similar structural and molecular organization of this neurogenic niche as shown for rodents.

In support, for this our marker gene expression analysis revealed that not only the same proteins were expressed marking the different stages of progenitor cells toward a neuronal phenotype but also that the expression patterns were similar to those in rodents clearly indicating that the processes during adult neurogenesis seem to be conserved between species. Being able to apply those antibodies used in the study for immunohistochemical analysis of adult neurogenesis will be of great importance to study the involved processes in more detail. This will allow to further evaluate similarities and/or differences in proliferation, migration, differentiation, and integration of newly generated neurons between rodents and new world monkeys and may also contribute to a better understanding of adult neurogenesis within humans. Nevertheless, some of the antibodies failed to detect the common marmoset proteins raising the question whether these proteins were not expressed or whether the antibodies were unable to recognize the marmoset protein. However, two of those undetected proteins were Mash1 and Olig2, both well characterized proneural bHLH transcription factors (Guillemot, 1999). For example, Mash1 was shown to induce neuronal differentiation in the embryonic and postnatal brain seems to be expressed in transient amplifying cells of the SVZ of adult rodents (Parras et al., 2004) and its deficiency was also found to result in loss of neuronal progenitors (Casarosa et al., 1999). Additionally, mammalian Mash1 was described as the homolog of the Drosophila achaete-scute gene and was shown to be conserved between invertebrates and vertebrates (Bertrand et al., 2002). Therefore, it seems rather unlikely that those transcription factors would not be expressed in the common marmoset brain but rather that they were not recognized by our antibodies indicating limitations in the work with marmoset brains. In other cases, such as for Tuj1, the expression was weaker as it would have been expected from expression profiles in rodent brains (Zhao et al., 2008). However, a study by Leuner et al. (2007) on adult neurogenesis in the common marmoset brain was based on expression of Tuj1 using the exact same antibody pointing out that Tuj1 indeed is expressed in new born neurons in common marmosets.

Interestingly, quantitative analysis of cells expressing DCX or NeuN within the neurogenic regions of the brain revealed substantial differences between rodents such as mice and the new world monkey. Within both the DG and the SVZ but even more prominent in the RMS of the marmoset significantly less DCX-positive cells were found while NeuN cell numbers were not different in the DG and SVZ. This raises the question whether ongoing adult neurogenesis occurs at even lower levels in higher species such as humans and further demonstrates potential limitations for the attempt of using endogenous neural stem cells in replacement therapies following injury to the human brain.

Additionally, we observed significantly less NeuN-positive cells within the marmoset OB when compared to mice. This is in agreement with studies demonstrating that the relative number of glia cells increases within evolutionary higher organized species (see review Klambt, 2009) and thus it was expected that more glia cells reside in the OB at the relative expense of neuronal cell numbers.

Finally, we observed an age-dependent decline of hippocampal neurogenesis in the common marmoset brain as previously reported (Leuner et al., 2007). Importantly, we extended our analysis using animals that had reached an age of senescence clearly showing that the reduction in new neurons continues into old age. Senescence has been shown to be associated with the onset of diseases to the brain such as Alzheimer’s and Parkinson’s disease and thus using animals of this age can more closely predict the neurogenic status of the human brain at an age relevant for the onset of neurodegeneration. Labeling endogenous DCX expression rather than BrdU incorporation for this analysis had several advantages. Since BrdU is thought to be incorporated only within a two hour window following administration BrdU-labeling usually underestimates the number of cells being in active cell cycle and thus does not provide conclusive evidence of the actual proliferative status within a brain. Furthermore it was suggested that BrdU can also be incorporated into dying cells because of DNA repair mechanisms, although others have argued against this possibility (Palmer et al., 2000; Cooper-Kuhn & Kuhn, 2002; Rakic, 2002; Taupin, 2007). Lastly, BrdU might be diluted to an undetectable level because of continuing cell divisions and thus newly generated cells would be underestimated depending on the proliferative potential of the cells (Hayes & Nowakowski, 2002). Therefore, as an endogenous marker DCX can visualize more accurately the number of cells that are at a stage to become mature neurons at any given time. In the hippocampus, DCX expression commences within the type IIb stage of transiently amplifying progenitor cells, remains in the stage of type III progenitor cells which will become postmitotic neurons and DCX is still transiently expressed when NeuN expression in the immature postmitotic neurons has already started (Kempermann et al., 2004). Thus, DCX expression encompasses not only one selective cell type but a wide spectrum of neuronal progenitor cells and young neurons. In their study, Leuner et al. (2007) demonstrated a decline of BrdU+/Tuj1+ immature neurons 3 weeks following BrdU application comparing young and middle-aged marmosets. However, they could not rule out whether this effect on neurogenesis was a result of decreased proliferation of neural progenitor cells or rather impaired survival of the newly generated neurons within the older animals. It has been shown that during adult neurogenesis a surplus of cells is generated of which a substantial number undergoes apoptotic cells death (Dayer et al., 2003) probably because of a lack of appropriate synaptic activity. Therefore, in our study by using DCX as an endogenous marker, it could clearly be shown that age-dependent decreased neurogenesis was associated with a decline of transiently amplifying progenitor cells in the common marmoset brain as has been demonstrated within rodents previously (Rao et al., 2006). Additionally, correlational analysis between the age of the animals and the decrease in DCX-expressing cells within the hippocampal GCL best fitted a negative exponential curve similarly to what has been reported in rodents (Ben Abdallah et al., 2010). This is in disagreement with the study by Leuner et al. (2007) who showed a linear decline of neurogenesis over a lifespan of 7 years within the common marmoset brain. However, this discrepancy might be explained by the fact that while Leuner et al. (2007) correlated only young and middle-aged animals, while we also included a third group of old marmosets that had already reached the age of senescence indicating that with aging the decay of neurogenesis is less pronounced than compared to younger animals.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contribution
  9. References
  10. Supporting Information

Animals

Common marmosets.  Marmoset monkeys (Callithrix jacchus) were raised in the institutional breeding facility of the Centre of Reproductive Medicine and Andrology (Prof. Dr. Stefan Schlatt). Breeding, maintenance, and experimental procedures were performed in accordance with the German Federal law on the Care and Use of laboratory animals. A license for the marmoset breeding colony was obtained from local authorities (Veterinär- und Lebensmittelüberwachungsamt der Stadt Münster). The license to sacrifice marmosets as tissue donors for scientific experiments was granted by the Landesamt für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen (8.87–50.10.46.09.018). Animals were obtained from local colonies at the Institute of Reproductive Medicine, University of Münster. Three hours prior to sacrifice, animals received a single dose of 100 mg/kg BrdU, were then deeply sedated using ketamine, and killed by exsanguination. Afterward, brains were obtained and processed for immunohistochemistry. For this work, the following animals were used: 1 male (age: 144 weeks), 8 females [age: 2 × 134 weeks (young group), 209, 227, and 352 weeks (middle-aged group), 423, 490, and 727 weeks (old group)].

C57BL/6 mice.  C57BL/6 mice were raised in the institutional breeding facility of the Center for Molecular Biology of Inflammation. Breeding, maintenance, and experimental procedures were performed in accordance with the German Federal law on the Care and Use of laboratory animals. For this study, three mice at an age of 10 weeks were used. Animals were sacrificed by anesthesia and transcardial perfusion with 50 mL PBS followed by 50 mL 4% paraformaldehyde (PFA). Each solution was injected into the heart with an injection rate of 4 mL/min. Brains were removed and processed for immunohistochemistry.

Autoradiographic analysis

For autoradiographic analysis of the lateral ventricle structure, animals were sacrificed, brains were removed and kept in Dulbecco’s modified Eagle’s medium (DMEM) until further processing. Fifty microliter radiolabeled diethylenetriaminpentaacetat (99mTc-DTPA, 50 MBq) was injected into the lateral ventricle. Upon injection, the brains were kept on ice for 15 min to allow the tracer to distribute equally within the ventricle structure. Brains were then snap-frozen in liquid nitrogen, and sagittal sections were obtained of either 10 or 40 μm thickness using a cryotome (Leica). Autoradiography was performed in selected sections for 8 h using a high-resolution digital autoradiography system (Micro Imager; Biospace Lab, Orléans, France). Remaining sections were measured using film autoradiography. For anatomical mapping, sections were stained using hematoxylin/eosin (H&E) staining, digitized with Nikon Eclipse 90i, NIS Image 3 (Nikon GmbH, Düsseldorf, Germany) and coregistered with digital autoradiographs (Matlab 7.9; Mathworks Inc., Natick, MA, USA).

Histological procedures

Hematoxylin/eosin staining.  For H/E labeling, brain sections were fixed for 5 min in 4% PBS-buffered PFA followed by rinsing in dH2O. Hematoxylin was applied for 10 min, sections were rinsed for 10 min under tap water followed by dH2O and 1% eosin solution for 7 min. Ascending alcohol steps were the following: 80%, 96%, 100%, and 100% isopropanol for 5 min and xylene for 10 min. Brain sections were mounted using Entellan.

Immunohistochemistry.  For immunohistochemical analysis, brains were dissected and postfixed over night with 4% PFA in PBS. Sagittal sections (40 μm) were prepared with a Vibratome (Leica, Wetzlar, Germany). Sections were blocked with blocking buffer (100 mm Tris buffer, 0.5% Triton X-100, 0.1% sodium azide, 0.1% sodium citrate and 5% normal goat serum). Blocking was followed by incubation with primary (Table S1) and secondary antibodies (Alexa-fluorophore conjugated antibodies; Invitrogen), both diluted in the blocking solution. Images were collected by confocal microscopy using cen software (Zeiss, Jena, Germany); image analysis was performed with the cen software, Adobe Photoshop and the image j software (NIH, Bethesda, MD, USA).

Quantification of total cell numbers and the level of DCX- or NeuN-positive cells

To compare total cell numbers and the relative abundance of ongoing neurogenesis within different regions of the adult brain between the common marmoset and the mouse confocal stack images stained for DCX and/or NeuN were analyzed. The selected regions included the hippocampal DG, SVZ, RMS, and the OB. First, the total number of cells per stack and region was counted by Hoechst staining, and numbers were expressed as cells/mm3. Additionally, DCX- and/or NeuN-positive cells were counted, and numbers were expressed as percent to the total cell number. For each region and marker, 3–7 confocal stack images were analyzed. Images were taken from three individual animals all of which were aged 2.8 years or 10 weeks for marmosets and mice, respectively. Statistical analysis was performed using t-testing for two independent groups, and data were considered significant when * 0.06, ** 0.05 and *** 0.001.

Cell counts and statistical analysis

Cell counts of DCX-positive cells within the hippocampus were performed throughout the entire DG including SGZ and GCL and are expressed as cell number per 40 μm brain section. Cells located more than one cell width away from the SGZ in the hilus were excluded from counting. For each animal, 3–12 comparable brain sections were used. For selection of brain sections, equivalent stereotaxic coordinates were used for each animal. This was ensured by automatic vibratome sectioning (Leica Vibratome 1200S, Leica, Wetzlar, Germany). Brain hemispheres were sectioned from lateral to medial with lateral being the starting point of 0 μm. Forty micrometer sections were cut repeatedly until the desired depth was reached. For each section, cells within the entire 40 μm using a 1 μm optical slice were counted using a 63× oil objective from Zeiss. Statistical analysis was performed using the Mann–Whitney U rank test followed by Bonferroni correction, and a P-value of = 0.01 was considered significant. For the analysis of linear or exponential correlation between cell numbers and age of the animals, trendlines were fitted to the graphs and R2-values were calculated. All analyses were undertaken using the sigmaplot software (Systat Software Inc., Erkrath, Germany).

Acknowledgments

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contribution
  9. References
  10. Supporting Information

The authors thank Irmgard Hoppe and Sandra Schröer for excellent technical assistance. This work was supported by research grants from the German Research Foundation (Emmy Noether Program and SFB629), Competence network stem cell research NRW, EU FP7 network EuroSyStem, Interdisciplinary Center of Clinical Research (IZKF; Münster; Germany; core unit SmAP) and the Medical Faculty of the University Münster.

Author contribution

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contribution
  9. References
  10. Supporting Information

E.C.B., conception and design, collection and/or assembly of data, data analysis and interpretation, manuscript writing, final approval of manuscript; S.S., collection of data, final approval of manuscript; S.H., M.S. and J.C.S., conception and design, data analysis and interpretation, final approval of manuscript; S.S., conception and design, final approval of manuscript.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contribution
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contribution
  9. References
  10. Supporting Information

Fig. S1 Autoradiographic analysis revealed no lateral ventricular extension in more medial parts of the common marmoset brain. Injection of radiolabeled DTPA into the lateral ventricle was analyzed in sagittal brain sections derived from parts further medial and revealed no ventricular extension. High resolution digital autoradiography illustrates distribution of the tracer within the LV as well as in the injection tract (A, autoradiography overlayed on a H/E stained section, A′, autoradiography alone). The arrowhead points at the location where a tube-like ventricular extension would have been expected. Overlay with H/E staining of the same section demonstrates the locations of surrounding brain structures. Scale bars: 5 mm (A,A′).

Fig. S2 Serial brain sections show no ventricular extension by film autoradiography analysis. Radiolabeled albumine was injection into the lateral ventricle of two marmoset brains and serial sagittal brain sections were analyzed. In both animals radioactivity could be detected within the LV but no distribution of the tracer into a potential ventricular extension was observed. The arrowheads point at the locations where a tube-like ventricular extension would have been expected (A, animal 1; B, animal 2). Cx, cortex; Cb, cerebellum; h, hippocampus; LV, lateral ventricle. Scale bars: 1 cm (A,B).

Fig. S3 NeuN/DCX-double-positive cells can be found in the olfactory bulb. Immunostaining was performed for DCX (A) and NeuN (B) on OB sections. CO-staining showed DCX-positive cells (green) that already started to express NeuN (red) (C, arrowheads). Scale bars: 10 µm (A–C).

Fig. S4 Decline of neurogenesis does not correlate linearly with aging. DCX-positive cell numbers within the DG were correlated to the age of the animals. A linear trend line was fitted to the graph. R2-value calculation revealed that there was only a weak linear correlation (r2 = −0.5).

Table S1 Table of antibodies tested on common marmoset brain sections

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
ACEL_639_sm_FigureS1.pdf824KSupporting info item
ACEL_639_sm_FigureS2.pdf450KSupporting info item
ACEL_639_sm_FigureS3.pdf402KSupporting info item
ACEL_639_sm_FigureS4.pdf274KSupporting info item
ACEL_639_sm_Legend.doc24KSupporting info item
ACEL_639_sm_TableS1.pdf216KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.