The generation, migration, and differentiation of olfactory neurons in the adult primate brain

Abstract

In adult rodents, neural progenitor cells in the subependymal (SZ) zone of the lateral cerebral ventricle generate neuroblasts that migrate in chains via the rostral migratory stream (RMS) into the olfactory bulb (OB), where they differentiate into interneurons. However, the existence of this neurogenic migratory system in other mammals has remained unknown. Here, we report the presence of a homologue of the rodent SZ/RMS in the adult macaque monkey, a nonhuman Old World primate with a relatively smaller OB. Our results—obtained by using combined immunohistochemical detection of a marker for DNA replication (5-bromodeoxyuridine) and several cell type-specific markers—indicate that dividing cells in the adult monkey SZ generate neuroblasts that undergo restricted chain migration over an extended distance of more than 2 cm to the OB and differentiate into granule interneurons. These findings in a nonhuman primate extend and support the use of the SZ/RMS as a model system for studying neural regenerative mechanisms in the human brain.

The generation of mammalian brain structures is restricted to developmental periods (1). However, investigations of neurogenesis in the rodent brain revealed two forebrain regions that continue producing new neurons well into adulthood: the hippocampal dentate gyrus (2–4) and the subventricular, or subependymal, zone (SZ) (5–7). Of these two regions, the SZ harbors the largest pool of dividing neuronal progenitor cells in the adult rodent brain (8, 9). Progenitor cells in the SZ generate immature neurons that aggregate to form an extensive network of neuroblast chains along the lateral wall of the lateral cerebral ventricle (10, 11). These chains of neuroblasts coalesce anteriorly to form a highly restricted migratory route, designated as the rostral migratory stream (RMS), which extends from the anterior SZ into the olfactory bulb (OB). Unlike the radial glial-guided migration used by young neurons during early brain development (12), neuroblasts undergoing “chain migration” in the adult SZ/RMS migrate along one another via neurophilic interactions (11, 13). These chains are ensheathed by tubes of flanking astrocytes, which delineate the RMS. Neuroblasts migrate rostrally within the RMS to enter the OB, whereupon they differentiate into local interneurons (6, 7, 10, 14).

However, although the SZ/RMS neurogenic migratory system often is presumed to be a generic feature of all adult mammalian brains (including humans), in fact, its existence has remained documented only in murine rodents. In these rodents, which have relatively large OBs and are predominantly nocturnal, this system is thought to play a role in odor discrimination (15). Consequently, the question has remained open of whether this olfactory stream is peculiar to macrosmatic rodents or also exists in microsmatic nonrodent species, particularly anthropoid primates (monkeys, apes, and humans). Compared with rodents, anthropoids have relatively small OBs with elongated olfactory peduncles and are largely diurnal (16). Moreover, both adult macaque monkeys and humans appear to generate fewer new neurons in the hippocampal dentate gyrus than do adult mice (17–19), which perhaps might reflect an overall decline in adult neurogenesis in the Old World primate brain that would include the SZ. Therefore, to begin assessing the potential of the SZ/RMS system as a model for developing neuronal replacement strategies for brain repair in humans (20, 21), it is imperative to determine whether this system exists in a primate forebrain with an olfactory system more similar to our own. Here, we have tested whether a neurogenic migratory homologue of the rodent SZ/RMS system is present in adult macaque monkeys—a nonhuman Old World anthropoid.

Materials and Methods

5-Bromodeoxyuridine (BrdU) Injections.

All animal care and experimentation were conducted in accordance with institutional guidelines. Eleven adult (5–16 years of age) macaque monkeys, both rhesus (Macaca mulatta) and cynomolgus (Macaca fascicularis) of both sexes, received i.v. injections of BrdU (Sigma) dissolved in 0.9% NaCl with 0.007 M NaOH. BrdU is a thymidine analogue that is incorporated into replicating DNA of dividing cells and thereby can be detected in these cells or their subsequent progeny by using immunohistochemistry (22, 23). Injections were given between 9 a.m. and noon. Three monkeys received a single BrdU injection and were killed 2 h (one monkey) or 27 days postinjection. The other eight monkeys received one daily injection for 5 consecutive days and were killed at 2 h or after 4, 12, 31, 32, 72, 75, or 97 days postinjection. The monkeys with postinjection survival times of 75 and 97 days received a dose of 75 mg/kg body weight per injection; all other monkeys received a dose of 50 mg/kg per injection. We confined our analysis to the region surrounding the anterior horn of the lateral cerebral ventricle, the olfactory peduncles, and the OB, which are the brain regions that would be possibly homologous to the anterior SZ and RMS described in the rodent.

Immunohistochemistry.

For immunoperoxidase staining, animals were anesthetized and perfused with 70% ethanol, and brains were blocked and postfixed overnight at 4°C. Blocks were dehydrated in graded alcohol solutions, cleared in xylene, embedded in paraffin, and serially sectioned at either 8 or 10 μm in the coronal plane. Sections were mounted on glass slides and processed as follows. For proliferating cell nuclear antigen (PCNA) immunoperoxidase staining, rehydrated sections were incubated in blocking serum and then incubated overnight at 4°C with a mouse anti-PCNA antibody (1:500; Roche Molecular Biochemicals). BrdU immunoperoxidase staining was performed as described (23). Briefly, rehydrated sections were placed in 2 M HCl for 1 h, rinsed, and incubated with a mouse anti-BrdU antibody (1:100; Becton Dickinson) for 30 min at room temperature. Either antibody was visualized by using a biotinylated horse anti-mouse IgG (1:200; Vector Laboratories), the Vector ABC Elite kit, and H₂O₂/diaminobenzidine with 0.02% cobalt chloride and 0.02% nickel ammonium sulfate to yield a black reaction product. Sections were counterstained by using 0.1% basic fuchsin.

For immunofluorescence triple-labeling for BrdU and cell-type markers, we perfused with 0.9% saline, followed by 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer (PB), pH 7.4. Blocks of brain tissue were postfixed in PFA-PB for 6 h at 4°C and then sunk in graded sucrose solutions to 30%. Blocks were frozen, and coronal cryostat sections (30 or 40 μm) were placed in PBS and immediately processed. To detect BrdU, the free-floating sections were pretreated to denature DNA by the following steps: 2 h incubation in 50% formamide/2× SSC [standard saline citrate (1× SSC = 0.15 M sodium chloride/0.015 M sodium citrate, pH 7] at 65°C; 5-min rinse in 2× SSC; 30-min incubation in 2 M HCl at 37°C, and 10-min rinse in 0.1 M boric acid, pH 8.5. Sections then were blocked and incubated for 48 h at 4°C with a pooled solution of rat anti-BrdU antibody (1:100; Accurate Scientific, Westbury, NY), rabbit anti-glial fibrillary acidic protein (GFAP, 1:1,000; Dako) and either mouse anti-NeuN antibody (1:125; Chemicon) or mouse anti-class III β-tubulin antibody (TuJ1, 1:400; a kind gift from A. Frankfurter, Univ. of Virginia, Charlottesville). After rinsing in blocking buffer, sections were incubated for 2 h in a pooled solution of Alexa 488-conjugated goat anti-rat IgG, Alexa 568-conjugated goat anti-mouse IgG (1:500; both from Molecular Probes) and Cy5-conjugated goat anti-rabbit IgG (1:200; Jackson ImmunoResearch).

For polysialylated neural cell adhesion molecule (PSA-NCAM) and TuJ1 double-immunostaining, sections were treated as described above, except that the DNA-denaturing pretreatment was omitted. Sections were incubated in a pooled solution of anti-meningococcus group B (MenB) mAb (1:1,000; a kind gift from G. Rougon, Faculté des Sciences de Luminy, Marseille, France) and TuJ1 antibody. Sections then were rinsed in blocking buffer and incubated in a pooled solution of Cy5-conjugated anti-mouse IgM (1:200; Jackson ImmunoResearch) and Alexa 488-conjugated goat anti-mouse IgG (1:500; Molecular Probes). Sections were mounted and coverslipped with mounting medium. Omitting primary antibodies from the immunohistochemistry processing steps eliminated fluorescence or peroxidase labeling of cells.

Bright-field images were obtained by using a charge-coupled device camera mounted on a Zeiss microscope. Fluorescent signals were imaged by using a confocal laser-scanning microscope (Zeiss LSM 510). For obtaining confocal micrographs, each fluorochrome dye within the same field was scanned separately by using a quasi-simultaneous mode; this eliminates the possibility of signal “bleeding,” which could generate false-positive results. To ensure that a cell that appeared to be double-labeled for BrdU and neuronal marker was not, instead, two closely apposed, single-labeled cells (24), we performed a z-stack analysis, examining optical sections less than 0.8 μm in the z axis. The emission signals of Alexa-488, Alexa-568, and Cy5 were assigned to the green, red, and blue channels, respectively, except in Fig. 3 e-g, in which Alexa-488 and Cy5 signals were assigned to the red and green channels, respectively.

Figure 3.

The rostral migratory stream in the adult macaque monkey olfactory peduncle (OP). (a) A sagittal plane reveals that BrdU-labeled nuclei (arrows, green) are aligned in a strict linear pattern within the core of the peduncle. GFAP immunoreactivity (blue) in the OP is especially pronounced along this pathway. These BrdU-labeled nuclei are immunonegative for the mature neuronal marker, NeuN (red), unlike the neurons seen in the orbitofrontal cortex (OFC) and the few scattered neurons at the periphery of the peduncle that belong to the anterior olfactory nucleus. (b–d) Images in same x, y, and z registration show chains of neuroblasts in the rostral migratory stream. Within the peduncle, chains of intensely stained TuJ1-positive cells (red) are restricted to the core and are aligned parallel to the longitudinal axis. (c) Some cells within these chains have elongated BrdU-positive nuclei (arrows, green) oriented in the presumed axis of migration. (d) These chains are immediately surrounded by a sheath of GFAP-positive fibers (blue) but are GFAP-negative. (e–g) The TuJ1-positive chains in the peduncle (e, red) also coexpress PSA-NCAM (f, green) as indicated by yellow fluorescence when the two signals are merged (g). [Bars = 100 μm (a) and 25 μm (b–g).]

Results

Neural Progenitor Cells.

In monkeys that survived 2 h after either a single or multiple injections of BrdU, we detected BrdU-labeled cells in SZ, olfactory peduncle, and OB, indicating proliferative activity. In coronal tissue sections of the SZ, BrdU-positive cells were distributed either singly or in nested clusters, predominantly along the lateral margin of the anterior horn of the lateral ventricle, between the ependymal lining and the parenchyma of the head of the caudate nucleus (Fig. 1 a and b). The nests were most concentrated in the ventrolateral SZ. In contrast, relatively fewer BrdU-positive cells were observed in the caudatopallial angle or in the SZ on the medial, or “septal,” margin of the lateral ventricle. In sagittal sections of the olfactory peduncle, elongated BrdU-positive nuclei were distributed in small linear groupings, oriented parallel to the tract. In the OB, BrdU-labeled cells were scattered primarily in the granule cell layer and white matter of the core and less frequently in the external plexiform layer and glomerular layer. The presence of occasional BrdU-labeled mitotic figures in the OB (Fig. 1 h and i) provided independent confirmation for resident proliferative activity within the bulb and indicated that BrdU immunoreactivity reflected cell division rather than DNA repair.

Figure 1.

Dividing and newly generated cells in the SZ and OB of adult macaque monkeys as detected by immunoperoxidase staining for BrdU or PCNA. Examples of clusters of BrdU-labeled cells 2 h after the last of five daily injections (a and b) and of PCNA-positive cells (c and d) in coronal sections of the anterior SZ along the striatal wall of the lateral ventricle (LV). (a) Although occasional immunopositive cells initially appeared to be part of the ependymal layer (E), further examination revealed them to be immediately subjacent to it and, thus, to belong to the SZ; no evidence for cell proliferation in the ependymal layer was observed. (e and f) PCNA-positive cells (arrows) among a stream of cells in the SZ closely associated with the basal aspect of the anterolateral ventricle. C, caudate nucleus; S, septal nuclei. Upper left area is dorsal. (f) The boxed field in e at higher magnification, showing that PCNA-positive cells are in the SZ but not the ependyma. (g) A sagittal section of the OB shows “strings” of elongated, PCNA-positive nuclei (arrows) in the white matter (WM) as it enters the core of the OB. GL, granule cell layer. (h and i) BrdU-labeled mitotic figures in the OB 2 h after five daily injections. (h) An early anaphase cell in the glomerular layer. (i) A late anaphase/early telophase cell in the external plexiform layer. [Bars = 20 μm (a–d and f), 100 μm (e), 50 μm (g), and 10 μm (h and i).]

To verify the presence of cell proliferation in the macaque SZ and olfactory peduncle and OB, we examined tissue sections from the same brains that were immunolabeled with antibodies against PCNA, an endogenous marker expressed in the nuclei of cells engaged in the cell cycle (25). The distribution of PCNA-immunoreactive cells was similar to that of the BrdU-labeled cells (Fig. 1 c–g). Nests of PCNA-positive cells were distributed in the SZ along the anterolateral and anteroventral margin of the lateral ventricle (Fig. 1 e and f). Small strings of elongated PCNA-positive nuclei extended from the anteroventral SZ, along the length of the olfactory peduncle, and into the core white matter of the OB (Fig. 1g). Single PCNA-positive cells were scattered throughout the aforementioned layers of the OB. There was no evidence of ependymal cells lining the ventricle that were immunoreactive for either BrdU or PCNA. These results indicate that adult macaque monkeys harbor proliferating cells in forebrain regions homologous to the SZ/RMS system in adult rodents.

The presence of BrdU-labeled cells in the monkey forebrain after 97 days postinjection (the longest survival period examined) indicated that at least some newly generated cells survive more than 3 months after incorporating BrdU. Such an extended survival argues against both the possibility of a rapid turnover of all new cells and of BrdU cytotoxicity (26) at the doses used in this study.

New Cells Express a Neuronal Phenotype.

To determine whether BrdU-labeled cells in the SZ, olfactory peduncle, and OB expressed a neuronal phenotype, we used triple-label immunofluorescence for BrdU and the following cell type-specific markers: GFAP, a marker for astrocytes (27), and either TuJ1, a mAb to neuron-specific class III β-tubulin, which is expressed early in young neurons (28), or NeuN, a mature neuronal marker (29). We used laser-scanning confocal microscopy to examine the phenotype of BrdU-labeled cells at 32, 75, and 97 days after the final of five daily BrdU injections. No cross-immunoreactivity was detected between GFAP and either of the two neuronal markers used.

Examination of parasagittal forebrain sections revealed a dense plexus of cells bordering the head of the caudate nucleus in the anterior SZ that were intensely immunopositive for TuJ1 (Fig. 2). Many of these TuJ1-positive cells had a BrdU-positive nucleus, indicating that these cells were newly generated. Some individual BrdU/TuJ1 double-labeled cells exhibited a complex morphology with ramified processes (Fig. 2 d–f). Other BrdU/TuJ1-positive cells displayed a bipolar morphology that was similar to that of immature migrating neurons (12, 30), having a thick “leading” process that terminated in a growth cone-like swelling and a thinner, opposing trailing process (Fig. 2 j–k). BrdU/TuJ1-positive cells appeared to aggregate into chains that were oriented roughly parallel to the striatal wall of the ventricle, similar to the neuroblast chains observed in adult rodents (10, 31) (Fig. 2 a–f). Moreover, these cells were surrounded by a dense meshwork of GFAP-positive fibers and cell bodies. Within the region surrounding the anterior lateral ventricle, these chains were restricted to the SZ region and did not extend into the caudate parenchyma—where TuJ1 immunostaining of mature neurons was much weaker—or into the overlying cerebral white matter. The monkey SZ also contained BrdU-positive cells that were immunonegative for GFAP and TuJ1 and had a relatively large, spherical nucleus (Fig. 2 d–f).

Figure 2.

Newly generated cells in the anterior SZ of the adult macaque display a neuroblast phenotype as revealed by triple-label immunofluorescence and confocal microscopy. (a–c) A sagittal section through the striatal wall of the LV provides an oblique view of the anterior SZ, revealing an extensive network of TuJ1-positive cells (red) that are distributed singly or in chain-like aggregates. Imaging in the same x, y, and z registration reveals that many of these cells are colabeled with BrdU (b and c, green) and are closely associated with GFAP-positive fibers (c, blue). The TuJ1-positive chains do not extend into the adjacent caudate nucleus (CN) or overlying cortical white matter. (d–f) The same field partially demarcated by the box in c at higher magnification shows chains and individual BrdU-labeled neuroblasts (arrows) and their proximity to the GFAP-immunopositive ependymal lining (E in f). A BrdU-labeled cell (arrowhead) that is immunonegative for both TuJ1 and GFAP may be a “type C” precursor. (g–i) A 0.6-μm-thick optical section of the same BrdU-labeled neuroblast indicated by the crossed arrow in d–f, confirming that the BrdU signal (h and i, green) is confined to the nucleus of the TuJ1-labled cell rather than to an adjacent TuJ1-negative cell. (j–l) An example of a TuJ1-postitive cell with a BrdU-labeled nucleus (arrowhead), exhibiting a bipolar morphology similar to that of a migrating neuron. A growth cone-like swelling (arrow) appears at the end of the putative thick “leading” process, and a thinner process “trails” behind. a–i show labeling 75 days after BrdU injections; j–l, 97 days after BrdU injections. (Bars = 50 μm.)

Within the core of the olfactory peduncle, BrdU/TuJ1-positive chains formed a narrow, highly restricted stream that extended linearly through the length of the olfactory tract (Fig. 3). This stream was bordered by a dense meshwork of GFAP-positive fibers (Fig. 3 a and d). This arrangement of cells is also characteristic of the RMS in rodents, in which TuJ1-immunoreactive chains of migrating neuroblasts are ensheathed by “tubes” of GFAP-positive astroglial processes (31, 32). Although many BrdU-labeled cells in the SZ and RMS were TuJ1-positive, none were found to express NeuN, which is consistent with the immature status of neuroblasts. NeuN immunofluorescence was confined to the nuclei and perikarya of mature neurons (Figs. 3a and 4).

Figure 4.

Newly generated cells in the adult macaque monkey OB 97 days after the last of five BrdU injections. (a–d) Granule neurons in the OB express NeuN (red), and astrocytes express GFAP (blue). (a and b) A cell in the granule cell layer that is labeled with BrdU (arrow, yellow green in b) also expresses NeuN (arrow, a), indicating a newly generated granule neuron. (c and d) An example of two BrdU-labeled nuclei (d, arrows, green) that did not emit a red fluorescence signal (c, arrows), demonstrating that the BrdU signal did not bleed into the red channel; these might be progenitors or newly generated nonneuronal cells. (Bars = 20 μm.)

In rodents, chains of neuroblasts in the SZ and RMS highly express PSA-NCAM (10, 32, 33). This molecule is important for the neurophilic migration of neuroblasts to the OB (8, 34). To determine whether the neuroblast chains in the adult monkey similarly express PSA-NCAM, we examined parasagittal forebrain sections that were double-stained for PSA-NCAM and TuJ1. TuJ1-positive chains in the monkey SZ and RMS were strongly copositive for PSA-NCAM (Fig. 3 e–g), similar to the staining pattern described in rodents.

At 32 or 75 days after BrdU injection, none of the BrdU-positive cells in the OB also expressed the mature neuronal marker, NeuN. Although some BrdU-positive cells in the granule cell layer had nuclei of similar size and shape to those of neighboring NeuN-positive granule neurons, they were immunonegative for NeuN and GFAP. Thus, we could not distinguish whether these BrdU-labeled cells might be immature granule neurons or nonneuronal cells. These observations suggested that if any new BrdU-labeled neurons in fact were present in the OB at these time points, they were not yet differentiated. However, at 75 days postinjection, a large population of BrdU/TuJ1-positive cells was present in the RMS as it entered the OB. This apparent surge of new cells was not observed at the earlier, 32-day survival time. By 97 days postinjection, some of the BrdU-positive cells in the granule cell layer were immunoreactive for NeuN and were similar in size and shape to adjacent NeuN-positive granule neurons, indicating the presence of newly generated cells that expressed a mature neuronal phenotype (Fig. 4). These results indicate that new olfactory interneurons are generated in the adult macaque monkey SZ but do not appear differentiated in the OB until 11–14 weeks later.

Discussion

The present results demonstrate the generation and chain migration of neuroblasts from the SZ to the OB in adult microsmatic primates. Previous documentation of this neurogenic migratory system has been limited to macrosmatic rodents. The very cell types that characterize the SZ/RMS in adult rodents also appear to be present in adult monkeys. The BrdU/TuJ1-positive cells in the SZ and RMS of adult monkeys resemble the migrating neuroblasts (type A cells) described in the rodent SZ/RMS (11, 32). In mice and monkeys alike, these cells form chains that are strongly immunopositive for both TuJ1 and PSA-NCAM and are immunonegative for GFAP. Likewise, the GFAP-positive cells that flank the neuroblast chains in monkeys probably are equivalent to the GFAP-positive astrocytes (type B cells) in the SZ and RMS of mice. In adult rodents, these glia ensheathe migrating neuroblasts, forming contiguous tubes that might restrict migration within the RMS (11, 31, 32). In the rodent SZ, type B astrocytes can function as multipotent stem cells, capable of generating type A neuroblasts and immature precursor cells (type C cells) (35–37). The BrdU-positive cells in the monkey SZ that were immunonegative for GFAP and TuJ1 may correspond to the type C precursor cells, which are scattered in small numbers in the adult rodent SZ (11, 32). These shared characteristics between mice and macaques indicate that a rostral migratory stream of neuroblasts is generated and maintained in the adult primate forebrain, homologous to that in rodents.

The present results also provide an estimate for the timing of the generation, migration, and differentiation of new granule neurons in the adult macaque monkey. Their delayed appearance in the bulb between 75 and 97 days after BrdU injections is consistent with the proposition that these cells originate from outside the bulb in the SZ and migrate into the bulb via the RMS before becoming mature neurons. This period is much longer than that reported in rodents. In adult mice, new olfactory interneurons require a total of at least 15 days to be generated, migrate 3–5 mm from the SZ to the OB, and develop a mature morphology (7). These cells migrate in the RMS at an average rate of 30 μm/hr. In the larger brains of macaque monkeys, the migratory route is considerably extended—the elongated olfactory peduncle alone is ≈20 mm long, which is longer than an entire mouse brain. Thus, even if rates of chain migration are comparable in mice and monkeys, one would not expect to see mature, SZ-derived neurons in the OB of monkeys until at least 30 days after being labeled in the SZ. That we did not detect new granule neurons until after 75 postinjection days may reflect slower rates of neuronal generation, migration, and/or differentiation in the primate. Such protracted development is consistent with the longer neuronal generation times in macaque monkeys vs. mice during fetal development (23) and with our observation that BrdU-labeled neuroblasts were still present in the adult monkey SZ for as long as 97 days postinjection. It is also possible that new granule neurons that potentially arrived and differentiated earlier than 75 days postinjection went undetected. Moreover, because we observed proliferating cells in the OB itself, we also cannot rule out the possibility that at least some of these new granule neurons were generated in situ and differentiated over a protracted time period. Such local neurogenesis, however, has not been reported in the rodent OB.

Previous studies—including those using BrdU or [³H]thymidine to label dividing cells—had shown evidence for proliferative activity in the adult SZ of the lateral ventricle in primate species ranging from New World monkeys (38) to Old World monkeys (39–42) to humans (17, 43). However, it was not determined whether these cells generate olfactory neuroblasts. The first indication that SZ cells in adult primates could be neurogenic came from in vitro evidence that temporal lobe SZ tissue from adult epileptic humans could generate cells with neuronal characteristics under tissue culture conditions (43–45). In support of these findings, two recent studies showed that some cells in the adult human anterior SZ are immunopositive for markers associated with a neuroblast phenotype, including PSA-NCAM and class III β-tubulin (46, 47). However, these immunohistochemical studies of postmortem tissue were unable to verify whether these cells were, in fact, newly generated; nor were they able to determine their fate. Our data confirm and extend these studies by providing evidence that dividing cells in the adult primate SZ normally generate new neurons that migrate in chains to the OB and differentiate into granule neurons. Together, these findings raise the possibility that an active SZ/RMS system is also present in humans.

Similar to rodents, chains of neuroblasts in the SZ and RMS of adult monkeys expressed PSA-NCAM, which has been shown in mice to mediate neuroblast migration to the OB (8, 34). This species similarity suggests that the very molecular cues that promote chain migration in the rodent olfactory stream also operate in the adult primate forebrain. It also was apparent that, as in rodents (11, 14, 32), migrating neuroblasts in adult monkeys were confined to the SZ and RMS and did not penetrate into the surrounding striatal or septal parenchyma or overlying cortical white matter, suggesting the presence of guidance cues that spatially restrict migration. Specific molecules that may play such a restrictive role in directing neuroblast migration from the SZ to the OB recently have been identified in rodents (48–50). By analogy to mechanisms that govern axonal pathfinding (51), it is likely that a combination of permissive and repulsive cues direct neuroblast migration in the adult mammalian brain and prevent widespread dispersion, even over the extended migratory distance of the primate RMS. Although the destination of SZ-derived neuroblasts normally is restricted to the OB, it may be possible to manipulate and redirect migrating neuroblasts to nonneurogenic brain regions suffering neuronal loss from trauma or neurodegenerative disease and to reestablish functional circuitry (52). Indeed, recent experimental evidence suggests that endogenous SZ-derived neuroblasts in adult mice actually can be induced in situ to repopulate injured neocortical areas—which are not normally neurogenic—and to establish axonal connections with appropriate thalamic targets (53). Conceivably, this effect could be potentiated by, for example, expanding the SZ progenitor pool through the introduction of mitogenic growth factors (54, 55). Successfully coercing the adult brain's own neural progenitor cells to compensate for neuronal loss in humans would circumvent the ethical and immunological problems posed by therapeutic strategies that rely on transplants of cells from human fetuses or nonhuman species. For therapeutic applications in humans, it will be essential to assess the capacity of the adult primate SZ/RMS system for compensatory plasticity in response to neuronal loss in distant brain regions. The present study establishes the macaque monkey as a phylogenetically proximate animal model in which to explore this issue as it relates to neural plasticity in humans.

Abbreviations

BrdU

5-bromodeoxyuridine

GFAP

glial fibrillary acidic protein

OB

olfactory bulb

PSA-NCAM

polysialylated neural cell adhesion molecule

RMS

rostral migratory stream

SZ

subependymal/subventricular zone

PCNA

proliferating cell nuclear antigen