CNS lesions stimulate adult neurogenic niches. Endogenous neural stem/progenitor cells represent a potential resource for CNS regeneration. Here, we investigate the response to unilateral focal laser-lesions applied to the visual cortex of juvenile rats. Within 3 days post-lesion, an ipsilateral increase of actively cycling cells was observed in cortical layer one and in the callosal white matter within the lesion penumbra. The cells expressed the neural stem/progenitor cell marker Nestin and the 473HD-epitope. Tissue prepared from the lesion area by micro-dissection generated self-renewing, multipotent neurospheres, while cells from the contralateral visual cortex did not. The newly formed neural stem/progenitor cells in the lesion zone might support neurogenesis, as suggested by the expression of Pax6 and Doublecortin, a marker of newborn neurons. We propose that focal laser-lesions may induce the emergence of stem/progenitor cells with neurogenic potential. This could underlie the beneficial effects of laser application in neurosurgery.
The CNS originates from self-renewing, multipotent neural stem cells that have traditionally been classified according to their morphological appearance during histogenesis (Bystron et al., 2008). The forebrain develops from a single layered neuroepithelium giving rise to radial glia cells that generate neurons during embryogenesis and oligodendrocytes and astroglia around birth (Merkle and Alvarez-Buylla, 2006). Postnatally the neural stem cells disappear or become quiescent, and those that remain do not undergo neurogenesis, with two marked exceptions. First, in the lateral ventricle wall of the forebrain subventricular zone, astrocytes of radial glia origin (Merkle et al., 2004) serve as neural stem cells and sustain olfactory bulb neurogenesis in the adult telencephalon (Imayoshi et al., 2008; Zhao et al., 2008). Second, granule neurons are sequentially born from neural stem cells in the subgranular layer of the hippocampal dentate gyrus of the adult forebrain (Kempermann et al., 2004). Both neurogenic regions of the adult brain are regarded as niches, or specialized environments that sustain stem cells (Mirzadeh et al., 2008; Shen et al., 2008; Tavazoie et al., 2008) and function as integrative entities (Scadden, 2006) for various physiological stimuli that regulate adult neurogenesis (Zhao et al., 2008).
The control of adult neurogenesis occurs mainly on the level of neural stem cell—or progenitor—proliferation and/or by determining the survival of newborn neurons (Kronenberg et al., 2003). Remarkably, several pathophysiological conditions affect adult neurogenesis. For example, experimental stroke increases neural stem/progenitor cell proliferation and attracts them to the lesion area (Arvidsson et al., 2002; Thored et al., 2006). Brain tumours contain endogenous, and attract transplanted, neural stem/progenitor cells (Aboody et al., 2000). Finally, the injection of defined excitotoxins results in stimulation of neurogenesis in the CA3 region of the hippocampus (Nakatomi et al., 2002). Thus, it appears that several CNS pathologies activate the neural stem cell compartment, which raises hopes that an endogenous repair mechanism may exist. The latter may be activated by laser-induced lesions, which have been reported to be beneficial compared with analogous interventions based on scalpel use during neurosurgery in human patients (Jallo et al., 2002). Along these lines, studies describe that laser-induced lesions of the visual cortex of rodents cause synaptic plasticity in the penumbra region (Mittmann and Eysel, 2001), which depends on NMDA-receptors (Huemmeke et al., 2004).
It is presently, however, not known whether the application of laser pulses activates neural stem/progenitor cells in the adult niches, or whether the observed changes of synaptic plasticity may result from neurogenesis in the penumbra of the lesion. Therefore, we set out to examine an established laser-lesion model of the rat visual cortex (Eysel et al., 1999) to address these issues.
Here, we document that the laser-induced lesion of the visual cortex causes a rapid increase of proliferating neural stem/progenitor cells within the lesion area, as revealed by the presence of long-term self-renewing and multipotent cells in neurosphere assays (Reynolds and Weiss, 1992). The cells express the 473HD-epitope, a complex chondroitinsulfate (Ito et al., 2005) marker of embryonic and adult neural stem/progenitor cells (von Holst et al., 2006) which is up-regulated in stab-wound lesions of adult rats (Dobbertin et al., 2003). The cells express the neurogenic transcription factor Pax6 (Heins et al., 2002) and Doublecortin, which is transiently expressed by immature neurons. Thus, the activation of an endogenous neural stem/progenitor cell compartment elicits potentially neurogenic progenitors that may contribute to elevated synaptic plasticity and beneficial effects in the context of neurosurgery.
Materials and Methods
Wistar rats were obtained from Charles River Laboratories (Sulzfeld, Germany). Wister rats were divided into six groups (n = 28): one sham-operated control group (n = 4) and five groups composed of animals sacrificed at increasing times post-lesion [1, 3, 5 and 7 days (post-lesion); n = 6 per group]. Treatment of all animals was in accordance with the German regulations for experimentation with vertebrate animals, and the local ethic committee's approval was obtained from the regional government for the experimental protocols used.
The following primary antibodies were used in this study. Monoclonal antibodies: 473HD (rat IgM), 487/LeX (rat IgM), anti-Nestin (mouse IgG, Chemicon, Hofheim, Germany), anti-βIII-tubulin (mouse IgG, Sigma, Schnelldorf, Germany), anti-BrdU (mouse IgG, Roche Diagnostics, Mannheim, Germany) and anti-Vimentin (mouse IgM, Sigma). Polyclonal antibodies (all rabbit) against: DSD-1-PG/phosphacan (referred to as pk-anti-phosphacan, Batch KAF 13(4)), GFAP (Sigma), NG2 (Chemicon), Doublecortin (Chemicon). Secondary antibodies: subclass specific CY2- or CY3-coupled anti-mouse, anti-rat and anti-rabbit antibodies (all from Dianova, Hamburg, Germany).
Unilateral infrared-laser lesions in the visual cortex of rats (n = 28, 21 days of age) were induced as previously described (Mittmann and Eysel, 2001). The animals were anaesthetized by intraperitoneal injection of chloral hydrate (4% w/v; 0.1 ml per 10 g bodyweight). The skull was exposed and cautiously drilled thin above the visual cortex. The dura mater was not exposed, excluding a mechanical damage of the pia mater and cortex. The lesion was applied to the cortex with an 810 nm infrared diode laser (OcuLight SLx, Iris Medical, USA) under visual control. Elongated multiple, partially overlapping round lesions were formed at 1 mm mediolateral width and 4 mm anteroposterior length, starting anterior of the lambda suture in the visual cortex (area 17/18b). Finally, the wound margins were closely attached and fixed with histoacryl-tissue glue (Braun-Dexon, Melsungen, Germany). Sham-operated rats of the same age served as controls.
At defined time points after laser lesion, rats were anaesthetized by asphyxiation in CO2, and the brains were removed. The lesioned and non-lesioned contralateral areas of the cerebral cortex were punched out with sterile cylinders (Ø 0.4 cm) and separately transferred to minimal essential medium (MEM, Sigma). Isolated tissues were enzymatically dissociated in digestions solution [0.125% (v/v) trypsin/EDTA (Invitrogen), 0.008% (w/v) DNaseI (Worthington), 70 U/ml collagenase, 160 U/ml hyaluronidase in DMEM (all Sigma)] for 15 min at 37°C and another 15 min at 37°C on a shaker. The acutely dissociated cortical cells were pelleted and resuspended in neurosphere medium as described (von Holst et al., 2006). Aliquots of 104 cells were plated per well in 24-well tissue culture dishes and cultured in the presence of EGF, bFGF [both at (20 ng/ml); Preprotech, Tebu, Offenbach, Germany] and 0.5 U/ml Heparin (Sigma) for 14 days in a humidified incubator at 37°C, 6% CO2. In the neurosphere bulk culture, 100 000 cells were cultured in 4 ml growth factor containing medium, following established procedures (Marshall et al., 2008). Care was taken to limit cell concentration to obtain near to clonal conditions (Singec et al., 2006). The medium was changed once before the total number of neurospheres was quantified after 14 div. For differentiation assays, individual neurospheres ranging from 150 to 250 µm in diameter were transferred into four-well dishes (Greiner, Frickenhausen, Germany) coated sequentially with poly-ornithine (10 µg/ml; Sigma) and laminin-1 (10 µg/ml; Tebu) and cultivated for another seven div. The differentiated neural cell types were identified using the cell-type selective antibodies described above (von Holst et al., 2006).
For immunohistochemistry, 1d, 3d, 5d or 7d after the laser pulse (days post-lesion) rats were anaesthetized by intraperitoneal injection of chloral hydrate (4% w/v; 0,1 ml/10 g bodyweight). After intracardial perfusion with 2% Heparin in Krebs–Ringer–Hepes buffer and subsequently with 4% PFA, the brains were removed and post fixed in 4% PFA for 24 h at 4°C. The fixed brains were cryoprotected, cut into 20 µm thick sections and processed for immunohistochemistry as described previously (Sirko et al., 2007). For cryosectioning, the neurospheres (150 µm diameter) were fixed with 4% PFA in PBS for 40 min at RT. The cryoprotected neurospheres were sectioned at 14 μm and processed for immunochemistry. The immunocytochemical stainings on acutely dissociated cells after 1 h, neurosphere differentiation assays and neurospheres sections were carried out as described (von Holst et al., 2006; Sirko et al., 2007).
At defined time points the lesioned animals were anaesthetized by asphyxiation in CO2, the lesioned and non-lesioned contralateral areas (Ø 0,2 cm) of the brain were punched out and snap frozen in liquid nitrogen. Total RNA was extracted and reverse transcribed as reported earlier (von Holst et al., 2007; Akita et al., 2008). The primers and PCR conditions used in this study have recently been described for semi-quantitative [Pax6 (von Holst et al., 2007)] and real-time analysis [all other primers employed here (Horvat-Brocker et al., 2008)].
In situ hybridization
In situ hybridization with digoxigenin-labelled probes detecting the Ptprz1 isoforms Phosphacan and RPTPβ long was performed as described previously (Heck et al., 2005). In short, lesioned rat brains were snap frozen in isopentane at −50°C for 2 min and then cryosectioned. The 18 µm-thick frontal brain sections were fixed in 4% paraformaldehyde (PFA) in 0.1 M PBS for 20 min, washed in PBS, treated with 1 mg/ml proteinase-K in 50 mM Tris–HCl (pH 7,0) and 5 mM EDTA (pH 8,0) for 10 min, and then refixed. Sections were hybridized overnight at 50°C followed by overnight incubation at 4°C with anti-DIG Fab fragments conjugated with alkaline phosphatase. DIG detection was performed using the NBT/BCIP substrate (Roche) until dark purple precipitates were visible. The specimens were mounted with PBS/glycerol (1:1) and documented with brightfield optics.
The immunostained preparations were analysed using a fluorescence microscope equipped with UV-epifluorescence (Axioplan 2 imaging, Zeiss). Images were captured with a digital camera (AxioCamHRc, Zeiss) and documented using the Axiovision 4.2 and/or 4.5 programs. Images of in situ hybridization were acquired using an Axiophot microscope equipped with a digital camera and with Nomarski technique (Zeiss). In some cases, confocal laser scanning microscopy was applied (LSM 510 meta, Zeiss). Standard phase contrast images of living cells or neurospheres were taken using a digital camera (DP10, Olympus) on an inverted CK40 microscope (Olympus).
Counting and statistical analysis
The proportion of single- and double-immunolabelled cells was determined by counting 200–400 individual cells per independent experiment and epitope. Statistical significance of differences observed between distinct experimental groups was assessed using an unpaired, two-tailed Student's t test.
The visual cortex of postnatal rats does not contain self-renewing, multipotent neural stem cells
The two neurogenic regions of the adult mammalian brain contain significant numbers of self-renewing multipotent neural stem/progenitor cells (Zhao et al., 2008). The stem cell properties are revealed in the neurosphere assay, where single cells proliferate in the presence of EGF or FGF-2 to form cellular aggregates that are composed of neural stem cells and various types of neural progenitors undergoing different stages of differentiation. When performed under controlled conditions, this assay permits conclusions regarding the frequency of these cells in a given tissue explant (Marshall et al., 2007). Neural stem/progenitor cells that generate neurons, astrocytes and oligodendrocytes in vitro and have the capacity to form neurospheres have also been isolated from non-neurogenic regions of the adult CNS (Palmer et al., 1995, 1999; Weiss et al., 1996). Therefore, we first determined whether the juvenile and adult visual cortex of rats contain cells that display the central neural stem/progenitor cell features of self-renewal and multipotency, especially in light of the original description of neurogenesis in the adult rat visual cortex (Kaplan, 1981). No neurospheres were obtained in response to EGF and/or FGF-2 from single cells isolated from the visual cortex of postnatal Day 21 (P21) and of 6-month-old animals under our preparation and culture conditions.
In contrast, subventricular zone-derived cells from the same rats readily produced self-renewing, multipotent neurospheres (Fig. 1). As expected, cells derived from this area in adults formed significantly fewer neurospheres than those derived from the P21 subventricular zone (0.6% versus. 2%, respectively) (Fig. 1). Both populations gave rise to secondary and tertiary neurospheres and generated βIII-tubulin positive neurons, GFAP-positive astrocytes and O4-positive oligodendrocytes (Fig. 1D). Thus, the postnatal visual cortex does not contain detectable neural stem/progenitor-type cells under physiological conditions.
The available evidence suggests that adult neural stem/progenitor cells express Nestin, GFAP and Vimentin (Lendahl et al., 1990; Gage et al., 1995; Luskin et al., 1997; Doetsch et al., 1999). Furthermore, the specific carbohydrate structures LewisX (LeX) and the 473HD-epitope are expressed by adult mouse subventricular zone astrocytes and transient amplifying progenitor cells. Both markers can be used to immunoisolate self-renewing, multipotent neurosphere forming cells (Capela and Temple, 2002; von Holst et al., 2006). In agreement with the outcome of the neurosphere assay, 473HD-/Nestin- or 473HD-/GFAP-double immunopositive cells were detected in subventricular zone suspensions from both ages, but were absent in cell preparations of the visual cortex (Fig. 1E).
Focal laser-lesions induce changes in visual cortex cell composition
The presence of multipotent, self-renewing neural stem/progenitor cells defines an important conceptual distinction between neurogenic and non-neurogenic CNS regions. Yet, this difference may be rooted in complex functional states rather than in an innate cell autonomous capacity. Here, we used the focal laser-lesion of the visual cortex (Mittmann and Eysel, 2001) to examine whether this pathological condition induces endogenous neural stem/progenitor cells and neurogenesis in a non-neurogenic area. The thermal lesions consisted of a core of coagulated tissue 1–1.5 mm in diameter and reached ventrally into the deep cortical layers, while age-matched sham-operated animals served as controls (Fig. 2A). At 1 day post-lesion (1 dpl) we found Nestin- and Vimentin-positive cells selectively on the ipsilateral side within the parenchyma of visual cortex layer one (LI) and in white matter (WM) areas (Fig. 2B). Nestin- and Vimentin-positive cells were juxtaposed in the corpus callosum (CC), while only few Vimentin-positive cells were detected in the subcallosal layers (WM). At 3 days post-lesion, immunoreactivity for Nestin and Vimentin selectively increased on the ipsilateral side throughout the cortical layers I–III as well as in the callosal and white matter areas, indicative of reactive gliosis (Frisen et al., 1995). In contrast, the contralateral side appeared unchanged, although a small number of Nestin-positive cells in LI were recorded (Fig. 2B). As expected, we observed and quantified a robust GFAP-activation in the penumbra between 1 and 5 days post-lesion, which was pronounced in LI and in white matter areas along the anterior-posterior axis Supplementary Fig. 1). Few of the Nestin-positive cells in the penumbra of the lesion expressed the 473HD-epitope that was prominent on extending processes (Fig. 2C). Consistent with this pattern, the core proteins RPTP-β and phosphacan of the Ptprz1 gene that carry the 473HD-epitope, were up-regulated 2- to 3-fold ipsilaterally from 1 to 7 days post-lesion (Supplementary Fig. 2). Taken together, the focal laser-lesion of the visual cortex caused preferentially on the ipsilateral side a reactive gliosis that was paralleled by the up-regulation of immature neural stem/progenitor markers in upper and deep cortical layers.
Focal laser-lesions increase proliferation and allow for neurogenesis
Next, we monitored in comparison to the contralateral side actively cycling cells in the visual cortex that had been labelled with a cumulative pulse of BrdU on 1, 3 and 5 days post-lesion (Fig. 3A). The number of BrdU-positive cells was augmented at 1 day. The total number of BrdU-positive cells increased on both sides at 3 and 5 days post-lesion, with a highly significant 5- to 10-fold expansion ipsilaterally (Fig. 3B and C). In all animals examined, the BrdU-positive cells were more abundant in the upper cortical layers and in the white matter of the injured visual cortex. Remarkably, within the first 3 days post-lesion most BrdU-positive cells localized to the penumbra, many of which expressed the 473HD-epitope on their surface. This was not the case for the contralateral BrdU-labelled cells (Fig. 3C). To determine the survival and fate of the BrdU-positive cells, rats were injected with BrdU for 2 days between 3 and 5 day post-lesion and sacrificed two weeks after lesion (Fig. 3D). On the ipsilateral side, three-times as many BrdU-positive cells were found at 14 days post-lesion, although their total number was smaller compared with the situation at 5 days, suggesting that approximately half of the newborn cells do not survive (Fig. 3F). To investigate whether newborn cells can differentiate into neurons, BrdU-positive cells were counter labelled for the immature neuronal marker Doublecortin (Dcx) at 5 and 14 days post-lesion (Fig. 3E). Although cycling cells were detected in both hemispheres, BrdU- and Dcx-double-positive cells occurred only ipsilaterally and their number significantly increased between 5 and 14 days post-lesion (Fig. 3G). This observation suggests that the laser lesion induced a protracted generation of newborn neurons in the visual cortex.
The lesion area contains cells with neural stem cell characteristics
To assess the composition of the lesion area on the cellular level of resolution, the marker profiles of cell suspensions prepared by micro-dissection from the ispi and contralateral visual cortex were compared. Comparable numbers of live cells were obtained at 1–7 days post-lesion from each hemisphere (Fig. 4A). Consistent with the results of immunohistochemistry, Nestin-positive cells that co-expressed GFAP, Vimentin or the 473HD-epitope were demonstrable ipsilaterally within 3 days post-lesion. Importantly, most of the BrdU- and Nestin-positive cells also surface expressed the 473HD-epitope (Fig. 4B). In agreement with the observation of newborn immature neurons, we recorded a significant elevation of the expression of the neurogenic transcription factor Pax6 (Heins et al., 2002) on the lesion side between 3 and 5 days post-lesion (Fig. 4C and D). As Pax6 by itself is sufficient to induce functional neurons (Berninger et al., 2007) this finding provides further evidence for neurogenesis within the lesioned visual cortex. This is consistent with a previous report that NMDA-currents characteristic for young neurons have been recorded in the lesion penumbra (Huemmeke et al., 2004). Interestingly, the alternatively spliced Pax6 isoform Pax6(5a), which increases proliferation without changing cell fate (Haubst et al., 2004), was also significantly up-regulated (Fig. 4C and D). In view of the expression of neural stem/progenitor cell markers in the lesion area, we tested the micro-dissected cell population for neurosphere formation and self-renewal in clonal density assays. Clearly, the lesion area of the visual cortex from 1 to 7 days post-lesion on contained neurosphere-forming cells, while suspensions prepared from the contralateral hemisphere generated dwarf-neurospheres at 3- to 4-fold lower frequency (Fig. 5A and B). The neurospheres obtained 3–7 days post-lesion from the lesion side could be expanded and propagated beyond five passages, which documents long-term self-renewal (Figs 5A, B and 6). In conjunction with the selective occurrence of 473HD-epitope- and BrdU-double-positive cells in the penumbra at 3 days post-lesion these findings are indicative for the emergence of neural stem/progenitor cells in the lesion territory (Fig. 3B). Different from this population, those BrdU-incorporating cells visible within 1 day post-lesion represent 473HD-epitope-negative progenitors with limited self-renewal capacity (Fig. 5B).
To establish multipotency, we examined the differentiation pattern of the lesion-derived neurospheres after plating on a laminin-1 substrate (Fig. 5C). The quantitative assessment of the relative numbers of GFAP-positive astrocytes, O4-positive oligodendrocytes and βIII-tubulin-positive neurons revealed that the major neural lineages were generated (Fig. 5D). In contrast, the dwarf-neurospheres obtained at 1 day post-lesion that lacked long-term self-renewal also failed with respect to neurogenesis and produced glial cells only (Fig. 5D).
Neurospheres are composite assemblies that contain various types of progenitors and differentiating cells (Marshall et al., 2007). To ascertain the presence of neural stem cells, we tested whether primary neurospheres derived from the lesioned postnatal visual cortex at 3, 5 and 7 days post-lesion could be passaged over longer periods. The endogenous neural stem/progenitor cells exhibited neurosphere-initiating capacity for more than nine passages (Fig. 6A) and thus satisfied the key criterium of long-term self-renewal (Louis et al., 2008). Furthermore, these neural stem/progenitor cells maintained multipotency after several passages in that they generated neurons, astrocytes and oligodendrocytes in vitro (Fig. 6B–D).
So far, juvenile rats were used for the laser lesions that may harbour a transient progenitor population that is missing in the adult. To exclude that the neurosphere forming cells described in our study derive from such a population, the experiments were repeated in 6-months-old animals. Cell suspensions were prepared from the adult lesioned and contralateral visual cortex at 5 days post-lesion. In this situation, our findings were equivalent to those collected with P26 visual cortices (corresponding to P21 animals at 5 days post-lesion). That is neurosphere-forming, expandable, self-renewing neural stem/progenitor cells with the characteristic antigenic profile occurred solely in the lesion penumbra, but not in the control hemisphere (Supplementary Fig. 3). Therefore, our key observation that focal laser-lesions elicit the appearance of neural stem cells with the potential to generate ex vivo immature neurons in non-neurogenic regions may be generalized to the whole period of ontogeny. This laser-related effect may prove beneficial in the clinical context, because mechanical lesions such as the stab wound have recently been reported to induce neural stem/progenitor cells without neurogenic potential in mice (Buffo et al., 2008).
In the present work, we could show that the visual cortex displays a remarkable ability to undergo plastic cellular changes after laser-induced lesion. This lesion type caused a rapid increase in an actively cycling progenitor cell population that expressed the radial glia markers Vimentin, Nestin, GFAP, Pax6 and the 473HD-epitope but not NG2 and gave rise to newborn neural stem/progenitor cells in the lesion area in vivo. Importantly, when micro-dissected from the lesion area these cells gave rise to self-renewing neurospheres ex vivo that generated neurons, oligodendrocytes and astrocytes. Such multipotent, self-renewing neurospheres were not obtained from the contralateral visual cortex. Thus, we propose that laser-lesions activate the neural stem cell compartment and may induce neurogenesis in cortical areas that are regarded as non-neurogenic. This phenomenon may explain the better tolerance for laser-based as compared with knife-operated interventions in neurosurgery. Similarly, neurogenesis has been reported in the penumbra of stroke in the human brain (Jin et al., 2006).
A central question raised by our work concerns the anatomical source of the novel neurogenic progenitor cell population that we have identified in the environment of the laser lesion. It has been documented that various CNS pathologies activate the neural stem cell niche in the anterior lateral ventricle wall in animal models as well as in humans (Komitova et al., 2005). The subventricular zone astrocytes undergo increased proliferation after injury and give rise to transient amplifying cells that are attracted to the lesion area. Depending on the insult these cells may undergo neurogenesis (Emsley et al., 2005) and can contribute to functional recovery in motor cortex (Kolb et al., 2007). The transient amplifying cells often migrate along fibre tracts. Interestingly, we recorded an increased number of BrdU-positive cells in the subcallosal area, which presumably originated in the adult subventricular zone. However, considering the large distance to the lesion side in the posterior visual cortex, it seems unlikely that the subventricular zone is the source of the neural stem/progenitor cells that could be obtained from the lesion area 3 days post-lesion.
The subgranular region in the dentate gyrus of the hippocampus (Glass et al., 2005) represents the other endogenous source of neurogenic progenitors in the adult. This cell population responds to various physiological stimuli, but has as yet not been described as source of neural stem/progenitor cells outside the hippocampal area. Rather, it appears that in the dentate gyrus these are restricted to a granular cell fate. Therefore, it seems unlikely that they contribute those emerging in the cortical laser-lesion territory. Likewise, subgranular zone neural stem cells did not contribute to neurogenesis in a kainate-induced lesion in CA3 (Nakatomi et al., 2002). In this latter work, neurogenesis and regeneration were attributed to a newly identified cell population in the posterior periventricular area. Although it has been argued that the occurrence of BrdU-positive neurons may be attributable to re-entry into S-phase of existing neurons undergoing prolonged cell death after hypoxia–ischemia (Kuan et al., 2004) and not regeneration, this position would be located in sufficient proximity to reach the lesion site within 3 days. However, our systematic analysis of BrdU- and GFAP-labelling along the anterior-posterior axis did not reveal an enhanced number of cells originating from the posterior periventricular area, rendering this region unlikely as origin of the neural stem/progenitor cells in the lesion area. More recently, a small cell population that gives rise to neurons during development has been identified in the marginal zone (Costa et al., 2007). Although this forebrain area has not been characterized as another neurogenic niche in the adult brain, it is interesting to note that we recorded a massive proliferation response in layer I of the visual cortex, which is rich in glial cells but sparse in neurons. Thus, quiescent glia from the marginal zone (i.e. layer I) could become activated and provide the source of neural stem/progenitor cells surrounding the lesion. Indeed, most traumatic insults give rise to reactive astrocytes that form the glial scar, as confirmed in our study. The astroglial origin of such reactive glial cells has recently been revealed in stab-wound lesions of the somatosensory/motor cortex using Cre-mediated lineage tracing (Buffo et al., 2008). These astroglial cells dedifferentiate and migrate to the lesion, where they form the glial scar, but do not undergo neurogenesis (Buffo et al., 2008). However, when removed from the lesion area these astroglial cells behaved like neural stem cells in the neurosphere model, in contrast to NG2- or PDGFRalpha-positive cells of the oligodendroglial lineage (Buffo et al., 2008). In our approach, the appearance of NG2-positive cells in the laser-lesioned parenchyma of the visual cortex was clearly different. NG2-positive cells were not found at 3 days post-lesion and appeared at later time points, most prominently in subpial layers (Supplementary Fig. 4). In addition to spatiotemporal and morphological differences in comparison with the radial glia-like neural stem/progenitor cells found in the penumbra of the lesion, only few NG2-positive cells were actively cycling (Supplementary Fig. 4). This clearly suggests that NG2 labels a different cell population. In the light of these findings, it seems reasonable to hypothesize that the neural stem/progenitor cells that emerged in our study are of astroglial origin, in agreement with increasing evidence for astroglial stem cells in the adult CNS (Laywell et al., 2000, 2007).
Interestingly, however, some of the cells we observed upon laser lesion expressed Doublecortin, in contrast to the situation in the stab wound lesion (Buffo et al., 2008). This protein is expressed by newborn immature neurons (Jin et al., 2001), which conforms with the original report on the occurrence of neuroblasts in the visual cortex. In fact, neurogenesis in the adult visual cortex of the rat had been observed nearly 30 years ago using electron microscopy of autoradiography sections (Kaplan, 1981, 2001). In that study, the frequency of newly generated neuroblasts was estimated at 1:104 cells, which may be too low for detection in control cortices using the ex vivo neurosphere assay (Marshall et al., 2008). Yet, it remains to be proven that newly born neurons fully differentiate and integrate into cortical networks in our paradigm.
A wealth of electrophysiological recordings supports the occurrence of complex neuronal differentiation and plasticity in the laser lesion, which could reflect laser-induced neurogenesis. This interpretation is consistent with reports that the elimination of distinct neuronal subpopulations by phototoxicity elicits neurogenesis and replacement of cortical neurons in the adult (Magavi et al., 2000; Chen et al., 2004). The exploitation of this regenerative capacity might represent a rewarding objective for further investigations.
Supplementary material is available at Brain online.
Center of Collaborative Research of the German Research Foundation (DFG, SFB 509 to U.E, T.M and A.F); German Ministry of Research and Technology (BMBF program ‘Stem cells for therapies of the CNS’, 01GN0504 to A.F.).
The authors thank Ute Neubacher and Anke Baar for excellent technical assistance.