- Email: [email protected]
Adult neurogenesis with 5-HT expression in lesioned goldfish spinal cord
Neuroscience 151 (2008) 1132–1141
ADULT NEUROGENESIS WITH 5-HT EXPRESSION IN LESIONED GOLDFISH SPINAL CORD A. TAKEDA, M. NAKANO, R. C. GORIS AND K. FUNAKOSHI*
genitor cells (NPCs) in the SVZ migrate through the rostral migratory stream to the olfactory bulb to differentiate into either granule neurons or periglomerular neurons, and those in the SGZ of the dentate gyrus migrate a short distance to become granule cells. Adult neurogenesis in the SVZ and SGZ is up-regulated by injury and pathologic stimulation in mammals (Parent, 2003). Newborn neurons might replace damaged cells after injury in the striatum, although few newborn neurons survive for a long time (Arvidsson et al., 2002). In the spinal cord, in vitro studies suggest that there are stem cells with self-renewal and multipotential features in mammals (Weiss et al., 1996). In non-pathologic conditions, proliferating cells in the ependymal and parenchymal layers are essentially glial progenitors producing astrocytes or oligodendrocytes (Horner et al., 2000). The proliferative activity of these progenitors is increased in response to spinal cord transection, but few neurons are generated from the endogenous NPCs (Yamamoto et al., 2001). Transplantation of adult-derived NPCs into the spinal cord after traumatic injury also fails to generate neurons in mammals (Cao et al., 2001; Vroemen et al., 2003; Pfeifer et al., 2004; Karimi-Abdolrezaee et al., 2006). The reason for the predominance of glial differentiation is not clear. One possibility might be the lack of an appropriate endogenous environment for neurogenesis in the spinal cord, because in vivo infusion of growth factors induces neurogenesis from endogenous NPCs (Karimi-Abdolrezaee et al., 2006). On the other hand, transplantation of embryonic stem cells or embryonic NPCs also induces differentiation into neurons, as well as into astrocytes and oligodendrocytes (McDonald et al., 1999; Ogawa et al., 2002). Adult neurogenesis is thought to be an evolutionarily conservative trait inherited from a common chordate ancestor (Zupanc, 2001a). In contrast to mammals, anamniotes (fishes and amphibians) have an enormous potential for producing new neurons in the adult CNS. In teleosts, neurogenesis is observed in various regions of the CNS, such as the olfactory bulb (Alonso et al., 1989; Byrd and Brunjes, 2001), optic tectum (Raymond and Easter, 1983), and cerebellum (Zupanc, 1999; Zupanc and Ott, 1999). Retinal ganglion cells in teleosts also continue to increase in number throughout their lives (Johns, 1981; Ito and Murakami, 1984). In the gymnotiform fish Apteronotus leptorhynchus, approximately 0.2% of the total population of the cells in the adult brain are in S phase. Many of these cells, produced in specific proliferation zones at the ventricular surface, migrate within a few weeks to specific target areas (Zupanc, 2001b). In the cerebellum, although
Department of Neuroanatomy, Yokohama City University School of Medicine, 3-9 Fukuura, Kanazawa-ku, Yokohama, Kanagawa, 2360004 Japan
Abstract—In contrast to mammals, spontaneous nerve regeneration occurs in the teleost spinal cord. In the present study, we examined whether neurogenesis is involved in posttraumatic regeneration in the goldfish spinal cord. In intact fish, many spinal cells positive for both a monoclonal neuronal marker (Hu) and bromodeoxyuridine (BrdU) were observed 24 h after i.p. injection of BrdU, suggesting that constant neurogenesis occurs in the goldfish spinal cord. After hemisection of the spinal cord, the number of spinal cells positive for Hu and BrdU was significantly increased around the lesion site. The number of Hu- and BrdU-positive cells reached the maximum level 7 days after hemisection. In intact fish, spinal cells positive for both Hu and BrdU were also observed 5 weeks after BrdU injection, suggesting that newborn neurons survive for a long time. Six weeks after hemisection, the number of surviving Hu- and BrdU-positive cells at the lesion site was significantly increased as compared with that in intact fish, and some of them were also positive for 5-HT. A retrograde tract tracing study showed that the 5-HTⴙ neurons were close to the regenerated axons passing through the lesion site. These results suggest that adult neurogenesis occurs in the goldfish spinal cord, and that neurogenesis is activated by spinal cord lesion. The newly produced neurons survive a long time at the lesion site, and might participate in the repair of injured tissue and in the regeneration of descending long axons beyond the lesion site. © 2008 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: regeneration, CNS, teleost, BrdU.
Adult neurogenesis, a process of producing new neurons throughout life, occurs in discrete regions of the CNS in mammals, i.e. the subventricular zone (SVZ) of the lateral ventricle and the subgranular zone (SGZ) of the dentate gyrus in the hippocampus (Ming and Song, 2005; Lledo et al., 2006). Newborn neurons generated from neural pro*Corresponding author. Tel: ⫹81-45-787-2569; fax: ⫹81-45-782-7251. E-mail address: [email protected] (K. Funakoshi). Abbreviations: BrdU, 5=-bromo-2=deoxyuridine; BSA, bovine serum albumin; DTAF, dichlorotriazinyl amino fluorescein; Elav, embryonic lethal abnormal visual; FGF, fibroblast growth factor; NDS, normal donkey serum; NPC, neural progenitor cell; PB, phosphate buffer; PBS, phosphate-buffered saline; PBST, 0.1 M phosphate-buffered saline containing 0.3% Triton X-100; PFA, paraformaldehyde; RDA, tetramethylrhodamine dextran amine; SGZ, subgranular zone; SNK, Student-Newman-Keuls; SVZ, subventricular zone; TPBS, 1.4 M NaCl, 27 mM KCl, 81 mM Na2HPO4-12H2O, 15 mM KH2PO4, 1% Tween-20; TUNEL, terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling.
0306-4522/08$32.00⫹0.00 © 2008 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2007.10.059
1132
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
half of newborn cells undergo apoptosis, some of the remaining cells differentiate into granule cells (Zupanc, 2001b). Injury induces high proliferative activity of cerebellar neurons, especially in areas close to the lesion. Newborn cells migrate to the lesion site to replace damaged cells, and some of them are incorporated into the cerebellar circuit as granule cells (Zupanc and Ott, 1999). The teleost spinal cord also has the ability to regenerate. In the weakly electric gymnotiform teleost Sternarchus (now Apteronotus) albifrons, the caudal segments of the spinal cord are regenerated after amputation of the tail (Anderson and Waxman, 1983). In this fish, neuronal precursors in the ependymal layer generate new neuroblasts that proliferate and differentiate into neurons in vitro (Anderson et al., 1994). Axonal projections in the spinal cord are also regenerated after mechanical injuries in teleosts. After the gap between the proximal and distal stumps is repaired, both ascending and descending spinal axonal projections regenerate through the lesion site in the goldfish (Sharma et al., 1993; Hanna et al., 1998) and zebrafish (Becker et al., 1997). In the goldfish, ependymal cells in the spinal cord are suggested to have high proliferative activity in response to traumatic injury (Yamada et al., 1997). The contribution of newborn neurons to spinal regeneration in this case, however, is largely unknown. The present study aimed primarily to examine whether neurogenesis is involved in regeneration of the spinal cord after injury. To detect newborn cells, 5=-bromo-2=deoxyuridine (BrdU), a thymidine analog incorporated into newly synthesized DNA in the S phase, was injected intraperitoneally in combination with immunoreactive studies with a neuronal marker, HuC/HuD. Hu proteins including HuC and HuD are RNA-binding proteins of the embryonic lethal abnormal visual (Elav) family, and function early in neuronal differentiation in vertebrates (Marusich et al., 1994). 5-HT is a neurotransmitter that has regulatory effects on cell proliferation within the neuroepithelium during development (Lauder, 1990). 5-HT also stimulates axonal growth during the developmental process (Lauder, 1993). To shed light on the potential role of 5-HT in tissue repair and axonal regeneration, we focused on the expression of 5-HT in neuronal cells in the spontaneous regeneration of the teleost spinal cord.
EXPERIMENTAL PROCEDURES The goldfish, Carassius auratus (body weight 15– 40 g; n⫽47), were obtained commercially (Nomoto Fish Farm Co. Ltd., Yokohama, Japan) and maintained in an aquarium at 25–28 °C. All procedures were performed according to the standards established by the NIH Guide for Care and Use of Laboratory Animals and the protocols were approved by the Yokohama City University Committee for Animal Research. All efforts were made to minimize the number of animals used and their suffering.
Immunohistochemical analysis of adult neurogenesis after spinal hemisection A total of 21 fish were used for fluorescent immunohistochemistry.
1133
Hemisection of the spinal cord Fifteen of the 21 fish were used for hemisection, and six as controls. The fish were deeply anesthetized with 0.02% 3-aminobenzoic acid ethyl ester (MS-222, Sigma Co., St. Louis, MO, USA) in water, and placed on ice. An incision of the dorsal skin and muscles was made in an appropriate region, and the post-temporal bone and vertebra were exposed. The rostral segments of the spinal cord were exposed after removal of the bones, and a frontal hemisection of the left side of the spinal cord was performed 500 m caudal to the first spinal nerve. The hemisection was performed by inserting the blades of a small scissors at right angles to the spinal surface along the posterior median septum. The wound was sutured and sealed with aerosol plastic dressing (Yoshitomi-Seiyaku, Osaka, Japan).
Injection of BrdU The fish (n⫽21) were anesthetized with 0.02% 3-aminobenzic acid ethyl ester (Sigma) in water. BrdU (100 g/g, Sigma) diluted in saline was injected intraperitoneally into intact fish (n⫽3), and hemisected fish 1 h (n⫽3), 2 days (n⫽3), 6 days (n⫽3), and 13 days (n⫽3) after hemisection, and they were allowed to survive 24 h in the aquarium. In addition, to examine the long-term survival of newly generated cells, fish were allowed to survive 5 weeks in the aquarium after BrdU (100 g/g) diluted in saline was injected intraperitoneally into intact fish (n⫽3), and into hemisected fish (n⫽3) 6 days after the operation.
Tissue preparation After various survival times, the fish (n⫽21) were anesthetized with 0.02% MS-222 in water, and perfused transcardially with 25 ml of saline containing 1% heparin, followed by 50 ml of 0.1 M phosphate buffer (PB, pH 7.4) containing 4% paraformaldehyde (PFA). The spinal cord including the hemisectioned level was removed immediately, postfixed in 0.1 M PB containing 4% PFA for 5– 6 h at 4 °C, and left overnight in 0.1 M PB containing 25% sucrose at 4 °C for cryoprotection. The spinal cord was embedded in Tissue-Tek OCT compound (Sakura, Tokyo, Japan), and frozen in liquid nitrogen. The frozen specimens were cut serially into 20-m thick horizontal sections in a cryostat (Microm, Walldorf, Germany), and thaw-mounted on gelatin-coated glass slides.
Multiple fluorescent immunohistochemistry Sections were dried for 1 h at room temperature, postfixed in 0.1 M PB containing 4% PFA for 30 min, and rinsed in 0.1 M phosphate-buffered saline (PBS, pH 7.4). The sections were incubated in 50 mM Tris–HCl (pH 8.0) for 30 min at 95 °C to activate the antigens and denature the DNA, and immediately cooled to room temperature. The sections were incubated with 0.1 M phosphate-buffered saline containing 0.3% Triton X-100 (0.1 M PBST, pH 7.4) for 10 min, and then incubated overnight at 4 °C in a moist chamber with a mixture of primary antibodies diluted with 1% normal donkey serum (NDS), 0.2% bovine serum albumin (BSA), and 0.1% NaN3 in 0.1 M PBST. Primary antibodies were used at the following dilution: sheep polyclonal anti-BrdU (25 g/ml, Exalpha Biologicals, Boston, MA, USA), mouse monoclonal antiHuC/D (20 g/ml, 16A11, Molecular Probes, Eugene, OR, USA), rabbit polyclonal anti-5-HT (1:1000, Sigma). Mouse monoclonal antibody against HuC/D (16A11, Molecular Probes) recognized the Elav family members HuC, HuD, and Hel-N1 (Szabo et al., 1991; Levine et al., 1993; Liu et al., 1995), which are all neuronal proteins expressed in both the CNS and peripheral nervous system of a variety of vertebrates, including the zebrafish (Byrd and Brunjes, 2001). After rinsing with PBST several times, the sections were then incubated for 3 h at room temperature with a mixture of secondary
1134
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
antibodies diluted with 1% NDS, 0.2% BSA, and 0.1% NaN3 in 0.1 M PBST. Secondary antibodies were used at the following dilution: dichlorotriazinyl amino fluorescein (DTAF) – conjugated donkey antimouse IgG (1:100; Chemicon), rhodamine (TRITC) -conjugated donkey anti-sheep IgG (1:100; Chemicon), and aminometylcoumarin (AMCA) -conjugated donkey anti-rabbit IgG (1:100; Chemicon). After rinsing with PBS several times, the sections were mounted with antifade reagent (SlowFade Gold; Molecular Probes). For quantitative analysis of labeled cells, we defined the lesion site as the spinal area extending from 0.5 mm rostral to 0.5 mm caudal to the incision level, because this was the size of the wound immediately after the incision. For controls, the spinal area from the level of the first spinal nerve to 1 mm caudal to the first spinal nerve was examined, because this was an area identical to that of the lesion site.
Observation and electronic imaging All the sections were analyzed under an epifluorescence microscope (Leica DMR; Leica, Wetzlar, Germany), and imaged with a CCD camera (Leica DC 200; Leica). Primary images of immunofluorescent reactivity were captured using DC Viewer Software (Leica). Contrast, brightness, and color adjustment were performed with Adobe Photoshop Software (Adobe, San Jose, CA, USA). For quantitative analysis, the number of labeled cells in which the nuclear contour could be distinguished was counted in each series of sections for immunofluorescent studies. The differences in the number of BrdU⫹ cells, Hu⫹/BrdU⫹ cells, Hu⫹/ BrdU⫹/5-HT⫹ cells in the lesion site after different survival times were analyzed with the Student-Newman-Keuls (SNK) test.
Western blotting An intact fish was used for verification of mouse monoclonal antibody against HuC/D (16A11, Molecular Probe). The spinal cord was dissected out from a deeply anesthetized goldfish, and 50 mg of the cord was homogenized in 500 m of blending buffer (Complete Lysis-M, Roche, Mannheim, Germany). Insoluble materials were removed by centrifugation at 12,000⫻g for 20 min at 4 °C. The supernatant was diluted six times with a buffer containing 125 mM Tris–HCl (pH 6.8), 10% 2-metacaptoethanol, 4% sodium dodecyl sulfate, 20% glycerol, 0.04% Bromophenol Blue, and heated for 5 min at 95 °C. The supernatant and molecular weight marker (Precision Plus Protein Standards, Bio-Rad, Hercules, CA, USA) were electrophoresed on polyacrylamide gel (e-PAGEL 10%, Atto, Tokyo, Japan). For immunoblotting, the separated proteins were transferred onto a nitrocellulose membrane (Protein BA85 Cellulosenitrat, Schlieicher & Schuell, Dassel, Germany) for 12 h at 4 °C at a constant voltage of 12 V. The membranes were then cut into strips, blocked with skim milk– TPBS (1.4 M NaCl, 27 mM KCl, 81 mM Na2HPO4-12H2O, 15 mM KH2PO4, 1% Tween-20) for 1 h, and incubated with TPBS containing mouse monoclonal antibody against HuC/D (5 g/ml, 16A11, Molecular Probes) overnight at 4 °C, followed by alkaline phosphate-conjugated anti-mouse IgG (1:15,000; Sigma) for 1 h at room temperature. The antibody was detected with nitro blue tetrazolium chloride/5-bromo-4-chloro-3-indolyl phosphate solution (Sigma).
Quantitative analysis of 5-HT cells Another 18 fish, i.e. control fish (n⫽3), and hemisected fish 1 day (n⫽3), 1 week (n⫽3), 3 weeks (n⫽3), 6 weeks (n⫽3), and 12 weeks (n⫽3) after hemisection of the spinal cord, were used for quantitative evaluation of 5-HT⫹ cells around the lesion site. After brief incubation with PBST, the sections were incubated overnight at 4 °C in a moist chamber with rabbit polyclonal anti5-HT (1:1000, Sigma). After rinsing with PBST, the sections were then incubated for 3 h at room temperature with donkey anti-rabbit
IgG (1:100; Chemicon). The primary and secondary antibodies were diluted with 1% NDS, 0.2% BSA, and 0.1% NaN3 in 0.1 M PBST. After rinsing with PBS, the sections were then incubated for 90 min at room temperature with rabbit peroxidase anti-peroxidase complex (1:100; Jackson Immunoresearch, West Grove, PA, USA) diluted with 1% NDS, 0.2% BSA, and 0.1% thimerosal in 0.1 M PBS. The peroxidase reactivity was demonstrated with 3,3=-diaminobenzidine (Sigma). The sections were then dehydrated in an ethanol series, cleared with xylene, and mounted with balsam. For quantitative analysis of labeled cells, we defined the lesion site as the spinal area extending from 0.5 mm rostral to 0.5 mm caudal to the incision. The differences in the number of 5-HT⫹ cells in the lesion site after different survival times were analyzed with the SNK test.
Combination of retrograde tract tracing and 5-HT immunohistochemistry To examine the spatial relationship between 5-HT cells and regenerating axons in the repaired tissue, we performed a combination of immunohistochemistry and retrograde tract tracing. First, the fish (n⫽2) were deeply anesthetized with 0.02% MS-222 (Sigma) in water and placed on ice. The spinal cord was exposed by an incision in the appropriate region of the dorsal skin and removal of the vertebrae, and frontal hemisection of the left side of the spinal cord was performed as described above. Six weeks after hemisection, the spinal cord was again exposed, and a strip of paper soaked in tetramethylrhodamine dextran amine (RDA; molecular weight⫽3000; 100 mg/ml; Molecular Probes) in PBS (0.1 M; pH 7.4) was applied to the proximal stump of the left side of the spinal cord at the level of the second spinal segment. After a survival period of 3– 4 days, the fish were re-anesthetized and perfused transcardially with saline containing 1% heparin, followed by 0.1 M PB (pH 7.4) containing 4% PFA. After postfixation and cryoprotection, the spinal cord including the lesion site was embedded in Tissue-Tek OCT compound and frozen. Then frontal sections were cut serially at 20 m in a cryostat, and were thaw-mounted on gelatin-coated glass slides. After brief incubation with PBST, the sections were incubated overnight at 4 °C in a moist chamber with rabbit polyclonal anti5-HT (1:1000, Sigma). After rinsing with PBST, the sections were incubated for 3 h at room temperature with DTAF-conjugated donkey anti-rabbit IgG (1:100; Chemicon). The primary and secondary antibodies were diluted with 1% NDS, 0.2% BSA, and 0.1% NaN3 in 0.1 M PBST. After rinsing with PBS, the sections were mounted with antifade reagent.
TUNEL method Because the incorporation of thymidine analog into DNA occurs not only during proliferative processes, but also during repair and apoptotic processes (Selden et al., 1993), demonstration of neurogenesis after brain injury requires evidence of the absence of apoptotic markers. To evaluate whether or how many neurons in the goldfish spinal cord undergo apoptosis after injury, horizontal sections from intact fish (n⫽1), and hemisected fish 1 day (n⫽1), 3 days (n⫽3), and 7 days (n⫽1) after hemisection were stained using the terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) method according to the manufacturer’s instructions (NeuroTACS II in situ Apoptosis Detection Kit, Trevigen, Gaithersburg, MD, USA).
RESULTS BrdU and Hu labeling after short-term survival In the intact spinal cord, BrdU-immunoreactive cells were observed in the gray matter 24 h after the BrdU injection (Fig. 1A). With immunohistochemistry using antibodies
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
1135
Fig. 1. (A) Many BrdU-labeled cells were detected in the intact spinal cord 1 day after a single BrdU injection. The majority of BrdU-positive cells were observed in the median and paramedian zones of the gray matter in the transverse sections. Dotted line: midline. (B) A cell (arrow) positive for both BrdU (red) and Hu (green) was observed in the intact spinal cord, 1 day after a single BrdU injection. (C) Numerous BrdU-labeled cells were observed in the spinal cord, particularly around the lesion site, 1 week after the left hemisection and 1 day after a single BrdU injection. Scissor mark: the site of the hemisection. Dotted line: midline. (D) Enlarged image of the white box in C. BrdU cells were observed in the ependyma surrounding the central canal (asterisk). (E) A cell (arrow) positive for both BrdU (red) and Hu (green) was observed in the spinal cord, 1 week after the left hemisection and 1 day after the BrdU injection. Scale bars⫽100 m in C (applies to A, C), 100 m in D, and 20 m in E (applies to B, E). (F) Western blotting of the intact spinal cord with anti-HuC/D. An immunoreactive band at 42– 43 kDa (arrowhead) was detected.
against HuC/D, approximately 35% of BrdU-labeled cells were also positive for HuC/D. BrdU⫹/Hu⫹-immunoreactive cells were scattered in the gray matter, not restricted to the region around the central canal. Huimmunoreactivity was observed in the cytoplasm of the cells, whereas BrdU-immunoreactivity was observed in the nucleus of the cells (Fig. 1B). In the hemisected spinal cord, numerous BrdU-labeled cells were observed 24 h after the BrdU injection (Fig. 1C). There were BrdU⫹ cells in the ependymal lining (Fig. 1D), but many labeled cells were also observed in the gray matter, particularly around the lesion site. The number of BrdU⫹ cells in the spinal cord was significantly increased. The maximum number of BrdU⫹ cells was recorded 1 week after hemisection, and after that the number of BrdU⫹ cells decreased (Fig. 2). There were many BrdU⫹ cells positive for HuC/D in the gray matter of the hemisected spinal cord (Fig. 1E), but not in the ependymal lining. The number of BrdU⫹/Hu⫹ cells was also significantly increased after hemisection, with the maximum level occurring 1 week after the hemisection (Fig. 3).
In control, no positive staining was observed when primary antibodies (anti-BrdU and anti-HuC/D) were omitted from the reaction. Western blotting In the sample of intact spinal cord, the anti-HuC/D reacted with a protein of 42– 43 kDa (Fig. 1F), which has been reported as the molecular weight of HuD (Szabo et al., 1991; Aronov et al., 2002; Ratti et al., 2006), suggesting that the anti-HuC/D specifically recognized the Hu protein in our study. BrdU and Hu labeling after long-term survival In the intact spinal cord, BrdU-immunoreactive cells were observed in the gray matter (Fig. 4A) and in the leptomeninges 5 weeks after the BrdU injection. A number of BrdU⫹ cells in the gray matter were also positive for HuC/D (Fig. 4B). Six weeks after spinal hemisection and 5 weeks after the BrdU injection, many BrdU-labeled cells were observed in the spinal cord, particularly around the lesion site (Fig. 4C). There were also BrdU⫹ cells in the ependy-
1136
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
the nucleus raphes pallidus and obscurus, and the caudal dorsomedial medullary group, and in the diencephalon nuclei including the posterior tuberal nucleus. Except for the descending axons, anterogradely labeled ascending axons might be present in the lesion site. The regenerated axons were in close apposition to 5-HT neurons (Fig. 6C). Apoptosis after spinal hemisection No cells were labeled by the TUNEL method in the spinal cord of intact fish (Fig. 8B). A few TUNEL-positive cells were seen around the lesion site of the spinal cord 1, 3, and 7 weeks after hemisection (Fig. 8C, E).
DISCUSSION
Increased number of serotonergic cells at the lesion site
Adult neurogenesis is thought to be an evolutionarily conservative trait inherited from a common chordate ancestor (Zupanc, 2001a). The ability for adult neurogenesis, however, was remarkably reduced in the branch leading to the mammals, i.e. neurogenesis seems restricted to the olfactory bulb and dentate gyrus in adult mammalian species. In these areas, exceptionally, the problems caused by new neurons during integration in the network might be overwhelmed by the benefits provided by new neurons. In contrast, in fish, new neurons are constantly produced in various brain regions such as the cerebellum and optic tectum. Adult neurogenesis also occurs in the cultured spinal cord of the gymnotiform fish Sternarchus albifrons, which can regenerate its spinal cord naturally (Anderson and Waxman, 1985). In the present study, BrdU mitotic labeling and immunohistochemistry using antibody against neuronal marker Hu revealed that many neuronal cells were in S phase in the intact goldfish spinal cord. There were only a few cells stained with the TUNEL method, suggesting that the incorporation of BrdU is not caused by
5-HT-immunoreactive cells in the spinal cord were classified morphologically into two distinct types: 5-HT neurons with ovoid to spindle-shaped perikarya of 5–10 m in diameter and some fine processes, and mast cells with round to oval cytoplasm of 5 m in diameter, devoid of any processes (Fig. 6A, B). These cells were frequently observed in the leptomeninges, and occasionally in the substance proper. The number of cells immunoreactive for 5-HT 1 week, 3 weeks, 6 weeks, and 12 weeks after hemisection was counted at the lesion site, and compared with that in the intact spinal cord. The number of 5-HT neurons significantly increased 3 weeks after hemisection, and then continued to increase (Fig. 7). The number of 5-HT⫹ mast cells also increased 3 weeks after hemisection (Fig. 7). In control, no positive staining was observed when primary antibodies were omitted from the reaction. After RDA application to the spinal cord, there were retrogradely labeled descending axons passing through the lesion site. Cell bodies of retrogradely labeled neurons were observed in several brainstem nuclei including the nucleus of the medial longitudinal fasciculus, the nucleus ruber, the nucleus of the lateral lemniscus, the nucleus reticularis superior, the nucleus reticularis medialis, the nucleus reticularis inferior, the nucleus raphes magnus,
Fig. 3. The number of BrdU⫹/Hu⫹ cells at the lesion site in the spinal cord 1 day, 3 days, 1 week, and 2 weeks after hemisection, and 1 day after single BrdU injection, and at the identical site of the intact spinal cord 1 day after BrdU injection (control). Values are means⫾S.E.M. Significant differences are indicated as ** (P⬍0.01) and * (P⬍0.05).
Fig. 2. The number of BrdU⫹ cells at the lesion site of the intact spinal cord (control), and in the spinal cord 1 day, 3 days, 1 week, and 2 weeks after hemisection. Values are means⫾S.E.M. Significant differences are indicated as ** (P⬍0.01) and * (P⬍0.05).
mal lining (Fig. 4D). The number of BrdU⫹ cells in the spinal cord 6 weeks after hemisection was significantly increased (Fig. 5). Many BrdU⫹ cells were also positive for HuC/D (Fig. 4E). The number of BrdU⫹/Hu⫹ cells was also significantly increased as compared with that in the intact fish (Fig. 5). Immunohistochemistry using antibodies against 5-HT revealed that some of the BrdU⫹/Hu⫹ cells were also 5-HT⫹ (Figs. 4F, 5).
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
1137
Fig. 4. (A) Many BrdU-labeled cells were detected mainly in the gray matter in the intact spinal cord, 5 weeks after a single BrdU injection. Dotted line: midline. (B) A cell (arrow) positive for both BrdU (red) and Hu (green) was observed in the intact spinal cord, 5 weeks after a single BrdU injection. (C) Numerous BrdU-labeled cells were observed in the spinal cord, particularly around the lesion site, 6 weeks after the left hemisection and 5 weeks after a single BrdU injection. Scissor mark: the site of the hemisection. Dotted line: midline. (D) Enlarged image of the white box in C. BrdU cells were observed in the ependyma surrounding the central canal (asterisk). (E) Cells (arrows) positive for both BrdU (red) and Hu (green) were observed in the attached lesion site, 6 weeks after the left hemisection and 5 weeks after single BrdU injection. (F) A cell (arrowhead in F1) positive for 5-HT was also positive for both BrdU (red) and Hu (green) (arrowhead in F2). Scale bars⫽100 m in C (applies to A, and C), 100 m in D, and 20 m in F2 (applies to B, E, F1, F2).
apoptosis. Therefore, it is likely that the NPCs are normally present in the spinal cord of goldfish, and that neurogenesis constantly occurs in the intact spinal cord of the adult goldfish. The small number of BrdU⫹/Hu⫹ neurons at the lesion site 5 weeks after the BrdU injection, however, suggests that most of the newly generated neurons could not survive for a long time. This condition is in contrast to
that in urodele amphibians, in which some newborn spinal neurons differentiate into motoneurons (Holder et al., 1991). Spinal hemisection produces a gap between the proximal and distal stumps. The gap is invaded by various cellular elements, forming scar tissue at the lesion site. The results of the present study demonstrated that the
1138
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
Fig. 5. The number of BrdU⫹ cells, BrdU⫹/Hu⫹ cells, and BrdU⫹/Hu⫹/ 5-HT⫹ cells in the lesion site of the intact spinal cord 5 weeks after BrdU injection (control), and in the spinal cord 6 weeks after hemisection and 5 weeks after BrdU injection. Values are means⫾S.E.M. Significant differences are indicated as ** (P⬍0.01), and * (P⬍0.05).
number of newly generated BrdU⫹/Hu⫹ neurons at the lesion site was significantly increased after hemisection, suggesting that neurogenesis was up-regulated in the spinal cord after the lesion. Newly generated neurons might have a role in tissue remodeling at the lesion site. The birthplace of these neurons is not clear in this study. There are few BrdU⫹/Hu⫹ cells in the ependymal zone, suggesting that newly generated neurons migrate rapidly to the lesion site, although they might be born in the ependyma.
Activation of adult neurogenesis following the spinal injury also occurred in the process of caudal spinal cord regeneration after tail amputation in Sternarchus albifrons (Anderson et al., 1994) and in urodele amphibians (Eager and Singer, 1972; Nordlander and Singer, 1978; Benraiss et al., 1999). In these animals, however, neurogenesis is not observed in the tissue remodeling process occurring at more cranial spinal levels after a lesion. Also, in the eel Anguilla anguilla, regenerated neurons are not observed in the gap replacement process after spinal transection (Dervan and Roberts, 2003). Therefore, the results of the present study primarily indicate that adult neurogenesis occurs not only in regeneration after tail amputation, but also in the gap replacement process after spinal lesion. The number of BrdU⫹/Hu⫹ newborn cells and of total BrdU⫹ cells in the present study was significantly higher 7 days after the lesion than at 1 day or 14 days after the lesion, suggesting that the rate of posttraumatic neurogenesis reaches the maximum level nearly 1 week after the lesion, and then declines. Although the precise mechanism for neurogenesis is not clear, several substances endogenously produced after injuries are suggested to stimulate adult neurogenesis in mammals. Growth factors such as fibroblast growth factor (FGF)-2 and vascular endothelial growth factor-B enhance neurogenesis after traumatic or ischemic injuries (Yoshimura et al., 2001; Wang et al., 2007). Heparin-binding epidermal growth factor-like growth factor, insulin-like growth factor-I, and nitric oxide are also suggested to increase neurogenesis (Anderson et al., 2002; Zhu et al., 2003; Jin et al., 2003). There is a possible role of FGF-2 in neurogenesis in tail cord regeneration after amputation. In the salamander, cell proliferation, including neurogenesis, parallels endogenous FGF-2 levels (Zhang et al., 2000). It is possible that growth factors such as FGF-2 are also involved in neurogenesis after spinal lesion in the goldfish.
Fig. 6. (A) Many 5-HT⫹ cells (arrows) were observed in the attached lesion site 6 weeks after hemisection. Scissor mark: the site of the hemisection. Dotted line: midline. (B) 5-HT⫹ neurons (arrows) and a 5-HT⫹ mast cell (arrowhead) were distinct in the lesion site 6 weeks after hemisection. (C) Many regenerated axons retrogradely labeled with RDA (arrowheads) were in close apposition to 5-HT⫹ cells (arrows) in the lesion site. Scale bars⫽100 m in A and C, 20 m in B.
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
1139
Fig. 7. The number of 5-HT⫹ neurons and mast cells in the lesion site of the intact spinal cord (control), and in the spinal cord 1 day, 1 week, 3 weeks, 6 weeks, and 12 weeks after hemisection. Values are means⫾S.E.M. Significant differences as compared with controls are indicated as ** (P⬍0.01), and * (P⬍0.05).
In the present study, a number of newborn neurons produced after hemisection survived for a long time at the lesion site, and some of them expressed 5-HT. Although the precise mechanism for 5-HT expression is also unclear, neurotrophins might play a role in 5-HT synthesis or 5-HT uptake. Brain-derived neurotrophic factor and basic FGF may induce the expression of tryptophan hydroxy-
lase, a synthesizing enzyme of 5-HT, and may enhance 5-HT uptake capacity in cultured raphe neurons (Galter and Unsicker, 1999). S-100 might also enhance 5-HT uptake capacity in cultured serotonergic neurons (Azmitia et al., 1990). Therefore, it is possible that growth factors released in the injured tissue might have a significant role in producing 5-HT neurons.
Fig. 8. (A) Many positive cells were observed in the positive control study of TUNEL method. (B) No TUNEL-positive cells were observed in the intact spinal cord. (C) Only a few TUNEL-positive cells were found in the attached lesion site of the spinal cord 1 week after the hemisection. Scissor mark: the site of the hemisection. (D) Enlarged image of the boxed area in A. Many TUNEL-positive cells (brown) were seen. (E) Enlarged image of the boxed area in C. A TUNEL-positive cell (arrow) was observed. Dotted line in A–C: midline. Scale bars⫽100 m in C (applies to A–C), 20 m in E (applies to D, E).
1140
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
The CNS of fish and urodele amphibians has the potential for spontaneous axonal regeneration after injury, even in the adult stage. In the goldfish, the descending spinal projections from several nuclei of the diencephalon, midbrain, and medulla oblongata regenerate beyond the lesion site within 4 –12 weeks after complete spinal cord transection (Sharma et al., 1993; Takeda et al., 2007). The present retrograde tract tracing study combined with immunohistochemistry demonstrated that 5-HT neurons at the lesion site have a close spatial relationship with RDAlabeled regenerated axons. It is possible that 5-HT neurons have a role in guiding regenerated axons through the lesion site. 5-HT may fine-tune axon guidance by inducing local collapse responses in regenerating axons in the snail Lymnaea stagnalis (Koert et al., 2001). It is also possible that 5-HT neurons are present in the course along which the regenerated axons pass only by chance, and that there is little functional relationship between 5-HT neurons and regenerated axons. Further detailed morphological studies are needed to clarify the relationship between 5-HT neurons and axonal regeneration. Acknowledgments—We are grateful to Mrs. M. Kobayashi for her expert technical aid. This study was supported, in part, by the Yokohama Foundation for Promotion of Medical Sciences.
REFERENCES Alonso JR, Lara J, Vecino E, Coveñas R, Aijón J (1989) Cell proliferation in the olfactory bulb of adult freshwater teleosts. J Anat 163:155–163. Anderson MJ, Waxman SG (1983) Caudal spinal cord of the teleost Sternarchus albifrons resembles regenerating spinal cord. Anat Rec 205:85–92. Anderson MJ, Waxman SG (1985) Neurogenesis in adult vertebrate spinal cord in situ and in vitro: a new model system. Ann N Y Acad Sci 457:213–233. Anderson MJ, Rossetto DL, Lorenz LA (1994) Neuronal differentiation in vitro from precursor cells of regenerating spinal cord of the adult teleosts Apteronotus albifrons. Cell Tissue Res 278:243–248. Anderson MF, Åberg MAI, Nilsson M, Eriksson PS (2002) Insulin-like growth factor-I and neurogenesis in the adult mammalian brain. Dev Brain Res 134:115–122. Aronov S, Aranda G, Behar L, Ginzburg I (2002) Visualization of translated tau protein in the axons of neuronal P19 cells and characterization of tau RNP granules. J Cell Sci 115:3817–3827. Arvidsson A, Collin T, Kirik D, Kokaia Z, Lindvall O (2002) Neuronal replacement from endogenous precursors in the adult brain after stroke. Nat Med 8:963–970. Azmitia EC, Dolan K, Whitaker-Azmitia PM (1990) S-100B but not NGF, EGF, insulin or calmodulin is a CNS serotonergic growth factor. Brain Res 516:354 –356. Becker T, Wullimann MF, Becker CG, Bernhardt RR, Schachner M (1997) Axonal regrowth after spinal cord transection in adult zebrafish. J Comp Neurol 377:577–595. Benraiss A, Arsanto JP, Coulon J, Thouveny Y (1999) Neurogenesis during caudal spinal cord regeneration in adult newts. Dev Genes Evol 209:363–369. Byrd CA, Brunjes PC (2001) Neurogenesis in the olfactory bulb of adult zebrafish. Neuroscience 105:793– 801. Cao QL, Zhang YP, Howard RM, Walters WM, Tsoulfas P, Whittemore SR (2001) Pluripotent stem cells engrafted into the normal or lesioned adult rat spinal cord are restricted to a glial lineage. Exp Neurol 167:48 –58.
Dervan AG, Roberts BL (2003) Reaction of spinal cord central canal cells to cord transection and their contribution to cord regeneration. J Comp Neurol 458:293–306. Eager M, Singer M (1972) The role of ependyma in spinal cord regeneration in the urodele, Triturus. Exp Neurol 37:422– 430. Galter D, Unsicker K (1999) Regulation of the transmitter phenotype of rostral and caudal groups of cultured serotonergic raphe neurons. Neuroscience 88:549 –559. Hanna GFB, Nawar NN, Sharma SC (1998) Regeneration of ascending spinal axons in goldfish. Brain Res 791:235–245. Holder N, Clarke JDW, Stephens N, Wilson SW, Orsi C, Bloomer T, Tonge DA (1991) Continuous growth of the motor system in the axolotl. J Comp Neurol 303:534 –550. Horner PJ, Power AE, Kempermann G, Kuhn HG, Palmer TD, Winkler J, Thal LG, Gage FH (2000) Proliferation and differentiation of progenitor cells throughout the intact adult rat spinal cord. J Neurosci 20:2218 –2228. Ito H, Murakami T (1984) Retinal ganglion cells in two teleost species, Sebastiscus marmoratus and Navodon modestus. J Comp Neurol 229:80 –96. Jin K, Xie L, Childs J, Sun Y, Mao XO, Logvinova A, Greenberg DA (2003) Cerebral neurogenesis is induced by intranasal administration of growth factors. Ann Neurol 53:405– 409. Johns PR (1981) Growth of fish retinas. Am Zool 21:447– 458. Karimi-Abdolrezaee S, Eftekharpour E, Wang J, Morshead CM, Fehlings MG (2006) Delayed transplantation of adult neural precursor cells promotes remyelination and functional neurological recovery after spinal cord injury. J Neurosci 26:3377–3389. Koert CE, Spencer GE, van Minnen J, Li KW, Geraerts WPM, Syed NI, Smit AB, van Kesteren RE (2001) Functional implications of neurotransmitter expression during axonal regeneration: serotonin, but not peptides, auto-regulate axon growth of an identified central neuron. J Neurosci 21:5597–5606. Lauder JM (1990) Ontogeny of the serotonergic system in the rat: serotonin as a developmental signal. Ann NY Acad Sci 660: 297–314. Lauder JM (1993) Neurotransmitters as growth regulatory signals: role of receptors and second messengers. Trends Neurosci 16: 233–240. Levine TD, Gao F, King PH, Andrews LG, Keene JD (1993) Hel-N1: an autoimmune RNA-binding protein with specificity for 3= uridylaterich untranslated regions of growth factor mRNAs. Mol Cell Biol 13:3494 –3504. Liu J, Dalmau J, Szabo A, Rosenfeld M, Huber J, Furneaux H (1995) Paraneoplastic encephalomyelitis antigens bind to the AU-rich elements of mRNA. Neurology 45:544 –550. Lledo P-M, Alonso M, Gubb MS (2006) Adult neurogenesis and functional plasticity in neuronal circuits. Nat Rev Neurosci 7:179 –193. Marusich MF, Furneaux HM, Henion PD, Weston JA (1994) Hu neuronal proteins are expressed in proliferating neurogenic cells. J Neurobiol 25:143–155. McDonald JW, Liu XZ, Qu Y, Liu S, Mickey SK, Turetsky D, Gottlieb DI, Choi DW (1999) Transplanted embryonic stem cells survive, differentiate and promote recovery in injured rat spinal cord. Nat Med 5:1410 –1412. Ming G, Song H (2005) Adult neurogenesis in the mammalian central nervous system. Annu Rev Neurosci 28:223–250. Nordlander RH, Singer M (1978) The role of ependyma in regeneration of the spinal cord of the urodele amphibian tail. J Comp Neurol 180:349 –374. Ogawa Y, Sawamoto K, Miyata T, Miyao S, Watanabe M, Nakamura M, Bregman BS, Koike M, Uchiyama Y, Toyama Y, Okano H (2002) Transplantation of in vitro-expanded fetal neural progenitor cells results in neurogenesis and functional recovery after spinal cord contusion injury in adult rats. J Neurosci Res 69:925–933. Parent JM (2003) Injury-induced neurogenesis in the adult mammalian brain. Neuroscientist 9:261–272.
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141 Pfeifer K, Vroemen K, Blesch A, Weidner N (2004) Adult neural progenitor cells provide a permissive guiding substrate for corticospinal axon growth following spinal cord injury. Eur J Neurosci 20:1695–1704. Ratti A, Fallini C, Cova L, Fantozzi R, Calzarossa C, Zennaro E, Pascale A, Quattrone A, Silani V (2006) A role for the ELAV RNA-binding proteins in neural stem cells: stabilization of Msi1 mRNA. J Cell Sci 119:1442–1452. Raymond PA, Easter Jr SS (1983) Postembryonic growth of the optic tectum in goldfish I. Location of germinal cells and numbers of neurons produced. J Neurosci 3:1077–1091. Selden JR, Dolbeare F, Clair JH, Nichols WW, Miller JE, Kleemeyer KM, Hyland RJ, DeLuca JG (1993) Statistical confirmation that immunofluorescent detection of DNA repair in human fibroblasts by measurement of bromodeoxyuridine incorporation is stoichiometric and sensitive. Cytometry 14:154 –167. Sharma SC, Jadhao AG, Prasada Rao PD (1993) Regeneration of supraspinal projection neurons in the adult goldfish. Brain Res 620:221–228. Szabo A, Dalmau J, Manley G, Rosenfeld M, Wong E, Henson J, Posner JB, Furneaux HM (1991) HuD, a paraneoplastic encephalomyelitis antigen, contains RNA-binding domains and is homologous to Elav and sex-lethal. Cell 67:325–333. Takeda A, Goris RC, Funakoshi K (2007) Regeneration of descending projections to the spinal motor neurons after spinal hemisection in the goldfish. Brain Res 1155:17–23. Vroemen M, Aigner L, Winkler J, Weidner N (2003) Adult neural progenitor cells grafts survive after acute spinal cord injury and integrate along axonal pathways. Eur J Neurosci 18:743–751. Wang Y, Jin K, Mao XO, Xie L, Banwait S, Marti HH, Greenberg DA (2007) VEGF-overexpressing transgenic mice show enhanced
1141
post-ischemic neurogenesis and neuromigration. J Neurosci Res 85:740 –747. Weiss S, Dunne C, Hewson J, Wohl C, Wheatley M, Peterson AC, Reynolds BA (1996) Multipotent CNS stem cells are present in the adult mammalian spinal cord and ventricular neuroaxis. J Neurosci 16:7599 –7609. Yamada H, Miyake T, Kitamura T (1997) Proliferation and differentiation of ependymal cells after transection of the carp spinal cord. Zool Sci 14:331–338. Yamamoto S, Yamamoto N, Kitamura T, Nakamura K, Nakafuku M (2001) Proliferation of parenchymal neural progenitors in response to injury in the adult rat spinal cord. Exp Neurol 172:115–127. Yoshimura S, Takagi Y, Harada J, Teramoto T, Thomas SS, Waeber C, Bakowska JC, Breakefield XO, Moskowitz MA (2001) FGF-2 regulation of neurogenesis in adult hippocampus after brain injury. Proc Natl Acad Sci U S A 98:5874 –5879. Zhang F, Clarke JDW, Ferretti P (2000) FGF-2 up-regulation and proliferation of neural progenitors in the regenerating amphibian spinal cord in vivo. Dev Biol 225:381–391. Zhu DY, Liu SH, Sun HS, Lu YM (2003) Expression of inducible nitric oxide synthase after focal cerebral ischemia stimulates neurogenesis in the adult rodent dentate gyrus. J Neurosci 23:223–229. Zupanc GKH (1999) Neurogenesis, cell death and regeneration in the adult gymnotiform brain. J Exp Biol 202:1435–1446. Zupanc GKH (2001a) A comparative approach towards the understanding of adult neurogenesis. Brain Behav Evol 58:246 –249. Zupanc GKH (2001b) Adult neurogenesis and neuronal regeneration in the central nervous system of teleost fish. Brain Behav Evol 58:250 –275. Zupanc GKH, Ott R (1999) Cell proliferation after lesions in the cerebellum of adult teleost fish: time course, origin, and type of new cells produced. Exp Neurol 160:78 – 87.
(Accepted 14 December 2007)
ADULT NEUROGENESIS WITH 5-HT EXPRESSION IN LESIONED GOLDFISH SPINAL CORD A. TAKEDA, M. NAKANO, R. C. GORIS AND K. FUNAKOSHI*
genitor cells (NPCs) in the SVZ migrate through the rostral migratory stream to the olfactory bulb to differentiate into either granule neurons or periglomerular neurons, and those in the SGZ of the dentate gyrus migrate a short distance to become granule cells. Adult neurogenesis in the SVZ and SGZ is up-regulated by injury and pathologic stimulation in mammals (Parent, 2003). Newborn neurons might replace damaged cells after injury in the striatum, although few newborn neurons survive for a long time (Arvidsson et al., 2002). In the spinal cord, in vitro studies suggest that there are stem cells with self-renewal and multipotential features in mammals (Weiss et al., 1996). In non-pathologic conditions, proliferating cells in the ependymal and parenchymal layers are essentially glial progenitors producing astrocytes or oligodendrocytes (Horner et al., 2000). The proliferative activity of these progenitors is increased in response to spinal cord transection, but few neurons are generated from the endogenous NPCs (Yamamoto et al., 2001). Transplantation of adult-derived NPCs into the spinal cord after traumatic injury also fails to generate neurons in mammals (Cao et al., 2001; Vroemen et al., 2003; Pfeifer et al., 2004; Karimi-Abdolrezaee et al., 2006). The reason for the predominance of glial differentiation is not clear. One possibility might be the lack of an appropriate endogenous environment for neurogenesis in the spinal cord, because in vivo infusion of growth factors induces neurogenesis from endogenous NPCs (Karimi-Abdolrezaee et al., 2006). On the other hand, transplantation of embryonic stem cells or embryonic NPCs also induces differentiation into neurons, as well as into astrocytes and oligodendrocytes (McDonald et al., 1999; Ogawa et al., 2002). Adult neurogenesis is thought to be an evolutionarily conservative trait inherited from a common chordate ancestor (Zupanc, 2001a). In contrast to mammals, anamniotes (fishes and amphibians) have an enormous potential for producing new neurons in the adult CNS. In teleosts, neurogenesis is observed in various regions of the CNS, such as the olfactory bulb (Alonso et al., 1989; Byrd and Brunjes, 2001), optic tectum (Raymond and Easter, 1983), and cerebellum (Zupanc, 1999; Zupanc and Ott, 1999). Retinal ganglion cells in teleosts also continue to increase in number throughout their lives (Johns, 1981; Ito and Murakami, 1984). In the gymnotiform fish Apteronotus leptorhynchus, approximately 0.2% of the total population of the cells in the adult brain are in S phase. Many of these cells, produced in specific proliferation zones at the ventricular surface, migrate within a few weeks to specific target areas (Zupanc, 2001b). In the cerebellum, although
Department of Neuroanatomy, Yokohama City University School of Medicine, 3-9 Fukuura, Kanazawa-ku, Yokohama, Kanagawa, 2360004 Japan
Abstract—In contrast to mammals, spontaneous nerve regeneration occurs in the teleost spinal cord. In the present study, we examined whether neurogenesis is involved in posttraumatic regeneration in the goldfish spinal cord. In intact fish, many spinal cells positive for both a monoclonal neuronal marker (Hu) and bromodeoxyuridine (BrdU) were observed 24 h after i.p. injection of BrdU, suggesting that constant neurogenesis occurs in the goldfish spinal cord. After hemisection of the spinal cord, the number of spinal cells positive for Hu and BrdU was significantly increased around the lesion site. The number of Hu- and BrdU-positive cells reached the maximum level 7 days after hemisection. In intact fish, spinal cells positive for both Hu and BrdU were also observed 5 weeks after BrdU injection, suggesting that newborn neurons survive for a long time. Six weeks after hemisection, the number of surviving Hu- and BrdU-positive cells at the lesion site was significantly increased as compared with that in intact fish, and some of them were also positive for 5-HT. A retrograde tract tracing study showed that the 5-HTⴙ neurons were close to the regenerated axons passing through the lesion site. These results suggest that adult neurogenesis occurs in the goldfish spinal cord, and that neurogenesis is activated by spinal cord lesion. The newly produced neurons survive a long time at the lesion site, and might participate in the repair of injured tissue and in the regeneration of descending long axons beyond the lesion site. © 2008 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: regeneration, CNS, teleost, BrdU.
Adult neurogenesis, a process of producing new neurons throughout life, occurs in discrete regions of the CNS in mammals, i.e. the subventricular zone (SVZ) of the lateral ventricle and the subgranular zone (SGZ) of the dentate gyrus in the hippocampus (Ming and Song, 2005; Lledo et al., 2006). Newborn neurons generated from neural pro*Corresponding author. Tel: ⫹81-45-787-2569; fax: ⫹81-45-782-7251. E-mail address: [email protected] (K. Funakoshi). Abbreviations: BrdU, 5=-bromo-2=deoxyuridine; BSA, bovine serum albumin; DTAF, dichlorotriazinyl amino fluorescein; Elav, embryonic lethal abnormal visual; FGF, fibroblast growth factor; NDS, normal donkey serum; NPC, neural progenitor cell; PB, phosphate buffer; PBS, phosphate-buffered saline; PBST, 0.1 M phosphate-buffered saline containing 0.3% Triton X-100; PFA, paraformaldehyde; RDA, tetramethylrhodamine dextran amine; SGZ, subgranular zone; SNK, Student-Newman-Keuls; SVZ, subventricular zone; TPBS, 1.4 M NaCl, 27 mM KCl, 81 mM Na2HPO4-12H2O, 15 mM KH2PO4, 1% Tween-20; TUNEL, terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling.
0306-4522/08$32.00⫹0.00 © 2008 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2007.10.059
1132
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
half of newborn cells undergo apoptosis, some of the remaining cells differentiate into granule cells (Zupanc, 2001b). Injury induces high proliferative activity of cerebellar neurons, especially in areas close to the lesion. Newborn cells migrate to the lesion site to replace damaged cells, and some of them are incorporated into the cerebellar circuit as granule cells (Zupanc and Ott, 1999). The teleost spinal cord also has the ability to regenerate. In the weakly electric gymnotiform teleost Sternarchus (now Apteronotus) albifrons, the caudal segments of the spinal cord are regenerated after amputation of the tail (Anderson and Waxman, 1983). In this fish, neuronal precursors in the ependymal layer generate new neuroblasts that proliferate and differentiate into neurons in vitro (Anderson et al., 1994). Axonal projections in the spinal cord are also regenerated after mechanical injuries in teleosts. After the gap between the proximal and distal stumps is repaired, both ascending and descending spinal axonal projections regenerate through the lesion site in the goldfish (Sharma et al., 1993; Hanna et al., 1998) and zebrafish (Becker et al., 1997). In the goldfish, ependymal cells in the spinal cord are suggested to have high proliferative activity in response to traumatic injury (Yamada et al., 1997). The contribution of newborn neurons to spinal regeneration in this case, however, is largely unknown. The present study aimed primarily to examine whether neurogenesis is involved in regeneration of the spinal cord after injury. To detect newborn cells, 5=-bromo-2=deoxyuridine (BrdU), a thymidine analog incorporated into newly synthesized DNA in the S phase, was injected intraperitoneally in combination with immunoreactive studies with a neuronal marker, HuC/HuD. Hu proteins including HuC and HuD are RNA-binding proteins of the embryonic lethal abnormal visual (Elav) family, and function early in neuronal differentiation in vertebrates (Marusich et al., 1994). 5-HT is a neurotransmitter that has regulatory effects on cell proliferation within the neuroepithelium during development (Lauder, 1990). 5-HT also stimulates axonal growth during the developmental process (Lauder, 1993). To shed light on the potential role of 5-HT in tissue repair and axonal regeneration, we focused on the expression of 5-HT in neuronal cells in the spontaneous regeneration of the teleost spinal cord.
EXPERIMENTAL PROCEDURES The goldfish, Carassius auratus (body weight 15– 40 g; n⫽47), were obtained commercially (Nomoto Fish Farm Co. Ltd., Yokohama, Japan) and maintained in an aquarium at 25–28 °C. All procedures were performed according to the standards established by the NIH Guide for Care and Use of Laboratory Animals and the protocols were approved by the Yokohama City University Committee for Animal Research. All efforts were made to minimize the number of animals used and their suffering.
Immunohistochemical analysis of adult neurogenesis after spinal hemisection A total of 21 fish were used for fluorescent immunohistochemistry.
1133
Hemisection of the spinal cord Fifteen of the 21 fish were used for hemisection, and six as controls. The fish were deeply anesthetized with 0.02% 3-aminobenzoic acid ethyl ester (MS-222, Sigma Co., St. Louis, MO, USA) in water, and placed on ice. An incision of the dorsal skin and muscles was made in an appropriate region, and the post-temporal bone and vertebra were exposed. The rostral segments of the spinal cord were exposed after removal of the bones, and a frontal hemisection of the left side of the spinal cord was performed 500 m caudal to the first spinal nerve. The hemisection was performed by inserting the blades of a small scissors at right angles to the spinal surface along the posterior median septum. The wound was sutured and sealed with aerosol plastic dressing (Yoshitomi-Seiyaku, Osaka, Japan).
Injection of BrdU The fish (n⫽21) were anesthetized with 0.02% 3-aminobenzic acid ethyl ester (Sigma) in water. BrdU (100 g/g, Sigma) diluted in saline was injected intraperitoneally into intact fish (n⫽3), and hemisected fish 1 h (n⫽3), 2 days (n⫽3), 6 days (n⫽3), and 13 days (n⫽3) after hemisection, and they were allowed to survive 24 h in the aquarium. In addition, to examine the long-term survival of newly generated cells, fish were allowed to survive 5 weeks in the aquarium after BrdU (100 g/g) diluted in saline was injected intraperitoneally into intact fish (n⫽3), and into hemisected fish (n⫽3) 6 days after the operation.
Tissue preparation After various survival times, the fish (n⫽21) were anesthetized with 0.02% MS-222 in water, and perfused transcardially with 25 ml of saline containing 1% heparin, followed by 50 ml of 0.1 M phosphate buffer (PB, pH 7.4) containing 4% paraformaldehyde (PFA). The spinal cord including the hemisectioned level was removed immediately, postfixed in 0.1 M PB containing 4% PFA for 5– 6 h at 4 °C, and left overnight in 0.1 M PB containing 25% sucrose at 4 °C for cryoprotection. The spinal cord was embedded in Tissue-Tek OCT compound (Sakura, Tokyo, Japan), and frozen in liquid nitrogen. The frozen specimens were cut serially into 20-m thick horizontal sections in a cryostat (Microm, Walldorf, Germany), and thaw-mounted on gelatin-coated glass slides.
Multiple fluorescent immunohistochemistry Sections were dried for 1 h at room temperature, postfixed in 0.1 M PB containing 4% PFA for 30 min, and rinsed in 0.1 M phosphate-buffered saline (PBS, pH 7.4). The sections were incubated in 50 mM Tris–HCl (pH 8.0) for 30 min at 95 °C to activate the antigens and denature the DNA, and immediately cooled to room temperature. The sections were incubated with 0.1 M phosphate-buffered saline containing 0.3% Triton X-100 (0.1 M PBST, pH 7.4) for 10 min, and then incubated overnight at 4 °C in a moist chamber with a mixture of primary antibodies diluted with 1% normal donkey serum (NDS), 0.2% bovine serum albumin (BSA), and 0.1% NaN3 in 0.1 M PBST. Primary antibodies were used at the following dilution: sheep polyclonal anti-BrdU (25 g/ml, Exalpha Biologicals, Boston, MA, USA), mouse monoclonal antiHuC/D (20 g/ml, 16A11, Molecular Probes, Eugene, OR, USA), rabbit polyclonal anti-5-HT (1:1000, Sigma). Mouse monoclonal antibody against HuC/D (16A11, Molecular Probes) recognized the Elav family members HuC, HuD, and Hel-N1 (Szabo et al., 1991; Levine et al., 1993; Liu et al., 1995), which are all neuronal proteins expressed in both the CNS and peripheral nervous system of a variety of vertebrates, including the zebrafish (Byrd and Brunjes, 2001). After rinsing with PBST several times, the sections were then incubated for 3 h at room temperature with a mixture of secondary
1134
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
antibodies diluted with 1% NDS, 0.2% BSA, and 0.1% NaN3 in 0.1 M PBST. Secondary antibodies were used at the following dilution: dichlorotriazinyl amino fluorescein (DTAF) – conjugated donkey antimouse IgG (1:100; Chemicon), rhodamine (TRITC) -conjugated donkey anti-sheep IgG (1:100; Chemicon), and aminometylcoumarin (AMCA) -conjugated donkey anti-rabbit IgG (1:100; Chemicon). After rinsing with PBS several times, the sections were mounted with antifade reagent (SlowFade Gold; Molecular Probes). For quantitative analysis of labeled cells, we defined the lesion site as the spinal area extending from 0.5 mm rostral to 0.5 mm caudal to the incision level, because this was the size of the wound immediately after the incision. For controls, the spinal area from the level of the first spinal nerve to 1 mm caudal to the first spinal nerve was examined, because this was an area identical to that of the lesion site.
Observation and electronic imaging All the sections were analyzed under an epifluorescence microscope (Leica DMR; Leica, Wetzlar, Germany), and imaged with a CCD camera (Leica DC 200; Leica). Primary images of immunofluorescent reactivity were captured using DC Viewer Software (Leica). Contrast, brightness, and color adjustment were performed with Adobe Photoshop Software (Adobe, San Jose, CA, USA). For quantitative analysis, the number of labeled cells in which the nuclear contour could be distinguished was counted in each series of sections for immunofluorescent studies. The differences in the number of BrdU⫹ cells, Hu⫹/BrdU⫹ cells, Hu⫹/ BrdU⫹/5-HT⫹ cells in the lesion site after different survival times were analyzed with the Student-Newman-Keuls (SNK) test.
Western blotting An intact fish was used for verification of mouse monoclonal antibody against HuC/D (16A11, Molecular Probe). The spinal cord was dissected out from a deeply anesthetized goldfish, and 50 mg of the cord was homogenized in 500 m of blending buffer (Complete Lysis-M, Roche, Mannheim, Germany). Insoluble materials were removed by centrifugation at 12,000⫻g for 20 min at 4 °C. The supernatant was diluted six times with a buffer containing 125 mM Tris–HCl (pH 6.8), 10% 2-metacaptoethanol, 4% sodium dodecyl sulfate, 20% glycerol, 0.04% Bromophenol Blue, and heated for 5 min at 95 °C. The supernatant and molecular weight marker (Precision Plus Protein Standards, Bio-Rad, Hercules, CA, USA) were electrophoresed on polyacrylamide gel (e-PAGEL 10%, Atto, Tokyo, Japan). For immunoblotting, the separated proteins were transferred onto a nitrocellulose membrane (Protein BA85 Cellulosenitrat, Schlieicher & Schuell, Dassel, Germany) for 12 h at 4 °C at a constant voltage of 12 V. The membranes were then cut into strips, blocked with skim milk– TPBS (1.4 M NaCl, 27 mM KCl, 81 mM Na2HPO4-12H2O, 15 mM KH2PO4, 1% Tween-20) for 1 h, and incubated with TPBS containing mouse monoclonal antibody against HuC/D (5 g/ml, 16A11, Molecular Probes) overnight at 4 °C, followed by alkaline phosphate-conjugated anti-mouse IgG (1:15,000; Sigma) for 1 h at room temperature. The antibody was detected with nitro blue tetrazolium chloride/5-bromo-4-chloro-3-indolyl phosphate solution (Sigma).
Quantitative analysis of 5-HT cells Another 18 fish, i.e. control fish (n⫽3), and hemisected fish 1 day (n⫽3), 1 week (n⫽3), 3 weeks (n⫽3), 6 weeks (n⫽3), and 12 weeks (n⫽3) after hemisection of the spinal cord, were used for quantitative evaluation of 5-HT⫹ cells around the lesion site. After brief incubation with PBST, the sections were incubated overnight at 4 °C in a moist chamber with rabbit polyclonal anti5-HT (1:1000, Sigma). After rinsing with PBST, the sections were then incubated for 3 h at room temperature with donkey anti-rabbit
IgG (1:100; Chemicon). The primary and secondary antibodies were diluted with 1% NDS, 0.2% BSA, and 0.1% NaN3 in 0.1 M PBST. After rinsing with PBS, the sections were then incubated for 90 min at room temperature with rabbit peroxidase anti-peroxidase complex (1:100; Jackson Immunoresearch, West Grove, PA, USA) diluted with 1% NDS, 0.2% BSA, and 0.1% thimerosal in 0.1 M PBS. The peroxidase reactivity was demonstrated with 3,3=-diaminobenzidine (Sigma). The sections were then dehydrated in an ethanol series, cleared with xylene, and mounted with balsam. For quantitative analysis of labeled cells, we defined the lesion site as the spinal area extending from 0.5 mm rostral to 0.5 mm caudal to the incision. The differences in the number of 5-HT⫹ cells in the lesion site after different survival times were analyzed with the SNK test.
Combination of retrograde tract tracing and 5-HT immunohistochemistry To examine the spatial relationship between 5-HT cells and regenerating axons in the repaired tissue, we performed a combination of immunohistochemistry and retrograde tract tracing. First, the fish (n⫽2) were deeply anesthetized with 0.02% MS-222 (Sigma) in water and placed on ice. The spinal cord was exposed by an incision in the appropriate region of the dorsal skin and removal of the vertebrae, and frontal hemisection of the left side of the spinal cord was performed as described above. Six weeks after hemisection, the spinal cord was again exposed, and a strip of paper soaked in tetramethylrhodamine dextran amine (RDA; molecular weight⫽3000; 100 mg/ml; Molecular Probes) in PBS (0.1 M; pH 7.4) was applied to the proximal stump of the left side of the spinal cord at the level of the second spinal segment. After a survival period of 3– 4 days, the fish were re-anesthetized and perfused transcardially with saline containing 1% heparin, followed by 0.1 M PB (pH 7.4) containing 4% PFA. After postfixation and cryoprotection, the spinal cord including the lesion site was embedded in Tissue-Tek OCT compound and frozen. Then frontal sections were cut serially at 20 m in a cryostat, and were thaw-mounted on gelatin-coated glass slides. After brief incubation with PBST, the sections were incubated overnight at 4 °C in a moist chamber with rabbit polyclonal anti5-HT (1:1000, Sigma). After rinsing with PBST, the sections were incubated for 3 h at room temperature with DTAF-conjugated donkey anti-rabbit IgG (1:100; Chemicon). The primary and secondary antibodies were diluted with 1% NDS, 0.2% BSA, and 0.1% NaN3 in 0.1 M PBST. After rinsing with PBS, the sections were mounted with antifade reagent.
TUNEL method Because the incorporation of thymidine analog into DNA occurs not only during proliferative processes, but also during repair and apoptotic processes (Selden et al., 1993), demonstration of neurogenesis after brain injury requires evidence of the absence of apoptotic markers. To evaluate whether or how many neurons in the goldfish spinal cord undergo apoptosis after injury, horizontal sections from intact fish (n⫽1), and hemisected fish 1 day (n⫽1), 3 days (n⫽3), and 7 days (n⫽1) after hemisection were stained using the terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) method according to the manufacturer’s instructions (NeuroTACS II in situ Apoptosis Detection Kit, Trevigen, Gaithersburg, MD, USA).
RESULTS BrdU and Hu labeling after short-term survival In the intact spinal cord, BrdU-immunoreactive cells were observed in the gray matter 24 h after the BrdU injection (Fig. 1A). With immunohistochemistry using antibodies
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
1135
Fig. 1. (A) Many BrdU-labeled cells were detected in the intact spinal cord 1 day after a single BrdU injection. The majority of BrdU-positive cells were observed in the median and paramedian zones of the gray matter in the transverse sections. Dotted line: midline. (B) A cell (arrow) positive for both BrdU (red) and Hu (green) was observed in the intact spinal cord, 1 day after a single BrdU injection. (C) Numerous BrdU-labeled cells were observed in the spinal cord, particularly around the lesion site, 1 week after the left hemisection and 1 day after a single BrdU injection. Scissor mark: the site of the hemisection. Dotted line: midline. (D) Enlarged image of the white box in C. BrdU cells were observed in the ependyma surrounding the central canal (asterisk). (E) A cell (arrow) positive for both BrdU (red) and Hu (green) was observed in the spinal cord, 1 week after the left hemisection and 1 day after the BrdU injection. Scale bars⫽100 m in C (applies to A, C), 100 m in D, and 20 m in E (applies to B, E). (F) Western blotting of the intact spinal cord with anti-HuC/D. An immunoreactive band at 42– 43 kDa (arrowhead) was detected.
against HuC/D, approximately 35% of BrdU-labeled cells were also positive for HuC/D. BrdU⫹/Hu⫹-immunoreactive cells were scattered in the gray matter, not restricted to the region around the central canal. Huimmunoreactivity was observed in the cytoplasm of the cells, whereas BrdU-immunoreactivity was observed in the nucleus of the cells (Fig. 1B). In the hemisected spinal cord, numerous BrdU-labeled cells were observed 24 h after the BrdU injection (Fig. 1C). There were BrdU⫹ cells in the ependymal lining (Fig. 1D), but many labeled cells were also observed in the gray matter, particularly around the lesion site. The number of BrdU⫹ cells in the spinal cord was significantly increased. The maximum number of BrdU⫹ cells was recorded 1 week after hemisection, and after that the number of BrdU⫹ cells decreased (Fig. 2). There were many BrdU⫹ cells positive for HuC/D in the gray matter of the hemisected spinal cord (Fig. 1E), but not in the ependymal lining. The number of BrdU⫹/Hu⫹ cells was also significantly increased after hemisection, with the maximum level occurring 1 week after the hemisection (Fig. 3).
In control, no positive staining was observed when primary antibodies (anti-BrdU and anti-HuC/D) were omitted from the reaction. Western blotting In the sample of intact spinal cord, the anti-HuC/D reacted with a protein of 42– 43 kDa (Fig. 1F), which has been reported as the molecular weight of HuD (Szabo et al., 1991; Aronov et al., 2002; Ratti et al., 2006), suggesting that the anti-HuC/D specifically recognized the Hu protein in our study. BrdU and Hu labeling after long-term survival In the intact spinal cord, BrdU-immunoreactive cells were observed in the gray matter (Fig. 4A) and in the leptomeninges 5 weeks after the BrdU injection. A number of BrdU⫹ cells in the gray matter were also positive for HuC/D (Fig. 4B). Six weeks after spinal hemisection and 5 weeks after the BrdU injection, many BrdU-labeled cells were observed in the spinal cord, particularly around the lesion site (Fig. 4C). There were also BrdU⫹ cells in the ependy-
1136
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
the nucleus raphes pallidus and obscurus, and the caudal dorsomedial medullary group, and in the diencephalon nuclei including the posterior tuberal nucleus. Except for the descending axons, anterogradely labeled ascending axons might be present in the lesion site. The regenerated axons were in close apposition to 5-HT neurons (Fig. 6C). Apoptosis after spinal hemisection No cells were labeled by the TUNEL method in the spinal cord of intact fish (Fig. 8B). A few TUNEL-positive cells were seen around the lesion site of the spinal cord 1, 3, and 7 weeks after hemisection (Fig. 8C, E).
DISCUSSION
Increased number of serotonergic cells at the lesion site
Adult neurogenesis is thought to be an evolutionarily conservative trait inherited from a common chordate ancestor (Zupanc, 2001a). The ability for adult neurogenesis, however, was remarkably reduced in the branch leading to the mammals, i.e. neurogenesis seems restricted to the olfactory bulb and dentate gyrus in adult mammalian species. In these areas, exceptionally, the problems caused by new neurons during integration in the network might be overwhelmed by the benefits provided by new neurons. In contrast, in fish, new neurons are constantly produced in various brain regions such as the cerebellum and optic tectum. Adult neurogenesis also occurs in the cultured spinal cord of the gymnotiform fish Sternarchus albifrons, which can regenerate its spinal cord naturally (Anderson and Waxman, 1985). In the present study, BrdU mitotic labeling and immunohistochemistry using antibody against neuronal marker Hu revealed that many neuronal cells were in S phase in the intact goldfish spinal cord. There were only a few cells stained with the TUNEL method, suggesting that the incorporation of BrdU is not caused by
5-HT-immunoreactive cells in the spinal cord were classified morphologically into two distinct types: 5-HT neurons with ovoid to spindle-shaped perikarya of 5–10 m in diameter and some fine processes, and mast cells with round to oval cytoplasm of 5 m in diameter, devoid of any processes (Fig. 6A, B). These cells were frequently observed in the leptomeninges, and occasionally in the substance proper. The number of cells immunoreactive for 5-HT 1 week, 3 weeks, 6 weeks, and 12 weeks after hemisection was counted at the lesion site, and compared with that in the intact spinal cord. The number of 5-HT neurons significantly increased 3 weeks after hemisection, and then continued to increase (Fig. 7). The number of 5-HT⫹ mast cells also increased 3 weeks after hemisection (Fig. 7). In control, no positive staining was observed when primary antibodies were omitted from the reaction. After RDA application to the spinal cord, there were retrogradely labeled descending axons passing through the lesion site. Cell bodies of retrogradely labeled neurons were observed in several brainstem nuclei including the nucleus of the medial longitudinal fasciculus, the nucleus ruber, the nucleus of the lateral lemniscus, the nucleus reticularis superior, the nucleus reticularis medialis, the nucleus reticularis inferior, the nucleus raphes magnus,
Fig. 3. The number of BrdU⫹/Hu⫹ cells at the lesion site in the spinal cord 1 day, 3 days, 1 week, and 2 weeks after hemisection, and 1 day after single BrdU injection, and at the identical site of the intact spinal cord 1 day after BrdU injection (control). Values are means⫾S.E.M. Significant differences are indicated as ** (P⬍0.01) and * (P⬍0.05).
Fig. 2. The number of BrdU⫹ cells at the lesion site of the intact spinal cord (control), and in the spinal cord 1 day, 3 days, 1 week, and 2 weeks after hemisection. Values are means⫾S.E.M. Significant differences are indicated as ** (P⬍0.01) and * (P⬍0.05).
mal lining (Fig. 4D). The number of BrdU⫹ cells in the spinal cord 6 weeks after hemisection was significantly increased (Fig. 5). Many BrdU⫹ cells were also positive for HuC/D (Fig. 4E). The number of BrdU⫹/Hu⫹ cells was also significantly increased as compared with that in the intact fish (Fig. 5). Immunohistochemistry using antibodies against 5-HT revealed that some of the BrdU⫹/Hu⫹ cells were also 5-HT⫹ (Figs. 4F, 5).
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
1137
Fig. 4. (A) Many BrdU-labeled cells were detected mainly in the gray matter in the intact spinal cord, 5 weeks after a single BrdU injection. Dotted line: midline. (B) A cell (arrow) positive for both BrdU (red) and Hu (green) was observed in the intact spinal cord, 5 weeks after a single BrdU injection. (C) Numerous BrdU-labeled cells were observed in the spinal cord, particularly around the lesion site, 6 weeks after the left hemisection and 5 weeks after a single BrdU injection. Scissor mark: the site of the hemisection. Dotted line: midline. (D) Enlarged image of the white box in C. BrdU cells were observed in the ependyma surrounding the central canal (asterisk). (E) Cells (arrows) positive for both BrdU (red) and Hu (green) were observed in the attached lesion site, 6 weeks after the left hemisection and 5 weeks after single BrdU injection. (F) A cell (arrowhead in F1) positive for 5-HT was also positive for both BrdU (red) and Hu (green) (arrowhead in F2). Scale bars⫽100 m in C (applies to A, and C), 100 m in D, and 20 m in F2 (applies to B, E, F1, F2).
apoptosis. Therefore, it is likely that the NPCs are normally present in the spinal cord of goldfish, and that neurogenesis constantly occurs in the intact spinal cord of the adult goldfish. The small number of BrdU⫹/Hu⫹ neurons at the lesion site 5 weeks after the BrdU injection, however, suggests that most of the newly generated neurons could not survive for a long time. This condition is in contrast to
that in urodele amphibians, in which some newborn spinal neurons differentiate into motoneurons (Holder et al., 1991). Spinal hemisection produces a gap between the proximal and distal stumps. The gap is invaded by various cellular elements, forming scar tissue at the lesion site. The results of the present study demonstrated that the
1138
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
Fig. 5. The number of BrdU⫹ cells, BrdU⫹/Hu⫹ cells, and BrdU⫹/Hu⫹/ 5-HT⫹ cells in the lesion site of the intact spinal cord 5 weeks after BrdU injection (control), and in the spinal cord 6 weeks after hemisection and 5 weeks after BrdU injection. Values are means⫾S.E.M. Significant differences are indicated as ** (P⬍0.01), and * (P⬍0.05).
number of newly generated BrdU⫹/Hu⫹ neurons at the lesion site was significantly increased after hemisection, suggesting that neurogenesis was up-regulated in the spinal cord after the lesion. Newly generated neurons might have a role in tissue remodeling at the lesion site. The birthplace of these neurons is not clear in this study. There are few BrdU⫹/Hu⫹ cells in the ependymal zone, suggesting that newly generated neurons migrate rapidly to the lesion site, although they might be born in the ependyma.
Activation of adult neurogenesis following the spinal injury also occurred in the process of caudal spinal cord regeneration after tail amputation in Sternarchus albifrons (Anderson et al., 1994) and in urodele amphibians (Eager and Singer, 1972; Nordlander and Singer, 1978; Benraiss et al., 1999). In these animals, however, neurogenesis is not observed in the tissue remodeling process occurring at more cranial spinal levels after a lesion. Also, in the eel Anguilla anguilla, regenerated neurons are not observed in the gap replacement process after spinal transection (Dervan and Roberts, 2003). Therefore, the results of the present study primarily indicate that adult neurogenesis occurs not only in regeneration after tail amputation, but also in the gap replacement process after spinal lesion. The number of BrdU⫹/Hu⫹ newborn cells and of total BrdU⫹ cells in the present study was significantly higher 7 days after the lesion than at 1 day or 14 days after the lesion, suggesting that the rate of posttraumatic neurogenesis reaches the maximum level nearly 1 week after the lesion, and then declines. Although the precise mechanism for neurogenesis is not clear, several substances endogenously produced after injuries are suggested to stimulate adult neurogenesis in mammals. Growth factors such as fibroblast growth factor (FGF)-2 and vascular endothelial growth factor-B enhance neurogenesis after traumatic or ischemic injuries (Yoshimura et al., 2001; Wang et al., 2007). Heparin-binding epidermal growth factor-like growth factor, insulin-like growth factor-I, and nitric oxide are also suggested to increase neurogenesis (Anderson et al., 2002; Zhu et al., 2003; Jin et al., 2003). There is a possible role of FGF-2 in neurogenesis in tail cord regeneration after amputation. In the salamander, cell proliferation, including neurogenesis, parallels endogenous FGF-2 levels (Zhang et al., 2000). It is possible that growth factors such as FGF-2 are also involved in neurogenesis after spinal lesion in the goldfish.
Fig. 6. (A) Many 5-HT⫹ cells (arrows) were observed in the attached lesion site 6 weeks after hemisection. Scissor mark: the site of the hemisection. Dotted line: midline. (B) 5-HT⫹ neurons (arrows) and a 5-HT⫹ mast cell (arrowhead) were distinct in the lesion site 6 weeks after hemisection. (C) Many regenerated axons retrogradely labeled with RDA (arrowheads) were in close apposition to 5-HT⫹ cells (arrows) in the lesion site. Scale bars⫽100 m in A and C, 20 m in B.
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
1139
Fig. 7. The number of 5-HT⫹ neurons and mast cells in the lesion site of the intact spinal cord (control), and in the spinal cord 1 day, 1 week, 3 weeks, 6 weeks, and 12 weeks after hemisection. Values are means⫾S.E.M. Significant differences as compared with controls are indicated as ** (P⬍0.01), and * (P⬍0.05).
In the present study, a number of newborn neurons produced after hemisection survived for a long time at the lesion site, and some of them expressed 5-HT. Although the precise mechanism for 5-HT expression is also unclear, neurotrophins might play a role in 5-HT synthesis or 5-HT uptake. Brain-derived neurotrophic factor and basic FGF may induce the expression of tryptophan hydroxy-
lase, a synthesizing enzyme of 5-HT, and may enhance 5-HT uptake capacity in cultured raphe neurons (Galter and Unsicker, 1999). S-100 might also enhance 5-HT uptake capacity in cultured serotonergic neurons (Azmitia et al., 1990). Therefore, it is possible that growth factors released in the injured tissue might have a significant role in producing 5-HT neurons.
Fig. 8. (A) Many positive cells were observed in the positive control study of TUNEL method. (B) No TUNEL-positive cells were observed in the intact spinal cord. (C) Only a few TUNEL-positive cells were found in the attached lesion site of the spinal cord 1 week after the hemisection. Scissor mark: the site of the hemisection. (D) Enlarged image of the boxed area in A. Many TUNEL-positive cells (brown) were seen. (E) Enlarged image of the boxed area in C. A TUNEL-positive cell (arrow) was observed. Dotted line in A–C: midline. Scale bars⫽100 m in C (applies to A–C), 20 m in E (applies to D, E).
1140
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141
The CNS of fish and urodele amphibians has the potential for spontaneous axonal regeneration after injury, even in the adult stage. In the goldfish, the descending spinal projections from several nuclei of the diencephalon, midbrain, and medulla oblongata regenerate beyond the lesion site within 4 –12 weeks after complete spinal cord transection (Sharma et al., 1993; Takeda et al., 2007). The present retrograde tract tracing study combined with immunohistochemistry demonstrated that 5-HT neurons at the lesion site have a close spatial relationship with RDAlabeled regenerated axons. It is possible that 5-HT neurons have a role in guiding regenerated axons through the lesion site. 5-HT may fine-tune axon guidance by inducing local collapse responses in regenerating axons in the snail Lymnaea stagnalis (Koert et al., 2001). It is also possible that 5-HT neurons are present in the course along which the regenerated axons pass only by chance, and that there is little functional relationship between 5-HT neurons and regenerated axons. Further detailed morphological studies are needed to clarify the relationship between 5-HT neurons and axonal regeneration. Acknowledgments—We are grateful to Mrs. M. Kobayashi for her expert technical aid. This study was supported, in part, by the Yokohama Foundation for Promotion of Medical Sciences.
REFERENCES Alonso JR, Lara J, Vecino E, Coveñas R, Aijón J (1989) Cell proliferation in the olfactory bulb of adult freshwater teleosts. J Anat 163:155–163. Anderson MJ, Waxman SG (1983) Caudal spinal cord of the teleost Sternarchus albifrons resembles regenerating spinal cord. Anat Rec 205:85–92. Anderson MJ, Waxman SG (1985) Neurogenesis in adult vertebrate spinal cord in situ and in vitro: a new model system. Ann N Y Acad Sci 457:213–233. Anderson MJ, Rossetto DL, Lorenz LA (1994) Neuronal differentiation in vitro from precursor cells of regenerating spinal cord of the adult teleosts Apteronotus albifrons. Cell Tissue Res 278:243–248. Anderson MF, Åberg MAI, Nilsson M, Eriksson PS (2002) Insulin-like growth factor-I and neurogenesis in the adult mammalian brain. Dev Brain Res 134:115–122. Aronov S, Aranda G, Behar L, Ginzburg I (2002) Visualization of translated tau protein in the axons of neuronal P19 cells and characterization of tau RNP granules. J Cell Sci 115:3817–3827. Arvidsson A, Collin T, Kirik D, Kokaia Z, Lindvall O (2002) Neuronal replacement from endogenous precursors in the adult brain after stroke. Nat Med 8:963–970. Azmitia EC, Dolan K, Whitaker-Azmitia PM (1990) S-100B but not NGF, EGF, insulin or calmodulin is a CNS serotonergic growth factor. Brain Res 516:354 –356. Becker T, Wullimann MF, Becker CG, Bernhardt RR, Schachner M (1997) Axonal regrowth after spinal cord transection in adult zebrafish. J Comp Neurol 377:577–595. Benraiss A, Arsanto JP, Coulon J, Thouveny Y (1999) Neurogenesis during caudal spinal cord regeneration in adult newts. Dev Genes Evol 209:363–369. Byrd CA, Brunjes PC (2001) Neurogenesis in the olfactory bulb of adult zebrafish. Neuroscience 105:793– 801. Cao QL, Zhang YP, Howard RM, Walters WM, Tsoulfas P, Whittemore SR (2001) Pluripotent stem cells engrafted into the normal or lesioned adult rat spinal cord are restricted to a glial lineage. Exp Neurol 167:48 –58.
Dervan AG, Roberts BL (2003) Reaction of spinal cord central canal cells to cord transection and their contribution to cord regeneration. J Comp Neurol 458:293–306. Eager M, Singer M (1972) The role of ependyma in spinal cord regeneration in the urodele, Triturus. Exp Neurol 37:422– 430. Galter D, Unsicker K (1999) Regulation of the transmitter phenotype of rostral and caudal groups of cultured serotonergic raphe neurons. Neuroscience 88:549 –559. Hanna GFB, Nawar NN, Sharma SC (1998) Regeneration of ascending spinal axons in goldfish. Brain Res 791:235–245. Holder N, Clarke JDW, Stephens N, Wilson SW, Orsi C, Bloomer T, Tonge DA (1991) Continuous growth of the motor system in the axolotl. J Comp Neurol 303:534 –550. Horner PJ, Power AE, Kempermann G, Kuhn HG, Palmer TD, Winkler J, Thal LG, Gage FH (2000) Proliferation and differentiation of progenitor cells throughout the intact adult rat spinal cord. J Neurosci 20:2218 –2228. Ito H, Murakami T (1984) Retinal ganglion cells in two teleost species, Sebastiscus marmoratus and Navodon modestus. J Comp Neurol 229:80 –96. Jin K, Xie L, Childs J, Sun Y, Mao XO, Logvinova A, Greenberg DA (2003) Cerebral neurogenesis is induced by intranasal administration of growth factors. Ann Neurol 53:405– 409. Johns PR (1981) Growth of fish retinas. Am Zool 21:447– 458. Karimi-Abdolrezaee S, Eftekharpour E, Wang J, Morshead CM, Fehlings MG (2006) Delayed transplantation of adult neural precursor cells promotes remyelination and functional neurological recovery after spinal cord injury. J Neurosci 26:3377–3389. Koert CE, Spencer GE, van Minnen J, Li KW, Geraerts WPM, Syed NI, Smit AB, van Kesteren RE (2001) Functional implications of neurotransmitter expression during axonal regeneration: serotonin, but not peptides, auto-regulate axon growth of an identified central neuron. J Neurosci 21:5597–5606. Lauder JM (1990) Ontogeny of the serotonergic system in the rat: serotonin as a developmental signal. Ann NY Acad Sci 660: 297–314. Lauder JM (1993) Neurotransmitters as growth regulatory signals: role of receptors and second messengers. Trends Neurosci 16: 233–240. Levine TD, Gao F, King PH, Andrews LG, Keene JD (1993) Hel-N1: an autoimmune RNA-binding protein with specificity for 3= uridylaterich untranslated regions of growth factor mRNAs. Mol Cell Biol 13:3494 –3504. Liu J, Dalmau J, Szabo A, Rosenfeld M, Huber J, Furneaux H (1995) Paraneoplastic encephalomyelitis antigens bind to the AU-rich elements of mRNA. Neurology 45:544 –550. Lledo P-M, Alonso M, Gubb MS (2006) Adult neurogenesis and functional plasticity in neuronal circuits. Nat Rev Neurosci 7:179 –193. Marusich MF, Furneaux HM, Henion PD, Weston JA (1994) Hu neuronal proteins are expressed in proliferating neurogenic cells. J Neurobiol 25:143–155. McDonald JW, Liu XZ, Qu Y, Liu S, Mickey SK, Turetsky D, Gottlieb DI, Choi DW (1999) Transplanted embryonic stem cells survive, differentiate and promote recovery in injured rat spinal cord. Nat Med 5:1410 –1412. Ming G, Song H (2005) Adult neurogenesis in the mammalian central nervous system. Annu Rev Neurosci 28:223–250. Nordlander RH, Singer M (1978) The role of ependyma in regeneration of the spinal cord of the urodele amphibian tail. J Comp Neurol 180:349 –374. Ogawa Y, Sawamoto K, Miyata T, Miyao S, Watanabe M, Nakamura M, Bregman BS, Koike M, Uchiyama Y, Toyama Y, Okano H (2002) Transplantation of in vitro-expanded fetal neural progenitor cells results in neurogenesis and functional recovery after spinal cord contusion injury in adult rats. J Neurosci Res 69:925–933. Parent JM (2003) Injury-induced neurogenesis in the adult mammalian brain. Neuroscientist 9:261–272.
A. Takeda et al. / Neuroscience 151 (2008) 1132–1141 Pfeifer K, Vroemen K, Blesch A, Weidner N (2004) Adult neural progenitor cells provide a permissive guiding substrate for corticospinal axon growth following spinal cord injury. Eur J Neurosci 20:1695–1704. Ratti A, Fallini C, Cova L, Fantozzi R, Calzarossa C, Zennaro E, Pascale A, Quattrone A, Silani V (2006) A role for the ELAV RNA-binding proteins in neural stem cells: stabilization of Msi1 mRNA. J Cell Sci 119:1442–1452. Raymond PA, Easter Jr SS (1983) Postembryonic growth of the optic tectum in goldfish I. Location of germinal cells and numbers of neurons produced. J Neurosci 3:1077–1091. Selden JR, Dolbeare F, Clair JH, Nichols WW, Miller JE, Kleemeyer KM, Hyland RJ, DeLuca JG (1993) Statistical confirmation that immunofluorescent detection of DNA repair in human fibroblasts by measurement of bromodeoxyuridine incorporation is stoichiometric and sensitive. Cytometry 14:154 –167. Sharma SC, Jadhao AG, Prasada Rao PD (1993) Regeneration of supraspinal projection neurons in the adult goldfish. Brain Res 620:221–228. Szabo A, Dalmau J, Manley G, Rosenfeld M, Wong E, Henson J, Posner JB, Furneaux HM (1991) HuD, a paraneoplastic encephalomyelitis antigen, contains RNA-binding domains and is homologous to Elav and sex-lethal. Cell 67:325–333. Takeda A, Goris RC, Funakoshi K (2007) Regeneration of descending projections to the spinal motor neurons after spinal hemisection in the goldfish. Brain Res 1155:17–23. Vroemen M, Aigner L, Winkler J, Weidner N (2003) Adult neural progenitor cells grafts survive after acute spinal cord injury and integrate along axonal pathways. Eur J Neurosci 18:743–751. Wang Y, Jin K, Mao XO, Xie L, Banwait S, Marti HH, Greenberg DA (2007) VEGF-overexpressing transgenic mice show enhanced
1141
post-ischemic neurogenesis and neuromigration. J Neurosci Res 85:740 –747. Weiss S, Dunne C, Hewson J, Wohl C, Wheatley M, Peterson AC, Reynolds BA (1996) Multipotent CNS stem cells are present in the adult mammalian spinal cord and ventricular neuroaxis. J Neurosci 16:7599 –7609. Yamada H, Miyake T, Kitamura T (1997) Proliferation and differentiation of ependymal cells after transection of the carp spinal cord. Zool Sci 14:331–338. Yamamoto S, Yamamoto N, Kitamura T, Nakamura K, Nakafuku M (2001) Proliferation of parenchymal neural progenitors in response to injury in the adult rat spinal cord. Exp Neurol 172:115–127. Yoshimura S, Takagi Y, Harada J, Teramoto T, Thomas SS, Waeber C, Bakowska JC, Breakefield XO, Moskowitz MA (2001) FGF-2 regulation of neurogenesis in adult hippocampus after brain injury. Proc Natl Acad Sci U S A 98:5874 –5879. Zhang F, Clarke JDW, Ferretti P (2000) FGF-2 up-regulation and proliferation of neural progenitors in the regenerating amphibian spinal cord in vivo. Dev Biol 225:381–391. Zhu DY, Liu SH, Sun HS, Lu YM (2003) Expression of inducible nitric oxide synthase after focal cerebral ischemia stimulates neurogenesis in the adult rodent dentate gyrus. J Neurosci 23:223–229. Zupanc GKH (1999) Neurogenesis, cell death and regeneration in the adult gymnotiform brain. J Exp Biol 202:1435–1446. Zupanc GKH (2001a) A comparative approach towards the understanding of adult neurogenesis. Brain Behav Evol 58:246 –249. Zupanc GKH (2001b) Adult neurogenesis and neuronal regeneration in the central nervous system of teleost fish. Brain Behav Evol 58:250 –275. Zupanc GKH, Ott R (1999) Cell proliferation after lesions in the cerebellum of adult teleost fish: time course, origin, and type of new cells produced. Exp Neurol 160:78 – 87.
(Accepted 14 December 2007)