SEARCH

SEARCH BY CITATION

Keywords:

  • brain;
  • teleost;
  • cell proliferation;
  • neuron;
  • regeneration;
  • nervous system

Abstract

  1. Top of page
  2. Abstract
  3. RESULTS
  4. DISCUSSION
  5. LITERATURE CITED

Neurogenesis was studied in the cerebellum of adult goldfish, to establish the phenomenon in this popular laboratory animal model. BrdU and proliferating cell nuclear antigen labeling revealed a high rate of cell proliferation within the molecular layer of the cerebellar corpus and valve. Most newborn cells expressed the neuronal marker beta-III-tubulin after 24 hr, supporting the goldfish cerebellum as an excellent paradigm to study vertebrate adult neurogenesis. Anat Rec, 2010. © 2010 Wiley-Liss, Inc.

The phenomenon of adult neurogenesis occurs more widespread and at a much higher rate in teleosts than in mammals and terrestrial vertebrates, and has, therefore, aroused considerable interest in the last two decades (Chapouton et al.,2007; Kaslin et al.,2008; Zupanc,2008). A large amount of information has been gathered principally through studies in zebrafish (Byrd and Brunjes,2001; Zupanc et al.,2005; Grandel et al.,2006; Ampatzis and Dermon,2007;), due to its importance as developmental model organism, and in gymnotiformes such as the brown ghost knifefish (Zupanc and Horschke,1995; Zupanc,1999), owing to the wealth of information on brain morphology and function available for this electrocommunicating species. However, for most physiological and behavioral studies related to adult neurogenesis, neither species is ideal, because the zebrafish brain and especially its neurons are too small for many types of experiment, and most researchers do not have easy access to gymnotiformes. The goldfish, in turn, is a well-established, omnipresent and hardy laboratory animal perfectly suited for most behavioral and (electro-) physiological trials. It has been the preferred teleost model for the study of visual function, and for learning and memory assays like those concerned with the vestibulo-ocular reflex. Although the putative relationship between learning, memory and adult neurogenesis is intriguing, few studies have addressed this topic in the central nervous system, with the notable exception of the retina (Boucher and Hitchcock,1998). The goldfish cerebellum was evaluated here as an accessible model system for studies of vertebrate adult neurogenesis, because this part of the brain is central to motor learning and memory, and arguably the site of most widespread neurogenesis in teleosts (Ampatzis and Dermon, 2007). To that end, BrdU, proliferating cell nuclear antigen (PCNA), and beta-III-tubulin immunohistochemistry were combined to compare the cerebellar neuronal proliferation pattern of the goldfish with that of previously scrutinized species like zebrafish and the brown ghost knifefish.

For this study, eight male spawning-stage comet goldfish (Carassius auratus) were obtained from a local breeder during the spring of 2008, maintained in the laboratory for up to 1 month under a natural light/dark cycle at 20 ± 2°C, and fed twice daily. Females were not analyzed, because possible gender differences were not the aim of this study. The fish were injected intraperitoneally with 50 μL/g of body weight of a 3 mg/mL solution of 5-bromo-2-deoxyuridine (BrdU, Sigma-Aldrich). This dosage of 150 mg/kg body weight has been shown previously to effectively label proliferating cells in zebrafish (Zupanc et al.,2005). After 2 or 24 hr, the animals were sacrificed and their brains were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) for 2 hr and cryoprotected in 30% sucrose in PBS overnight. Sagittal and coronal sections of the cerebellum were incubated in 2 M HCl at 37°C for 30 min and washed in 0.1 M borate buffer, pH 8.5 and subsequently in PBS. The sections were incubated overnight at 4°C in the following primary antibodies: mouse monoclonal anti-BrdU (B2531, Sigma-Aldrich, diluted 1:500), rabbit anti-proliferating cell nuclear antigen (PCNA, sc-7907, Chemicon, diluted 1:500), and rabbit anti-beta-III-tubulin (ab18207, Abcam, 1:1,000). Subsequently, the sections were incubated for 1 hr at 20°C in the respective secondary antibodies, goat anti-rabbit-FITC (Chemicon, diluted 1:1,000), goat anti-mouse-Cy3 (Jackson ImmunoResearch, 1:500) and donkey anti-rabbit-HRP (Jackson ImmunoResearch, 1:1,000). HRP-labeled sections were treated with 0.3% hydrogen peroxide for 30 min before revelation with a commercial diaminobenzidine kit (K2368, Dako). The specificity of the beta-III-tubulin and PCNA immunoreactions was controlled with Western blots according to standard protocols, and omission of the primary antibodies was used to control the specificity of the secondary antisera. Marked cells were counted on photographs of randomly chosen cerebellar sections and compared to the total number of cells in equivalent areas, determined by DAPI staining. Results are expressed as mean percentage of labeled cells ± standard error (SE). All experimental procedures were approved by the bioethics committee of the University of Valparaiso and in accordance with the bioethics regulation of the Chilean Research Council (CONICYT).

RESULTS

  1. Top of page
  2. Abstract
  3. RESULTS
  4. DISCUSSION
  5. LITERATURE CITED

PCNA immunolabeling of coronal sections through the cerebellar corpus and valve revealed that proliferating cells localized to the molecular layer at the border with the granule cell layer (Fig. 1A,H). PCNA-positive cells were generally of large size (14–20-μm soma diameter) and of round shape (Fig. 1B). Isolated labeled cells were also found within the granule cell layer of the cerebellar corpus but always close to the border with the molecular layer. The granule cell layer of the cerebellar valve did not contain any PCNA-positive cells. Two hours after injection of BrdU, postmitotic cells were detected in the molecular layer of the cerebellar corpus and valve (Fig. 1C,J). The majority of BrdU-positive cells localized to the outer part of the molecular layer, and only a few labeled cells were found at the limit with the granule cell layer, which remained devoid of stained cells at this stage. Among the total number of cells visualized by DAPI-staining within the molecular layer, 56 ± 17% and 35 ± 13% (mean ± SE) were BrdU-positive in the cerebellar corpus and valve, respectively. BrdU-labeled nuclei were generally round and intensely stained, but a subgroup had weakly labeled or elongated nuclei. Compared to the cerebellum, the adjacent tectum opticum contained BrdU-positive cells at a much lower density. Twenty-four hours after BrdU administration, the BrdU-positive cells of the cerebellar corpus and valve largely remained within the molecular layer (Fig. 1E,L). Here, they accounted for 69 ± 15% and 49 ± 16% of cells visualized by DAPI-staining, in corpus and valve, respectively. Isolated BrdU-positive cells were now detected within the granule cell layer of the cerebellar valve. Double labeling with an antibody against neuron-specific beta-III-tubulin demonstrated that the vast majority (93 ± 5% in the corpus and 85 ± 7% in the valve) of BrdU-positive cells also expressed beta-III-tubulin after 24 hr, supporting their neuronal lineage (Fig. 1G,N).

thumbnail image

Figure 1. BrdU and PCNA labeling in the cerebellar corpus and valve of the goldfish. A: PCNA immunoreactivity in a coronal section through the cerebellar corpus. Note isolated PCNA-positive cells within the granule cell layer (arrow). Inset: PCNA Western blot. B: Detail of the molecular and granule cell layer border, showing large PCNA-positive ovoid cell bodies. C and D: BrdU immunofluorescence (red) and Nomarski image of a sagittal section through the molecular layer of the cerebellar corpus, 2 hr post-BrdU. For orientation, DAPI-stained sections through cerebellar corpus and valve are shown to the left. E and F: BrdU (red) and neuron-specific beta-III-tubulin (green) labeling of a parasagittal section through the cerebellar corpus, 24 hr post-BrdU. Inset: beta-III-tubulin Western blot. G: Colocalization of BrdU and beta-III-tubulin immunoreactivity (yellow) was widespread (arrow). Inset: Double-labeled cells at higher magnification. H and I: PCNA-labeled coronal sections through the cerebellar valve. Clustered PCNA-positive cells are found near the molecular and granule cell layer border (arrow). J and K: BrdU-immunofluorescence (red) and Nomarski image of a parasagittal section through the cerebellar valve, 2 hr after BrdU injection. Isolated cells in the tectum opticum were also BrdU-positive (arrow). L and M: BrdU (red) and neuron-specific beta-III-tubulin (green) labeling of a parasagittal section through the cerebellar valve, 24 hr post-BrdU. N: Colocalization of BrdU and beta-III-tubulin immunoreactivity (yellow). Inset: varying degrees of co-localization are evident at higher magnification. CCemol, molecular layer of the cerebellar corpus; CCegra, granule cell layer of the cerebellar corpus. VCmol, molecular layer of the cerebellar valve; VCgra, granule cell layer of the cerebellar valve, TeO, tectum opticum. Scale bars: A: 200 μm; B: 30 μm, C, E: 50 μm, H: 300 μm, I, J: 100 μm, L: 50 μm.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. RESULTS
  4. DISCUSSION
  5. LITERATURE CITED

Teleost adult neurogenesis has been studied among others in stickleback (Ekstrom et al.,2001), brown ghost (Zupanc,1999), guppy (Birse et al.,1980), goldfish (Meyer,1978; Sullivan et al.,1997), crucian carp (Margotta et al.,2002), and trout and medaka (Nguyen et al.,2001; Candal et al.,2005a,b). Here, we analyzed cellular proliferation in the cerebellum of male spawning-stage goldfish. Our data show PCNA-labeled dividing cells at the boundary of molecular and granule cell layer, and postmitotic cells in the molecular layer 2 and 24 hr after BrdU injection. Most of these newborn cells expressed beta-III-tubulin after 24 hr, suggesting a neuronal cell fate. These data confirm and extend previous data from other species to the goldfish, an ideal laboratory model for physiological and behavioral experiments. Compared with birds and mammals, cellular proliferation within the nervous system of adult fishes is known to occur at considerably higher levels throughout the animal's life (Alvarez-Buylla and Lois,1995; Eriksson et al.,1998; Dawley et al.,2000; Zikopoulos et al.,2000; Byrd and Brunjes,2001; Ekstrom et al.,2001; Zupanc,2001; Zupanc et al.,2005). New neurons are not only generated in brain structures considered homologous to the mammalian olfactory bulb and hippocampus, but also in dozens of additional areas (Zupanc, 2006). Although the replacement of neurons damaged by injury or disease does generally not occur in the mammalian central nervous system, fishes present an enormous capacity of cellular regeneration of their brain and spinal cord (Meyer et al.,1985; Zupanc et al.,1998; Zupanc and Ott,1999; Zupanc,2008; Zupanc,2009). In teleost adult neurogenesis, newly generated cells may either remain close to their proliferation zone, as occurs in the tectum opticum (Zupanc,2006), or migrate during the postmitotic days and weeks from the proliferation zone to their respective target areas. As shown in gymnotiformes and zebrafish, newborn neurons in the cerebellar corpus and valve migrate over large distances from their specific proliferation zone in the molecular layer to their target area in the granule cell layer (Zupanc et al.,1996; Zupanc et al.,2005). During this migration, the cells are probably guided by the fibers of radial glial cells (Zupanc and Clint,2003). Reaching their target area, ∼ 50% of the young cerebellar cells die through apoptosis (Soutschek and Zupanc,1996), whereas the remaining cells integrate into the cerebellar circuitry and probably survive for the remainder of the fish's life. Most of these newly integrated cells are thought to differentiate into mature neurons, whereas others acquire glial properties (Zupanc et al.,1996; Ott et al.,1997). A new and especially fascinating topic is the putative relationship between adult neurogenesis and environmental factors, such as diet, learning, and exercise (Leuner et al.,2006; Imayoshi et al.,2008; Zhao et al.,2008; Stangl and Thuret,2009; Lucassen et al.,2010). Their high rates of adult neurogenesis make teleosts ideal model systems for the study of neurogenic modulation by environmental factors, and the goldfish is well suited among teleosts for physiological and behavioral studies due to its comparatively large brain and ease of training. Recently, the reproductive pheromone PGF has been linked to neurogenic effects in goldfish, which is the first direct evidence for a modulation of adult neurogenesis by external signals in this species (Chung-Davidson et al.,2008). It will be interesting to see whether neurogenesis in the adult goldfish cerebellum is influenced by behavioral paradigms targeting specific cerebellar functions, such as motion control and spatial cognition (Rodriguez et al.,2005).

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. RESULTS
  4. DISCUSSION
  5. LITERATURE CITED