- Email: [email protected]
Mouse nestin protein localizes in growth cones of P19 neurons and cerebellar granule cells
Neuroscience Letters 302 (2001) 89±92
www.elsevier.com/locate/neulet
Mouse nestin protein localizes in growth cones of P19 neurons and cerebellar granule cells Ye Yan, Jing Yang, Wei Bian, Naihe Jing* Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 320 Yue Yang Road, Shanghai 200031, China Received 5 October 2000; received in revised form 20 February 2001; accepted 22 February 2001
Abstract The neuronal growth cone, a highly motile structure at the distal tip of growing axons, contains ®lamentous actin and microtubules as its main cytoskeletal components. Using immunocytochemistry, we observed that nestin, which is the predominant intermediate ®lament protein in neuroepithelial cells and young neurons of the developing brain, appears to be strongly expressed in neurites and growth cones of neurons differentiating from P19 embryonic carcinoma cells in vitro. Double-staining of nestin and microtubule-associated protein-2 as well as nestin and growth-associated protein-43 revealed that nestin protein localizes in neurites and the central regions of growth cones of primary cultures of cerebellar granule cells from postnatal day 6 mice. These results suggest a role for nestin in growth cone guidance during axon elongation. q 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Growth cone; Nestin; Growth-associated protein-43; P19 neurons; Cerebellar granule cells
The neuronal growth cone may be viewed as a signal transduction device for recognizing extracellular guidance signals and translating them into directed neurite growth. Two major cytoskeletal components, ®lamentous actin (Factin) and microtubules, are crucial for mediating neurite outgrowth and axon guidance [15,19,21]. At least two distinct F-actin populations are found in the peripheral cytoplasmic domain of the growth cone: (1) a dense actin meshwork forms the underlying structure of lamellipodia; (2) tight micro®lament bundles span the width of lamellipodia and extend into ®lopodia [2,14]. Microtubules are prominent in the neurite shaft and central domain of the growth cone, and they extend to the membrane of the leading edges and to the base of ®lopodia. Microtubules adopt three characteristic distributions that correlate with growth cone behavior: (1) dispersed and splayed throughout much of the growth cone; (2) looped and apparently contorted by compression; and (3) bundled into tight arrays [20]. The subcellular distribution of neuronal intermediate ®laments (IFs) has been studied for the most part in vitro. For example, neuro®lament-L was detected immunocytochemically in neurites of nerve growth factor induced PC12 * Corresponding author. Tel.: 186-21-64374430; fax: 186-2164338357. E-mail address: [email protected] (N. Jing).
neurons [12], and a-internexin was similarly found in the neuritic processes of cultured granule cell neurons [4]. Other neuronal IFs such as Xenopus neuronal intermediate ®lament protein [3] and tanabin [7] from Xenopus laevis have also been identi®ed. The IF associated earliest with neuronal development is nestin, which is expressed at high levels in neuroectodermal cells that have the potential to develop into both neurons and glias [10,13]. Nestin formed gently curved ®bers and was found in a wave-like pattern within the cytoplasm of neural precursor cell line ST15A [13]. However, the subcellular distribution of neuronal IFs in the growth cone is largely unknown. In the present study, immuno¯uorescence staining was employed to localize nestin in growth cones of neurons derived from P19 embryonic carcinoma (EC) cells and in the central regions of growth cones of cerebellar granule cells in vitro. The mouse EC cell line P19, was cultured as described [17]. To induce neuronal differentiation, P19 cells were allowed to aggregate by plating them in a bacterial-grade Petri Dish (Fisher, USA) at a seeding density of 1 £ 10 5 cells/ml in the presence of 10 26 M all-trans-retinoic acid (RA, Sigma, USA). After 4 days, aggregates were dissociated by trypsin digestion, and cells were grown on polyl-lysine coated (10 mg/ml) tissue culture plates (Corning, USA) or glass coverslips. Cells were seeded at a density of 1 £ 10 4 cells/cm 2 in N2 serum-free medium containing
0304-3940/01/$ - see front matter q 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S03 04 - 394 0( 0 1) 01 66 4- 0
90
Y. Yan et al. / Neuroscience Letters 302 (2001) 89±92
5 mg/ml insulin, 50 mg/ml human transferrin, 20 nM progesterone, 60 mM putrescine, 30 nM sodium selenite and 1 mg/ ml ®bronectin (Gibco, USA). Primary cultures of cerebellar granule cell neurons were prepared as described [6]. Brie¯y, cerebella from postnatal day 6 (P6) ICR mice were dissected and dissociated by trypsin digestion. The resulting cell suspension was seeded onto poly-l-lysine (10 mg/ml) coated glass coverslips in 35 mm diameter culture dishes at a density of 2.5±3.0 £ 10 6 cells/dish. Cells were maintained in Dulbecco's modi®ed Eagle's medium/F12 medium (Gibco) supplemented with 10% fetal bovine serum (FBS, Gibco). After 48 h, the medium was changed to 10% FBS containing 5 mM cytosine arabinoside (Sigma) for 24 h to suppress the growth of non-neuronal cells. Cells grown on the glass coverslips were ®xed in 4% paraformaldehyde in phosphate-buffered saline (PBS) at 48C for 30 min and then in a mixture of 50% methanol, 25% chloroform, and 25% acetone at 2208C for 30 min. Cells on the coverslips were rehydrated sequentially in decreasing concentrations of methanol (95, 75, 50, and 30%) and ®nally in PBS. Non-speci®c binding of antibodies was blocked by incubation of cells in PBS containing 0.5% normal goat serum plus 1% bovine serum albumin for 30 min at room temperature. Working dilutions of primary antibodies were applied to cells at 48C overnight. Three primary antibodies were used: (1) rabbit anti-nestin (1:200, anti-Nestin, [11]); (2) mouse monoclonal anti-microtubule-associated protein 2 (1:200, anti-MAP2; Sigma); (3) mouse monoclonal antigrowth-associated protein-43 (1:200, anti-GAP-43; Sigma). Normal rabbit IgG (1:200, Zymed, USA) and normal mouse IgG (1:200, Zymed) were used as negative controls. After a rinse in PBS, cells on the coverslips were incubated with a secondary antibody (1:200, FITC-conjugated goat anti-rabbit IgG, Tago, USA) for 40 min at room temperature. Double-staining was performed with antiNestin detected by LRSC-conjugated goat anti-rabbit IgG (1:200, Dainova, Germany), and either anti-MAP2 or antiGAP-43 detected by DTAF-conjugated goat anti-mouse IgG (1:200, Dainova). The coverslips were mounted with Mowiol 4±88 media (Hoechst, Germany), and immunoreactivity was visualized by ¯uorescence microscopy. Immunocytochemistry localized the subcellular distribution of nestin protein in P19 EC cells and in RA-induced P19 neurons (Fig. 1). In non-induced cells, nestin immunoreactivity showed a typical cytoskeletal network distribution characteristic of intermediate ®lament proteins (Fig. 1A). After RA-induced neuronal differentiation, P19 cell morphology changed from epithelial-like to neuronal-like with decreased cytoplasmic volume and long neuritic processes. Differentiating P19 neurons tended to form cell-clusters and aggregates. The pattern of nestin immunoreactivity in these cells also changed dramatically from a uniform distribution to a predominantly cytoplasmic localization (Fig. 1B,C). Strong nestin immunoreactivity appeared in the axon hillocks as well as in the neurites and growth cones (Fig. 1B). In cells with long, thin neurites
extending toward their neighbors, nestin immunoreactivity was weak in the neurite, but strong within the paintbrush shaped growth cone (Fig. 1C). These observations suggest that nestin protein may also exist in neuronal growth cones in vivo. Primary cultures of cerebellar granule cells from postnatal day 6 (P6) mice were used to con®rm growth cone localization of nestin protein in neurons. Microtubule-associated protein 2 (MAP2) was used as a neuronal marker (Fig. 2). Granule cells with short neurites displayed multi-®ngered growth cones (Fig.
Fig. 1. Nestin expression in P19 cells. Non-induced P19 cells were stained with the anti-nestin antibody (A). After RA-induction of neuronal differentiation, cells were replated in N2 medium for 2 days and stained with the anti-nestin antibody (B,C). Note that nestin immunoreactivity was evenly distributed in a pattern that was typical for cytoskeletal networks (A). After induction, cells extended long neuritic processes, and nestin immunoreactivity was strong in the growth cones of these neurons (B,C). Bar, 10 mm.
Y. Yan et al. / Neuroscience Letters 302 (2001) 89±92
91
using normal rabbit IgG instead of anti-Nestin and normal mouse IgG instead of anti-MAP2 or anti-GAP-43 were both negative (insets in Fig. 3). Nestin protein was ®rst detected in the endfeet of radially oriented neuroepithelial cells at the pial surface in the neural tube after closure on mouse embryonic day 9 (E9) and rat E11 [10]. These cells span the neural tube, and they are strongly immunoreactive for nestin [5,10,11]. In primary
Fig. 2. Localization of nestin and MAP2 in the growth cones of cerebellar granule cells. Granule cells from P6 mouse cerebella (A) were cultured for 3 days in vitro, and they were doublestained with anti-Nestin (red in B) and anti-MAP2 (green in C). Note that the central region of the growth cone (arrow in A) was positive for both nestin and MAP2. Bar, 6 mm.
2A). Double-staining revealed that cells immunoreactive with anti-Nestin were anti-MAP2 positive and that their immunoreactivities colocalized in the axon hillocks, neurites, and palm-like central region of growth cones (Fig. 2B,C). Furthermore, growth-associated protein-43 (GAP43) was used to visualize the ®ne structure of the growth cone [1] (Fig. 3). Double-staining showed that both nestin and GAP-43 localized in the cell bodies, neurites and growth cones of granule cells (Fig. 3B,C). Nestin immunoreactivity mainly distributed in the central region of growth cones (Fig. 3B), while GAP-43 immunoreactivity appeared not only in the central domain, but also in the peripheral ®nger-like ®lopodia of growth cones (Fig. 3C). Controls
Fig. 3. Localization of nestin and GAP-43 in the growth cones of cerebellar granule cells. Granule cells from P6 mouse cerebella (A) were cultured for 3 days in vitro, and double-stained with anti-Nestin (red in B) and anti-GAP-43 (green in C). Controls were shown as insets, which were stained with normal rabbit IgG (inset in B) or normal mouse IgG (inset in C). Note that nestin immunoreactivity was localized in the central region of the growth cone (arrow in A), while GAP-43 immunoreactivity was distributed not only in the central domain, but also in the peripheral ®nger-like ®lopodia of the growth cone. Control cells were both negative. Bar, 10 mm.
92
Y. Yan et al. / Neuroscience Letters 302 (2001) 89±92
cultures of the neural tube from E10 mouse embryos, neuroepithelial cells contain nestin protein in their cytoplasmic processes but not in ®ne neurites [11]. Evidence of nestin expression in developing mouse brains is well documented, but to our knowledge localization of nestin protein in neuronal growth cones has not been reported previously. Growth cones at the tips of growing axons are highly motile, specialized structures that rapidly extend and retract ®nger-like ®lopodia to search the surrounding environment for directional cues. Cytoskeletal rearrangements also play a functional role during axonal outgrowth. F-actin and microtubules are well known cytoskeletal components of growth cones. Thus far tanabin is the only IF protein known to be in the growth cones of a neuronal subpopulation in Xenopus embryos [7]. In the present study, the evidence indicates that nestin protein localizes in the growth cones of P19 neurons, as well as in the central regions of growth cones of cerebellar granule cells. It should also be noted that tanabin has a short N-terminal and a long C-terminal domain, which is similar to that of nestin. The long carboxyl-terminal domains of neuro®lament-M (NF-M) and -H (NF-H) are thought to form cross-bridge structures [9,16]. Electron microscopy shows that microtubules and neuro®laments are cross-linked, and actin micro®laments form a cortical network just under the surface membrane of axons [8]. Like NF-M and NF-H, the long carboxyl-terminus of nestin and tanabin may serve as cross-bridges or spacers between IF and other cytoskeletal components in the growth cone. Furthermore, ®ve sequence-related proteins in the plakin family are found on IFs and ®lament attachment sites at the plasma membrane [18]. Their ability to form crossbridges between IFs and other cytoskeletal components including actin and microtubules suggests potential functions as regulators of cytoskeletal assembly and IF network integrity. Accordingly, intermediate ®lament-associated proteins may mediate connections between nestin, F-actin, and microtubules during cytoskeletal reorganizations of the growth cone. We thank Robert Shiurba and Steve Kulich for the critical reading of this manuscript. This work was supported by grants from the National Natural Science Foundation of China (39870283, 39930090) and the National Basic Research Program (G1999054000) of China. [1] Benowitz, L.I. and Routtenberg, A., GAP-43: an intrinsic determinant of neuronal development and plasticity, Trends Neurosci., 20 (1997) 84±91. [2] Bridgman, P.C. and Dailey, M.E., The organization of myosin and actin in rapid frozen nerve growth cones, J. Cell Biol., 108 (1989) 95±109. [3] Charnas, L.R., Szaro, B.G. and Gainer, H., Identi®cation and developmental expression of a novel low molecular weight neuronal intermediate ®lament protein expressed in Xenopus laevis, J. Neurosci., 12 (1992) 3010±3024.
[4] Chien, C.L., Mason, C.A. and Liem, R.K.H., a-internexin is the only neuronal intermediate ®lament expressed in developing cerebellar granule neurons, J. Neurobiol., 29 (1996) 304±318. [5] Frederiksen, K. and McKay, R.D.G., Proliferation and differentiation of rat neuroepithelial precursor cells in vivo, J. Neurosci., 8 (1988) 1144±1151. [6] Freshney, R.I., Culture of Animal Cells: A Manual of Basic Technique, Alan R Liss Inc, New York, 1987, pp. 277±279. [7] Hemmati-Brivanlou, A., Mann, R.W. and Harland, R.M., A protein expressed in the growth cones of embryonic vertebrate neurons de®nes a new class of intermediate ®lament protein, Neuron, 9 (1992) 417±428. [8] Hirokawa, N., Cross-linker system between neuro®laments, microtubules, and membranous organelles in frog axons revealed by the quick-freeze, deep-etching method, J. Cell Biol., 94 (1982) 129±142. [9] Hirokawa, N., Glicksman, M.A. and Willard, M.B., Organization of mammalian neuro®lament polypeptides within the neuronal cytoskeleton, J. Cell Biol., 98 (1984) 1523±1536. [10] Hock®eld, S. and McKay, R.D.G., Identi®cation of major cell classes in the developing mammalian nervous system, J. Neurosci., 5 (1985) 3310±3328. [11] Jing, N.H., Kitani, H., Morio, I., Sakakura, T., Tomooka, Y. and Shiurba, R., Expression of intermediate ®lament nestin during mouse brain development, Chin. J. Physiol. Sci., 12 (1996) 1±8. [12] Lee, V.M.Y. and Page, C., The dynamics of nerve growth factor-induced neuro®lament and vimentin ®lament expression and organization in PC12 cells, J. Neurosci., 4 (1984) 1705±1714. [13] Lendahl, U., Zimmerman, L.B. and McKay, R.D.G., CNS stem cells express a new class of intermediate ®lament protein, Cell, 60 (1990) 585±595. [14] Lewis, A.K. and Bridgman, P.C., Nerve growth cone lamellipodia contain two populations of actin ®laments that differ in organization and polarity, J. Cell Biol., 119 (1992) 1219± 1243. [15] Lin, C.H., Thompson, C.A. and Forscher, P., Cytoskeletal reorganization underlying growth cone motility, Curr. Opin. Neurobiol., 4 (1994) 640±647. [16] Nakagawa, T., Chen, J., Zhang, Z., Kanai, Y. and Hirokawa, N., Two distinct functions of the carboxyl-terminal tail domain of NF-M upon neuro®lament assembly: crossbridge formation and longitudinal elongation of ®laments, J. Cell Biol., 129 (1995) 411±429. [17] Rudnicki, M.A. and McBurney, M.W., Cell culture methods and induction of differentiation of embryonal carcinoma cell lines, In E.J. Robertson (Ed.), Teratocarcinomas and Embryonic Stem Cells: A Practical Approach, IRL Press, Washington, DC, 1987, pp. 19±50. [18] Ruhrberg, C. and Watt, F.M., The plakin family: versatile organizers of cytoskeletal architecture, Curr. Opin. Genet. Dev., 7 (1997) 392±397. [19] Suter, D.M. and Forscher, P., An emerging link between cytoskeletal dynamics and cell adhesion molecules in growth cone guidance, Curr. Opin. Neurobiol., 8 (1998) 106±116. [20] Tanaka, E.M. and Kirschner, M.W., Microtubule behavior in the growth cones of living neurons during axon elongation, J. Cell Biol., 115 (1991) 345±363. [21] Tanaka, E. and Sabry, J., Making the connection: cytoskeletal rearrangements during growth cone guidance, Cell, 83 (1995) 171±176.
www.elsevier.com/locate/neulet
Mouse nestin protein localizes in growth cones of P19 neurons and cerebellar granule cells Ye Yan, Jing Yang, Wei Bian, Naihe Jing* Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 320 Yue Yang Road, Shanghai 200031, China Received 5 October 2000; received in revised form 20 February 2001; accepted 22 February 2001
Abstract The neuronal growth cone, a highly motile structure at the distal tip of growing axons, contains ®lamentous actin and microtubules as its main cytoskeletal components. Using immunocytochemistry, we observed that nestin, which is the predominant intermediate ®lament protein in neuroepithelial cells and young neurons of the developing brain, appears to be strongly expressed in neurites and growth cones of neurons differentiating from P19 embryonic carcinoma cells in vitro. Double-staining of nestin and microtubule-associated protein-2 as well as nestin and growth-associated protein-43 revealed that nestin protein localizes in neurites and the central regions of growth cones of primary cultures of cerebellar granule cells from postnatal day 6 mice. These results suggest a role for nestin in growth cone guidance during axon elongation. q 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Growth cone; Nestin; Growth-associated protein-43; P19 neurons; Cerebellar granule cells
The neuronal growth cone may be viewed as a signal transduction device for recognizing extracellular guidance signals and translating them into directed neurite growth. Two major cytoskeletal components, ®lamentous actin (Factin) and microtubules, are crucial for mediating neurite outgrowth and axon guidance [15,19,21]. At least two distinct F-actin populations are found in the peripheral cytoplasmic domain of the growth cone: (1) a dense actin meshwork forms the underlying structure of lamellipodia; (2) tight micro®lament bundles span the width of lamellipodia and extend into ®lopodia [2,14]. Microtubules are prominent in the neurite shaft and central domain of the growth cone, and they extend to the membrane of the leading edges and to the base of ®lopodia. Microtubules adopt three characteristic distributions that correlate with growth cone behavior: (1) dispersed and splayed throughout much of the growth cone; (2) looped and apparently contorted by compression; and (3) bundled into tight arrays [20]. The subcellular distribution of neuronal intermediate ®laments (IFs) has been studied for the most part in vitro. For example, neuro®lament-L was detected immunocytochemically in neurites of nerve growth factor induced PC12 * Corresponding author. Tel.: 186-21-64374430; fax: 186-2164338357. E-mail address: [email protected] (N. Jing).
neurons [12], and a-internexin was similarly found in the neuritic processes of cultured granule cell neurons [4]. Other neuronal IFs such as Xenopus neuronal intermediate ®lament protein [3] and tanabin [7] from Xenopus laevis have also been identi®ed. The IF associated earliest with neuronal development is nestin, which is expressed at high levels in neuroectodermal cells that have the potential to develop into both neurons and glias [10,13]. Nestin formed gently curved ®bers and was found in a wave-like pattern within the cytoplasm of neural precursor cell line ST15A [13]. However, the subcellular distribution of neuronal IFs in the growth cone is largely unknown. In the present study, immuno¯uorescence staining was employed to localize nestin in growth cones of neurons derived from P19 embryonic carcinoma (EC) cells and in the central regions of growth cones of cerebellar granule cells in vitro. The mouse EC cell line P19, was cultured as described [17]. To induce neuronal differentiation, P19 cells were allowed to aggregate by plating them in a bacterial-grade Petri Dish (Fisher, USA) at a seeding density of 1 £ 10 5 cells/ml in the presence of 10 26 M all-trans-retinoic acid (RA, Sigma, USA). After 4 days, aggregates were dissociated by trypsin digestion, and cells were grown on polyl-lysine coated (10 mg/ml) tissue culture plates (Corning, USA) or glass coverslips. Cells were seeded at a density of 1 £ 10 4 cells/cm 2 in N2 serum-free medium containing
0304-3940/01/$ - see front matter q 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S03 04 - 394 0( 0 1) 01 66 4- 0
90
Y. Yan et al. / Neuroscience Letters 302 (2001) 89±92
5 mg/ml insulin, 50 mg/ml human transferrin, 20 nM progesterone, 60 mM putrescine, 30 nM sodium selenite and 1 mg/ ml ®bronectin (Gibco, USA). Primary cultures of cerebellar granule cell neurons were prepared as described [6]. Brie¯y, cerebella from postnatal day 6 (P6) ICR mice were dissected and dissociated by trypsin digestion. The resulting cell suspension was seeded onto poly-l-lysine (10 mg/ml) coated glass coverslips in 35 mm diameter culture dishes at a density of 2.5±3.0 £ 10 6 cells/dish. Cells were maintained in Dulbecco's modi®ed Eagle's medium/F12 medium (Gibco) supplemented with 10% fetal bovine serum (FBS, Gibco). After 48 h, the medium was changed to 10% FBS containing 5 mM cytosine arabinoside (Sigma) for 24 h to suppress the growth of non-neuronal cells. Cells grown on the glass coverslips were ®xed in 4% paraformaldehyde in phosphate-buffered saline (PBS) at 48C for 30 min and then in a mixture of 50% methanol, 25% chloroform, and 25% acetone at 2208C for 30 min. Cells on the coverslips were rehydrated sequentially in decreasing concentrations of methanol (95, 75, 50, and 30%) and ®nally in PBS. Non-speci®c binding of antibodies was blocked by incubation of cells in PBS containing 0.5% normal goat serum plus 1% bovine serum albumin for 30 min at room temperature. Working dilutions of primary antibodies were applied to cells at 48C overnight. Three primary antibodies were used: (1) rabbit anti-nestin (1:200, anti-Nestin, [11]); (2) mouse monoclonal anti-microtubule-associated protein 2 (1:200, anti-MAP2; Sigma); (3) mouse monoclonal antigrowth-associated protein-43 (1:200, anti-GAP-43; Sigma). Normal rabbit IgG (1:200, Zymed, USA) and normal mouse IgG (1:200, Zymed) were used as negative controls. After a rinse in PBS, cells on the coverslips were incubated with a secondary antibody (1:200, FITC-conjugated goat anti-rabbit IgG, Tago, USA) for 40 min at room temperature. Double-staining was performed with antiNestin detected by LRSC-conjugated goat anti-rabbit IgG (1:200, Dainova, Germany), and either anti-MAP2 or antiGAP-43 detected by DTAF-conjugated goat anti-mouse IgG (1:200, Dainova). The coverslips were mounted with Mowiol 4±88 media (Hoechst, Germany), and immunoreactivity was visualized by ¯uorescence microscopy. Immunocytochemistry localized the subcellular distribution of nestin protein in P19 EC cells and in RA-induced P19 neurons (Fig. 1). In non-induced cells, nestin immunoreactivity showed a typical cytoskeletal network distribution characteristic of intermediate ®lament proteins (Fig. 1A). After RA-induced neuronal differentiation, P19 cell morphology changed from epithelial-like to neuronal-like with decreased cytoplasmic volume and long neuritic processes. Differentiating P19 neurons tended to form cell-clusters and aggregates. The pattern of nestin immunoreactivity in these cells also changed dramatically from a uniform distribution to a predominantly cytoplasmic localization (Fig. 1B,C). Strong nestin immunoreactivity appeared in the axon hillocks as well as in the neurites and growth cones (Fig. 1B). In cells with long, thin neurites
extending toward their neighbors, nestin immunoreactivity was weak in the neurite, but strong within the paintbrush shaped growth cone (Fig. 1C). These observations suggest that nestin protein may also exist in neuronal growth cones in vivo. Primary cultures of cerebellar granule cells from postnatal day 6 (P6) mice were used to con®rm growth cone localization of nestin protein in neurons. Microtubule-associated protein 2 (MAP2) was used as a neuronal marker (Fig. 2). Granule cells with short neurites displayed multi-®ngered growth cones (Fig.
Fig. 1. Nestin expression in P19 cells. Non-induced P19 cells were stained with the anti-nestin antibody (A). After RA-induction of neuronal differentiation, cells were replated in N2 medium for 2 days and stained with the anti-nestin antibody (B,C). Note that nestin immunoreactivity was evenly distributed in a pattern that was typical for cytoskeletal networks (A). After induction, cells extended long neuritic processes, and nestin immunoreactivity was strong in the growth cones of these neurons (B,C). Bar, 10 mm.
Y. Yan et al. / Neuroscience Letters 302 (2001) 89±92
91
using normal rabbit IgG instead of anti-Nestin and normal mouse IgG instead of anti-MAP2 or anti-GAP-43 were both negative (insets in Fig. 3). Nestin protein was ®rst detected in the endfeet of radially oriented neuroepithelial cells at the pial surface in the neural tube after closure on mouse embryonic day 9 (E9) and rat E11 [10]. These cells span the neural tube, and they are strongly immunoreactive for nestin [5,10,11]. In primary
Fig. 2. Localization of nestin and MAP2 in the growth cones of cerebellar granule cells. Granule cells from P6 mouse cerebella (A) were cultured for 3 days in vitro, and they were doublestained with anti-Nestin (red in B) and anti-MAP2 (green in C). Note that the central region of the growth cone (arrow in A) was positive for both nestin and MAP2. Bar, 6 mm.
2A). Double-staining revealed that cells immunoreactive with anti-Nestin were anti-MAP2 positive and that their immunoreactivities colocalized in the axon hillocks, neurites, and palm-like central region of growth cones (Fig. 2B,C). Furthermore, growth-associated protein-43 (GAP43) was used to visualize the ®ne structure of the growth cone [1] (Fig. 3). Double-staining showed that both nestin and GAP-43 localized in the cell bodies, neurites and growth cones of granule cells (Fig. 3B,C). Nestin immunoreactivity mainly distributed in the central region of growth cones (Fig. 3B), while GAP-43 immunoreactivity appeared not only in the central domain, but also in the peripheral ®nger-like ®lopodia of growth cones (Fig. 3C). Controls
Fig. 3. Localization of nestin and GAP-43 in the growth cones of cerebellar granule cells. Granule cells from P6 mouse cerebella (A) were cultured for 3 days in vitro, and double-stained with anti-Nestin (red in B) and anti-GAP-43 (green in C). Controls were shown as insets, which were stained with normal rabbit IgG (inset in B) or normal mouse IgG (inset in C). Note that nestin immunoreactivity was localized in the central region of the growth cone (arrow in A), while GAP-43 immunoreactivity was distributed not only in the central domain, but also in the peripheral ®nger-like ®lopodia of the growth cone. Control cells were both negative. Bar, 10 mm.
92
Y. Yan et al. / Neuroscience Letters 302 (2001) 89±92
cultures of the neural tube from E10 mouse embryos, neuroepithelial cells contain nestin protein in their cytoplasmic processes but not in ®ne neurites [11]. Evidence of nestin expression in developing mouse brains is well documented, but to our knowledge localization of nestin protein in neuronal growth cones has not been reported previously. Growth cones at the tips of growing axons are highly motile, specialized structures that rapidly extend and retract ®nger-like ®lopodia to search the surrounding environment for directional cues. Cytoskeletal rearrangements also play a functional role during axonal outgrowth. F-actin and microtubules are well known cytoskeletal components of growth cones. Thus far tanabin is the only IF protein known to be in the growth cones of a neuronal subpopulation in Xenopus embryos [7]. In the present study, the evidence indicates that nestin protein localizes in the growth cones of P19 neurons, as well as in the central regions of growth cones of cerebellar granule cells. It should also be noted that tanabin has a short N-terminal and a long C-terminal domain, which is similar to that of nestin. The long carboxyl-terminal domains of neuro®lament-M (NF-M) and -H (NF-H) are thought to form cross-bridge structures [9,16]. Electron microscopy shows that microtubules and neuro®laments are cross-linked, and actin micro®laments form a cortical network just under the surface membrane of axons [8]. Like NF-M and NF-H, the long carboxyl-terminus of nestin and tanabin may serve as cross-bridges or spacers between IF and other cytoskeletal components in the growth cone. Furthermore, ®ve sequence-related proteins in the plakin family are found on IFs and ®lament attachment sites at the plasma membrane [18]. Their ability to form crossbridges between IFs and other cytoskeletal components including actin and microtubules suggests potential functions as regulators of cytoskeletal assembly and IF network integrity. Accordingly, intermediate ®lament-associated proteins may mediate connections between nestin, F-actin, and microtubules during cytoskeletal reorganizations of the growth cone. We thank Robert Shiurba and Steve Kulich for the critical reading of this manuscript. This work was supported by grants from the National Natural Science Foundation of China (39870283, 39930090) and the National Basic Research Program (G1999054000) of China. [1] Benowitz, L.I. and Routtenberg, A., GAP-43: an intrinsic determinant of neuronal development and plasticity, Trends Neurosci., 20 (1997) 84±91. [2] Bridgman, P.C. and Dailey, M.E., The organization of myosin and actin in rapid frozen nerve growth cones, J. Cell Biol., 108 (1989) 95±109. [3] Charnas, L.R., Szaro, B.G. and Gainer, H., Identi®cation and developmental expression of a novel low molecular weight neuronal intermediate ®lament protein expressed in Xenopus laevis, J. Neurosci., 12 (1992) 3010±3024.
[4] Chien, C.L., Mason, C.A. and Liem, R.K.H., a-internexin is the only neuronal intermediate ®lament expressed in developing cerebellar granule neurons, J. Neurobiol., 29 (1996) 304±318. [5] Frederiksen, K. and McKay, R.D.G., Proliferation and differentiation of rat neuroepithelial precursor cells in vivo, J. Neurosci., 8 (1988) 1144±1151. [6] Freshney, R.I., Culture of Animal Cells: A Manual of Basic Technique, Alan R Liss Inc, New York, 1987, pp. 277±279. [7] Hemmati-Brivanlou, A., Mann, R.W. and Harland, R.M., A protein expressed in the growth cones of embryonic vertebrate neurons de®nes a new class of intermediate ®lament protein, Neuron, 9 (1992) 417±428. [8] Hirokawa, N., Cross-linker system between neuro®laments, microtubules, and membranous organelles in frog axons revealed by the quick-freeze, deep-etching method, J. Cell Biol., 94 (1982) 129±142. [9] Hirokawa, N., Glicksman, M.A. and Willard, M.B., Organization of mammalian neuro®lament polypeptides within the neuronal cytoskeleton, J. Cell Biol., 98 (1984) 1523±1536. [10] Hock®eld, S. and McKay, R.D.G., Identi®cation of major cell classes in the developing mammalian nervous system, J. Neurosci., 5 (1985) 3310±3328. [11] Jing, N.H., Kitani, H., Morio, I., Sakakura, T., Tomooka, Y. and Shiurba, R., Expression of intermediate ®lament nestin during mouse brain development, Chin. J. Physiol. Sci., 12 (1996) 1±8. [12] Lee, V.M.Y. and Page, C., The dynamics of nerve growth factor-induced neuro®lament and vimentin ®lament expression and organization in PC12 cells, J. Neurosci., 4 (1984) 1705±1714. [13] Lendahl, U., Zimmerman, L.B. and McKay, R.D.G., CNS stem cells express a new class of intermediate ®lament protein, Cell, 60 (1990) 585±595. [14] Lewis, A.K. and Bridgman, P.C., Nerve growth cone lamellipodia contain two populations of actin ®laments that differ in organization and polarity, J. Cell Biol., 119 (1992) 1219± 1243. [15] Lin, C.H., Thompson, C.A. and Forscher, P., Cytoskeletal reorganization underlying growth cone motility, Curr. Opin. Neurobiol., 4 (1994) 640±647. [16] Nakagawa, T., Chen, J., Zhang, Z., Kanai, Y. and Hirokawa, N., Two distinct functions of the carboxyl-terminal tail domain of NF-M upon neuro®lament assembly: crossbridge formation and longitudinal elongation of ®laments, J. Cell Biol., 129 (1995) 411±429. [17] Rudnicki, M.A. and McBurney, M.W., Cell culture methods and induction of differentiation of embryonal carcinoma cell lines, In E.J. Robertson (Ed.), Teratocarcinomas and Embryonic Stem Cells: A Practical Approach, IRL Press, Washington, DC, 1987, pp. 19±50. [18] Ruhrberg, C. and Watt, F.M., The plakin family: versatile organizers of cytoskeletal architecture, Curr. Opin. Genet. Dev., 7 (1997) 392±397. [19] Suter, D.M. and Forscher, P., An emerging link between cytoskeletal dynamics and cell adhesion molecules in growth cone guidance, Curr. Opin. Neurobiol., 8 (1998) 106±116. [20] Tanaka, E.M. and Kirschner, M.W., Microtubule behavior in the growth cones of living neurons during axon elongation, J. Cell Biol., 115 (1991) 345±363. [21] Tanaka, E. and Sabry, J., Making the connection: cytoskeletal rearrangements during growth cone guidance, Cell, 83 (1995) 171±176.