Regulation of gliogenesis in the central nervous system by the JAK-STAT signaling pathway

Science

Volumn 278, Issue 5337, 1997, Pages 477-483

  Bonni, Azad ^a   Sun, Yi ^a   Nadal Vicens, Mireya ^a   Bhatt, Ami ^a   Frank, David A ^b   Rozovsky, Irina ^c   Stahl, Neil ^d   Yancopoulos, George D ^d   Greenberg, Michael E ^a  

^a BOSTON CHILDREN S HOSPITAL   (United States)

^b HARVARD MEDICAL SCHOOL   (United States)

^c UNIVERSITY OF SOUTHERN CALIFORNIA   (United States)

^d REGENERON PHARMACEUTICALS INC   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

CILIARY NEUROTROPHIC FACTOR; CYTOKINE; MITOGEN ACTIVATED PROTEIN KINASE;

ANIMAL CELL; ARTICLE; ASTROCYTE; BRAIN DEVELOPMENT; CELL DIFFERENTIATION; EMBRYO; GLIA CELL; NONHUMAN; PRIORITY JOURNAL; RAT; REGULATORY MECHANISM; SIGNAL TRANSDUCTION;

ANIMALS; ANTIGENS, CD; ASTROCYTES; CELL DIFFERENTIATION; CELL DIVISION; CELL LINEAGE; CELLS, CULTURED; CEREBRAL CORTEX; CILIARY NEUROTROPHIC FACTOR; CYTOKINE RECEPTOR GP130; DIMERIZATION; DNA-BINDING PROTEINS; GLIAL FIBRILLARY ACIDIC PROTEIN; GROWTH INHIBITORS; INTERLEUKIN-6; JANUS KINASE 1; LEUKEMIA INHIBITORY FACTOR; LEUKEMIA INHIBITORY FACTOR RECEPTOR ALPHA SUBUNIT; LYMPHOKINES; MEMBRANE GLYCOPROTEINS; NERVE GROWTH FACTORS; NERVE TISSUE PROTEINS; PLATELET-DERIVED GROWTH FACTOR; PROTEIN-TYROSINE KINASES; RATS; RECEPTOR PROTEIN-TYROSINE KINASES; RECEPTOR, CILIARY NEUROTROPHIC FACTOR; RECEPTORS, CYTOKINE; RECEPTORS, NERVE GROWTH FACTOR; RECEPTORS, OSM-LIF; SIGNAL TRANSDUCTION; STAT1 TRANSCRIPTION FACTOR; STAT3 TRANSCRIPTION FACTOR; STEM CELLS; TRANS-ACTIVATORS;

ANIMALIA; JANUS; MAMMALIA;

EID: 0030708523     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.278.5337.477     Document Type: Article

References (48)

1. D. L. Turner and C. L. Cepko, Nature 328, 131 (1987).
2. C. Walsh and C. L. Cepko, Science 241, 1342 (1988).
3. R. Wetts and S. E. Fraser, ibid. 239, 1142 (1988).
4. B. P. Williams, J. Read, J. Price, Neuron 7, 685 (1991).
5. A. A. Davis and S. Temple, Nature 372, 263 (1994).
6. B. Barres and M. C. Raff, Neuron 12, 935 (1994).
7. S. K. McConnell, ibid. 15, 761 (1995).
8. R. D. McKay, Science 276, 66 (1997).
9. N. Y. Ip et al., Neuron 10, 89 (1993).
10. Cerebral cortices of E14 or E17 Long Evans rats were dissected and enzymatically dissociated in Hanks' buffered salt solution (HBSS; Gibco) with cysteine-activated papain (10 U/ml). Cortices were then washed with trypsin inhibitor, placed in plating medium, and dissociated into single cells with a 5-ml pipette. E14 cortical cells were plated at 2000 cells per square millimeter in L15 medium [International Chemical and Nuclear (ICN)] supplemented with N2 (Gibco) and bFGF (20 ng/ml; Amgen) on cover slips in 24-well plates that had been coated overnight with poly-D-lysine and laminin. E17 cortical cells were plated at 5000 cells per square millimeter in basal Eagle's medium (BME; Sigma) supplemented with 5% fetal calf serum.
11. U. Lendahl, L. B. Zimmerman, R. D. MacKay, Cell 60, 585 (1990).
12. A. Ghosh and M. E. Greenberg, Neuron 15, 89 (1995).
13. B. P. Williams et al., ibid. 18, 553 (1997).
14. K. K. Johe, T. G. Hazel, T. Muller, M. M. Dugich-Djordjevic, R. D. G. McKay, Genes Dev. 10, 3129 (1996).
15. A. Bonni, Y. Sun, M. Nadal-Vicens, M. E. Greenberg, unpublished observations.
16. The number of astrocytes (see Fig. 1B, lower) was greater in CNTF-treated E17 + 3 DIV cortical cultures than in CNTF-treated E14 + 4 DIV cultures [n = 3, analysis of variance (ANOVA), P < 0.001]. The number of astrocytes in untreated cultures E17 + 3 DIV and E14 + 4 DIV was similar.
17. Cultures were infected at the time of plating with a replication-defective retrovirus containing a β-galactosidase gene (at a dilution that led to 20 to 50 clones per well in a 24-well plate). Cultures were left untreated or treated with CNTF (100 ng/ml) or PDGF (20 ng/ml) starting at E14 + 1 DIV. Three days later, cultures were fixed and subjected to indirect immunofluorescence with a rabbit antiserum to β-galactosidase (1:500; 5′-3′, Inc.), guinea pig antibody to GFAP (anti-GFAP) and the TuJ1 mouse monoclonal antibody that labels neuron-specific β-tubulin (1:5000). Within each clone, precursor cells were identified as double GFAP-negative and TuJ1-negative cells. Neurons were identified as TuJ1-positive cells, and astrocytes were identified as GFAP-positive cells. The numbers of clones that were analyzed in control, CNTF-treated, and PDGF-treated cultures were 358, 442, and 315, respectively. In clones that were classified as precursor clones, 100% of the cells were precursor cells. In clones that were classified as astrocytic, >80% of the cells were astrocytes and the remainder were precursor cells. In clones that were classified as neuronal, >60% of the cells were neuronal and the remainder were precursor cells. In clones that were classified as mixed, on average 44% of the cells were astrocytes, 17% were neurons, and 39% were precursor cells. The number of cells per clone was similar in control cultures (4.8 ± 0.2), in CNTF-treated cultures (4.6 ± 0.2), and in PDGF-treated cultures (4.5 ± 0.2). Similar results were obtained in cultures that were analyzed with antibodies to β-galactosidase, guinea pig anti-GFAP, and the mouse monoclonal antibody to nestin, where precursor cells were identified as nestin-positive, GFAP-negative cells. Analysis of the various clone cohorts 4 DIV was used as a gauge for the behavior of precursor cells at the time of initial treatment with the various factors. Thus, the minimum percentage of precursor cells that have the potential to differentiate into astrocytes or neurons and that respond to CNTF was calculated as A - B, where A is the proportion of precursor cells that have the potential to differentiate into astrocytes, represented by the percentage of astrocyte-containing clones in the CNTF-treated cultures; B is the maximum proportion of precursor cells that may have lost the potential to differentiate into neurons and could differentiate into astrocytes, represented by the percentage of precursor clones in PDGF-treated cultures (where the greatest amount of neuronal differentiation occurred). To allow detection of living proliferating precursor cells, cultures were also infected at the time of plating with a retrovirus expressing GFP. GFP expression was detected within the precursor cells at 2 DIV. Cohorts of GFP-expressing clones (with an average of 1.2 cells per clone) were then ten untreated (106 clones) or treated with CNTF (104 clones) and were monitored every 3 hours until 4 DIV, when the cultures were fixed and subjected to indirect immunofluorescence. The survival rates of the monitored cell clones at 4 DIV in cultures that were untreated and treated with CNTF were 94.3% and 94.2%, respectively. Within the surviving clones, >98% of cells survived. A subset of GFP-expressing clones (81 in control cultures and 67 in CNTF-treated cultures) was analyzed at 4 DIV for the presence of astrocytes or neurons in separate wells by indirect immunofluorescence with a rabbit antiserum to GFAP or the TuJ1 mouse monoclonal antibody. Control cultures contained no astrocytic clones, whereas 73.5% clones contained neurons (>80% of cells within these clones were neurons). By contrast, in CNTF-treated cultures 66% of clones contained astrocytes (>80% of cells within these clones were astrocytes) and the number of neuron-containing clones was reduced by 35%.
18. L. J. Richards et al., Eur. J. Neurosci. 8, 291 (1996).
19. S. Davis et al., Science 260, 1805 (1993).
20. S. Davis and G. D. Yancopoulos, Curr. Opin. Cell Biol. 5, 281 (1993).
21. C. Ware et al., Development 121, 1283 (1995).
22. C. Lütticken et al., Science 263, 89 (1994).
23. N. Stahl et al., ibid., p. 92.
24. J. E. Darnell Jr., I. M. Kerr, G. R. Stark, ibid. 264, 1415 (1994).
25. J. N. Ihle and I. M. Kerr, Trends Genet. 11, 69 (1995).
26. N. Stahl et al., Science 267, 1349 (1995).
27. A. Bonni, D. A. Frank, C. Schindler, M. E. Greenberg, ibid. 262, 1575 (1993).
28. 759 with Phe. The EL construct contains wild-type LIFRβ cytoplasmic sequences (aa 834 to 1097) fused to the extracellular domain of the EGFR. The dY3-5 mutation contains a deletion of LIFRβ (aa 980 to 1097) that removes three COOH-terminal tyrosine motifs within LIFRβ that are critical for its ability to activate STAT3 in transfected COS cells (20). The EGt chimeric protein contains a truncation of gp130 (aa 758 to 918) and EGtY4 is an EGt protein with an appended tyrosine motif (Tyr-Leu-Pro-Gln) derived from Y4 in gp130. The mutations of gp130 and LIFRβ within the chimeric receptors have no effect on the amount of expression of the chimeric receptors (23).
29. K. Nakajima et al., EMBO J. 15, 3651 (1996).
30. Neither mutation of STAT3 (in STAT3F or STAT3D) had an effect on the expression of the protein when tested in transiently transfected COS cells.
31. M. A. Kahn et al., J. Neurochem. 68, 1413 (1997). The ability of CNTF to induce the GFAP promoter was reproducible but modest. However, if the GFAP mRNA and protein are stable then modest changes in transcriptional levels of the GFAP gene over prolonged times would lead to dramatic changes in the amount of GFAP mRNA and protein. In addition, the basal level of the transfected GFAP-luciferase reporter may be higher than the basal level of the endogenous GFAP gene in cortical precursor cells, possibly because of the presence of multiple copies of the reporter gene in transfected cells and activation of the reporter by the transfection procedure. An increase in the basal level of GFAP promoter activity would then be expected to reduce the level of CNTF induction of the promoter.
32. F. Besnard et al., J. Biol. Chem. 266, 18877 (1991).
33. D. F. Condorelli et al., J. Neurosci. Res. 39, 694 (1994).
34. W. Liedtke et al., Neuron 17, 607 (1996).
35. M. A. McCall et al., Proc. Natl. Acad. Sci. U.S.A. 93, 6361 (1996).
36. K. Shibuki et al., Neuron 16, 587 (1996).
37. D. A. Frank et al., Proc. Natl. Acad. Sci. U.S.A. 92, 7779 (1995).
38. Indirect immunofluorescence was done as described (27).
39. Transfections were done as described [Z. Xia et al., J. Neurosci 16, 5425 (1995)].
40. In the experiments shown in Fig. 2A, the β-galactosidase signal was detected with a mouse monoclonal antibody to β-galactosidase (Promega) diluted 1:300, followed by a goat antibody to mouse immunoglobulin G (IgG) conjugated to Cy3 [Biological Detection Systems (BDS)]. The nestin signal was defected with a rabbit antiserum to nestin (1:5000) followed by a goat antibody to rabbit IgG conjugated to Cy2 (BDS). The GFAP signal was detected with a guinea pig anti-GFAP (Advanced ImmunoChemical) diluted 1:500, followed by a donkey antibody to guinea pig IgG conjugated to AMCA diluted 1:200 (Jackson ImmunoResearch).
41. Cortical cultures (E17 + 3 DIV ) were transfected with a chimeric protein together with the GFAP luciferase reporter gene [containing 1876 nucleotides of the 5′ regulatory region of the GFAP gene fused to the luciferase gene in pGL3 (Promega)] and the EF-CAT plasmid (containing the bacterial chloramphenicol acetyltransferase downstream of the elongation factor 1α promoter) to serve as an internal control for transfection efficiency. Lysates of transfected cultures were left untreated or stimulated with the appropriate ligand [EGF (30 ng/ml) or NT-3 (50 ng/ml)] for 12 hours and were then analyzed for luciferase (Promega kit) and CAT (Dupont kit) activities. Induction with CNTF treatment was determined after luciferase activity was normalized with CAT activity in all transient expression assays in which the GFAP promoter was tested except in the case of transfections with the TG series of chimeric proteins because NT-3 was found to influence expression of the control EF-CAT gene.
42. Immunoprecipitations were done as described (26). Protein lysates were immunoprecipitated with an antiserum to JAK1 [Upstate Biotechnology, Inc. (UBI)], separated by SDS-polyacrylamide gel electrophoresis (PAGE), and immunoblotted with antibodies to phosphotyrosine (PY20) (ICN) and 4G10 (UBI). Antibody binding was detected by ECL (Amersham) with a secondary antibody conjugated to horseradish peroxidase. Protein immunoblot analysis was done as described [A. Bonni et al., Mol. Cell. Neurosci. 6, 168 (1995)].
43. The antibodies to phosphorylated STAT3 were generated as described (37).
44. The mouse monoclonal antibody to nestin was obtained from the Developmental Studies Hybridoma Bank maintained by the Department of Pharmacology and Molecular Sciences, Johns Hopkins University School of Medicine, Baltimore, MD, and the Department of Biological Sciences, University of Iowa, Iowa City, under contract N01-HD-62915 from the National Institute of Child Health and Human Development.
45. Similar results were obtained with the antibody to phosphorylated STAT3 (15). CNTF also induced STAT tyrosine phosphorylation in E14 + 0 DIV cortical cultures (15).
46. Northern blot RNA analysis was done as described [Y. Sun et al., Mol. Cell. Neurosci. 7, 152 (1996)].
47. DNA mobility-shift assays were done as described (27).
48. KA97 plasmid; D. L. Feinstein for GFAP cDNA; C. A. Walsh for the β-galactosidase-expressing retrovirus; T. D. Palmer and F. H. Gage for the GFP-expressing retrovirus; S. Vasquez for technical assistance; and D. Levy, G. Corfas, J. A. Loeb, T. Vartanian, G. D. Fischbach, B. A. Barres, and members of the Greenberg laboratory for helpful discussions and critical reading of the manuscript. Supported by an NIH RO1 grant (CA43855; M.E.G.) and an MRRC grant (NIHP30-HD 18655). Animal care was in accordance with institutional guidelines.