Neuroscience Letters, 133 (1991) 29-32
29
© 1991 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/91/$ 03.50
NSL 08183
Retrograde axonal transport of the GTP-binding protein Gi : a potential neurotrophic intra-axonal messenger I.A. H e n d r y and M.F. Crouch Division of Neuroscience, John Curtin School of Medical Research, Australian National University, Canberra (Australia) (Received 9 August 1991; Accepted 13 August 1991)
Key words: Guanosine triphosphate-binding protein; Retrograde axonal transport; Neurotrophic messenger When a neuronal target is to provide information to the nucleus of the neurone innervating it, it faces the problem of getting its message up the long length of axon separating the cell body from the site of receptor activation at the terminal. The retrograde axonal transport of the neurotrophic molecule, nerve growth factor (NGF), provided one possible mechanism for this information transfer in the sympathetic nervous system. However, some neurotrophic molecules are not retrogradely transported, indicating the message is carded back by a different mechanism. In this paper, we examined such a novel mechanism mediated by the retrograde axonal transport of the • subunit of the second messenger protein, Gi. It is proposed that some non-transported neurotrophic molecules may produce a stable second messenger that is itself transported to the nucleus to convey the target derived information for survival.
Developing peripheral neurones require neurotrophic support from their target tissue and die in its absence [11]. The mechanism of the information transfer from the target tissue constitutes a special case of signal transduction, as the site of receptor activation at the nerve terminal is separated from the nucleus by a long length of axon. It is now generally accepted that target tissues produce limited amounts of neurotrophic molecules (retrophins) which are retrogradely transported back to the cell body where they exert a survival effect [13]. Recently, however, a number of putative neurotrophic molecules have been described which do not appear to be retrogradely transported, for example acidic and basic fibroblast growth factor (aFGF, bFGF) [9, 12]. It is possible that the retrograde transport of a secondary molecule may be the neurotrophic signal from the nerve terminal after receptor stimulation by these factors. Support for a potential role of the GTP-binding protein Gi as a transportable messenger comes from studies on another cell type. Stimulation of Balb/c 3T3 cells by insulin, epidermal growth factor (EGF) or phorbol dibutyrate induced the intracellular translocation of the subunit of Gi (Gi~). This is characterized by an initial rapid movement of Gia from the membrane to perinu-
Correspondence: I.A. Hendry, Division of Neuroscience, John Curtin School of Medical Research, Australian National University, Canberra ACT 2601, Australia.
clear sites and later in the cell cycle by an association of Gi~ with the nuclear chromatin throughout mitosis [5]. This raises the possibility that activation of tyrosine kinase receptors may lead to a modification of Gi~ which results in its translocation to the nucleus. A similar modification in neurones also may result in the translocation of Gi~ from the terminal membrane via the axons to the cell body. In order to determine whether G-protein second messenger systems may be involved in neuronal survival, the effect of pertussis toxin on aFGF-induced neuronal survival was assessed. Ciliary or sensory ganglia from 8day-old chick embryos were dissociated and cultured in Dulbecco modified Eagles Medium containing 1 ktg/ml aFGF as described previously [2] and in the presence or absence of 200 ng/ml pertussis toxin for 24 h. The
TABLE I EFFECT OF PERTUSSIS TOXIN ON aFGF-PROMOTED NEURONAL SURVIVAL Values represent mean and S.E.M. of phase bright cells in each well after 24 h. Number of surviving neurones per well
Ciliary neurones Sensory neurones
Control
Pertussis toxin
1003 ___63.4 957 ___ 38.7
358 ___ 60.7 495 + 7.7
30
A
B
affinity-purified antibodies to the C-terminal decapeptide of Gia (anti-Gi~) [6, 10] and FITC-labelled secondary anti-rabbit antibodies (Wellcome, U.K.) as previously described [5]. Both sensory and ciliary neurones grown in culture in the presence of aFGF each have Gia immunoreactivity present in the growth cones (Fig. 1A). Neurones in unstimulated control cultures died over the same period and thus unstimulated levels could not be determined. Higher power sections of axons using the confocal microscope show immunofluorescent particles containing Gi~ in the axons of these neurones (Fig. 1B). In order to determine in vivo whether there was indeed retrograde axonal transport of the Gi~ we ligated the sciatic nerve of adult mice in the mid thigh region and waited for various lengths of time before the nerve was removed and sectioned. The nerve was sectioned longitudinally in 7-pm slices on a cryostat and processed for Gi~ immunohistochemistry sequentially with anti-Gi~, biotinylated secondary antibodies (Amersham, U.K.) and the
A
m
B Fig. 1. lmmunofluorescent staining of cultured neurones with antibodies to Gi~. A: low power fluorescence microscopy of a sensory neurone showing the presence of Gi~ in the cell body, growth cones and axons. B: high power confocal section through the axon bouton of a ciliary neurone showing the vesicular localization of the G~. Bar, A = 25 #m, B=2/~m.
number of surviving neurones was counted using phase contrast microscopy in 4 fields representing 9% of the well [2]. The neurotrophic activity of aFGF in causing the survival of both sensory and parasympathetic neurones was reduced by pertussis toxin (Table I). Pertussis toxin ADP-ribosylates the GTP-binding proteins, Go and Gi, in ciliary neurones [6] suggesting that one or both of these GTP-binding proteins is required for the full expression of the neurotrophic activity of aFGF. To further ascertain the likelihood that the GTP-binding proteins may play a role in such neurotrophic activity, the presence of Gi~ was determined in cultured neurones. Dissociated sensory and ciliary neurones were cultured in the presence of 1 #g/ml aFGF and, after 24 h, cellular Gi~ was visualized by immunofluorescence with
C
Fig. 2. Immunofluorescence of the adult mouse sciatic nerve ligated in the mid thigh region. A: photo-montage showing accumulation of immunoreactive G~ in both the proximal and distal stump 4 h after ligation (bar= 100 #m). B: nerve immediately after ligation showing the absence of immunoreactivity due to simple damage. C: distal stump showing immunofluorescence is blocked by the specific G~ decapeptide.
31
PEPTIDE
-
+
MOL.Wt. kDa 94 67
43 ii
Gi,c
30 :iiiii~iiii
20
14
i~ii!ii! i Fig. 3. Immunoprecipitationof Gi=from mousesciaticnerve. Proteins were immunoprecipitatedfrom a homogenateof a lengthof the ligated nerve with the anti-Gi=, resolved by SDS-PAGE and the gel silver stained. Molecularweightmarkers are indicated as is the precipitated Gi=.
labelling visualized with Texas red-streptavidin (Amersham, U.K.). Immunostaining with anti-Gi= after 4 h showed the accumulation of immunoreactive material on both the proximal and distal sides (Fig. 2A) of the ligature with the majority of the accumulation being on the distal side. The staining for Gi= was not present immediately after ligation (Fig. 2B), was barely visible after 2 h, increased up to 8 h and persisted for at least 24 h after ligation. The immunofluorescence was blocked by previous incubation of the antibodies with the specific Gi= decapeptide (Fig. 2C). Other peptides, including one specific for Gz ( G Q N N L K Y I G L C , single letter amino acid code) [8] failed to block the fluorescence, further demonstrating the specificity of the anti-Gi=. Taken together these results demonstrate that Gi= is not only transported to the terminal via anterograde transport after its synthesis in the cell body, but also undergoes a
retrograde axonal transport returning it from the periphery to the cell body. The specificity of the antisera and the presence of Gi= in the nerve was confirmed by immunoprecipitation. Sciatic nerve was dissected from the mid thigh region of a mouse, cut into 1-mm pieces, and placed into an octylglucoside-containing buffer [5]. The tissue was homogenized with a teflon pestle and the homogenate centrifuged for 30 min at 4°C at 16,000 g. The supernatant was used for immunoprecipitation of Gi= as previously described [5] either with untreated anti-Gi= (-) or with anti-Gi= blocked by 75 pg/ml of the immunizing peptide ( + ) . It can be seen that a protein of molecular weight corresponding to Gi= is precipitated by the antibodies and this precipitation alone is inhibited by the Gi= peptide (Fig. 3). Gi= may be transported as a membrane-bound species in coated vesicles, such as occurs for the retrograde transport of N G F [18]. Such vesicles are suggested by the confocal microscopic studies (Fig. 1B). Alternatively, Gi= may be modified so as to leave the membrane as occurs on activation of platelets [7] and 3T3 cells [5]. The modification of Gi= that must occur to initiate translocation may be due to phosphorylation [3, 16, 17], the removal of a membrane anchoring molecule such as myristate [4, 5, 19] or attachment of other molecules. The nature of this modification of Gi~ is unknown at present, This study suggests there may be at least two classes of target tissue-derived neurotrophic factors. The first, typified by the N G F family of molecules, are internalized at the nerve terminal together with their receptor, and the factor-receptor complex is then transported to the cell body [14]. As intracellular injection of N G F does not mimic its neurotrophic effect [15], it is thought that N G F itself is not the neurotrophic message but it must act via receptor-generated messengers. We postulate that messengers for this class of neurotrophic factors are relatively labile and need to be produced close to the nucleus. Molecules such as a F G F constitute a second class which is not internalized, perhaps due to their binding to the extracellular matrix [1, 20]. These can interact with receptors on the neuronal terminal to generate a stable second messenger which then acts as the retrograde signal transported via the axon to the cell body to act on the nucleus. Gi= appears to fulfil the criteria necessary for such a receptor-activated signal. We wish to thank Mrs. M. Preston and Mrs. K. Heydon for skillful technical assistance. l Baird,A. and Ling,N., Fibroblastgrowth factorsare presentin the extracellulairmatrixproducedby endothelialcellsin vitro: implications for a role of heparinase-like enzymesin the neovascularresponse, Biochem.Biophys,Res. Commun., 142 (1987) 428-435.
32 2 Bonyhady, R.E., Hendry, 1.A., Hill, C.E. and McLennan, I.S., Characterisation of a cardiac muslce factor required for the survival of cultured parasympathetic neurones, Neurosci. Lett., 18 (1980) 197-201. 3 Bushfield, M., Pyne, N.J. and Houslay, M.D., Changes in the phosphorylation state of the inhibitory guanine-nucleotide-binding protein Gi-2 in hepatocytes from lean (Fa/Fa) and obese (fa/fa) Zucker rats, Eur. J. Biochem., 192 (1990) 537-542. 4 Buss, J.E., Mumby, S.M., Casey, P.J., Gilman, A.G. and Sefton, B.M., Myristoylated alpha subunits of guanine nucleotide-binding regulatory proteins, Proc. Natl. Acad. Sci. U.S.A., 84 (1987) 7493 7497. 5 Crouch, M.F., Growth Factor induced cell division is paralleled by translocation of Gi, to the nucleus, FASEB J., 5 (1991) 200-206. 6 Crouch, M.F., Belford, D.A., Milburn, P.J. and Hendry, I.A., Pertussis toxin inhibits EGF-, phorbol ester- and insulin-stimulated DNA synthesis in BALB/c3T3 cells: evidence for post-receptor activation of Gi,, Biochem. Biophys. Res. Commun., 167 (1990) 1369-1376. 7 Crouch, M.F., Winegar, D.A. and Lapetina, E.G., Epinephrine induces changes in the subcellular distribution of the inhibitory GTPbinding protein Gi~.2 and a 38-kDa phosphorylated protein in the human platelet, Proc. Natl. Acad. Sci. U.S.A., 86 (1989) 17761780. 8 Dreyfus, C.F., Friedman, W.J., Markey, K.A. and Black, I.B., Depolarizing stimuli increase tyrosine hydroxylase in the mouse locus coeruleus in culture, Brain Res., 379 (1986) 216-222. 9 Ferguson, I.A., Schweitzer, J.B. and Johnson, E.M., Basic fibroblast growth factor: receptor-mediated internalization, metabolism, and anterograde axonal transport in retinal ganglion cells, J. Neurosci., 10 (1990) 2176-2189. 10 Goldsmith, P., Gierschik, P., Milligan, G., Unson, C.G., Vinitsky, R., Malech, H.L. and Spiegel, A.M., Antibodies directed against synthetic peptides distinguish between GTP-binding proteins in neutrophil and brain, J. Biol. Chem., 262 (1987) 14683-14688.
11 Hamburger, V., The effects of wing bud extirpation on the development of the central nervous system in chick embryos, J. Exp. Zool., 68 (1934) 449-494. 12 Hendry, I.A. and Belford, D.A., Lack of retrograde axonal transport of the heparin binding growth factors by chick ciliary neurones, Int. J. Dev. Neurosci., 9 (1991) 243-250. 13 Hendry, I.A. and Hill, C.E., Retrograde axonal transport of target tissue-derived macromolecules, Nature, 287 (1980) 647-649. 14 Hendry, I.A., Strckel, K., Thoenen, H. and Iversen, L.L., The retrograde axonal transport of nerve growth factor, Brain Res., 68 (1974) 103-12l. 15 Heumann, R., Schwab, M. and Thoenen, H., A second messenger required for nerve growth factor biological activity?, Nature, 292 (1981) 838-840. 16 Houslay, M.D., Bushfield, M. and Milligan, G., Phosphorylation of Gi in intact cells, Trends Biol. So., 15 (1990) 13. 17 Rosen, A., Keenan, K.F., Thelen, M., Nairn, A.C. and Aderem, A., Activation of protein kinase C results in the displacement of its myristoylated, alanine-rich substrate from punctate structures in macrophage filopodia, J. Exp. Med., 172 (1990) 1211-1215. 18 Schwab, M.E., Ultrastructural localization of a nerve growth factor horseradish peroxidase (NGF-HRP) coupling product after retrograde axonal transport in adrenergic neurons, Brain Res., 130 (1977) 190-196. 19 Utsumi, T., Yoshinaga, K., Koga, D., Ide, A., Nobori, K., Okimasu, E., Terada, S. and Utsumi, K., Association of a myristoylated protein with a biological membrane and its increased phosphorylation by protein kinase C, FEBS Lett., 238 (1988) 13-16. 20 Weiner, H.L. and Swain, J.L., Acidic fibroblast growth factor mRNA is expressed by cardiac myocytes in culture and the protein is localized to the extracellular matrix, Proc. Natl. Acad. Sci. U.S.A., 86 (1989) 2683-2687.
29
© 1991 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/91/$ 03.50
NSL 08183
Retrograde axonal transport of the GTP-binding protein Gi : a potential neurotrophic intra-axonal messenger I.A. H e n d r y and M.F. Crouch Division of Neuroscience, John Curtin School of Medical Research, Australian National University, Canberra (Australia) (Received 9 August 1991; Accepted 13 August 1991)
Key words: Guanosine triphosphate-binding protein; Retrograde axonal transport; Neurotrophic messenger When a neuronal target is to provide information to the nucleus of the neurone innervating it, it faces the problem of getting its message up the long length of axon separating the cell body from the site of receptor activation at the terminal. The retrograde axonal transport of the neurotrophic molecule, nerve growth factor (NGF), provided one possible mechanism for this information transfer in the sympathetic nervous system. However, some neurotrophic molecules are not retrogradely transported, indicating the message is carded back by a different mechanism. In this paper, we examined such a novel mechanism mediated by the retrograde axonal transport of the • subunit of the second messenger protein, Gi. It is proposed that some non-transported neurotrophic molecules may produce a stable second messenger that is itself transported to the nucleus to convey the target derived information for survival.
Developing peripheral neurones require neurotrophic support from their target tissue and die in its absence [11]. The mechanism of the information transfer from the target tissue constitutes a special case of signal transduction, as the site of receptor activation at the nerve terminal is separated from the nucleus by a long length of axon. It is now generally accepted that target tissues produce limited amounts of neurotrophic molecules (retrophins) which are retrogradely transported back to the cell body where they exert a survival effect [13]. Recently, however, a number of putative neurotrophic molecules have been described which do not appear to be retrogradely transported, for example acidic and basic fibroblast growth factor (aFGF, bFGF) [9, 12]. It is possible that the retrograde transport of a secondary molecule may be the neurotrophic signal from the nerve terminal after receptor stimulation by these factors. Support for a potential role of the GTP-binding protein Gi as a transportable messenger comes from studies on another cell type. Stimulation of Balb/c 3T3 cells by insulin, epidermal growth factor (EGF) or phorbol dibutyrate induced the intracellular translocation of the subunit of Gi (Gi~). This is characterized by an initial rapid movement of Gia from the membrane to perinu-
Correspondence: I.A. Hendry, Division of Neuroscience, John Curtin School of Medical Research, Australian National University, Canberra ACT 2601, Australia.
clear sites and later in the cell cycle by an association of Gi~ with the nuclear chromatin throughout mitosis [5]. This raises the possibility that activation of tyrosine kinase receptors may lead to a modification of Gi~ which results in its translocation to the nucleus. A similar modification in neurones also may result in the translocation of Gi~ from the terminal membrane via the axons to the cell body. In order to determine whether G-protein second messenger systems may be involved in neuronal survival, the effect of pertussis toxin on aFGF-induced neuronal survival was assessed. Ciliary or sensory ganglia from 8day-old chick embryos were dissociated and cultured in Dulbecco modified Eagles Medium containing 1 ktg/ml aFGF as described previously [2] and in the presence or absence of 200 ng/ml pertussis toxin for 24 h. The
TABLE I EFFECT OF PERTUSSIS TOXIN ON aFGF-PROMOTED NEURONAL SURVIVAL Values represent mean and S.E.M. of phase bright cells in each well after 24 h. Number of surviving neurones per well
Ciliary neurones Sensory neurones
Control
Pertussis toxin
1003 ___63.4 957 ___ 38.7
358 ___ 60.7 495 + 7.7
30
A
B
affinity-purified antibodies to the C-terminal decapeptide of Gia (anti-Gi~) [6, 10] and FITC-labelled secondary anti-rabbit antibodies (Wellcome, U.K.) as previously described [5]. Both sensory and ciliary neurones grown in culture in the presence of aFGF each have Gia immunoreactivity present in the growth cones (Fig. 1A). Neurones in unstimulated control cultures died over the same period and thus unstimulated levels could not be determined. Higher power sections of axons using the confocal microscope show immunofluorescent particles containing Gi~ in the axons of these neurones (Fig. 1B). In order to determine in vivo whether there was indeed retrograde axonal transport of the Gi~ we ligated the sciatic nerve of adult mice in the mid thigh region and waited for various lengths of time before the nerve was removed and sectioned. The nerve was sectioned longitudinally in 7-pm slices on a cryostat and processed for Gi~ immunohistochemistry sequentially with anti-Gi~, biotinylated secondary antibodies (Amersham, U.K.) and the
A
m
B Fig. 1. lmmunofluorescent staining of cultured neurones with antibodies to Gi~. A: low power fluorescence microscopy of a sensory neurone showing the presence of Gi~ in the cell body, growth cones and axons. B: high power confocal section through the axon bouton of a ciliary neurone showing the vesicular localization of the G~. Bar, A = 25 #m, B=2/~m.
number of surviving neurones was counted using phase contrast microscopy in 4 fields representing 9% of the well [2]. The neurotrophic activity of aFGF in causing the survival of both sensory and parasympathetic neurones was reduced by pertussis toxin (Table I). Pertussis toxin ADP-ribosylates the GTP-binding proteins, Go and Gi, in ciliary neurones [6] suggesting that one or both of these GTP-binding proteins is required for the full expression of the neurotrophic activity of aFGF. To further ascertain the likelihood that the GTP-binding proteins may play a role in such neurotrophic activity, the presence of Gi~ was determined in cultured neurones. Dissociated sensory and ciliary neurones were cultured in the presence of 1 #g/ml aFGF and, after 24 h, cellular Gi~ was visualized by immunofluorescence with
C
Fig. 2. Immunofluorescence of the adult mouse sciatic nerve ligated in the mid thigh region. A: photo-montage showing accumulation of immunoreactive G~ in both the proximal and distal stump 4 h after ligation (bar= 100 #m). B: nerve immediately after ligation showing the absence of immunoreactivity due to simple damage. C: distal stump showing immunofluorescence is blocked by the specific G~ decapeptide.
31
PEPTIDE
-
+
MOL.Wt. kDa 94 67
43 ii
Gi,c
30 :iiiii~iiii
20
14
i~ii!ii! i Fig. 3. Immunoprecipitationof Gi=from mousesciaticnerve. Proteins were immunoprecipitatedfrom a homogenateof a lengthof the ligated nerve with the anti-Gi=, resolved by SDS-PAGE and the gel silver stained. Molecularweightmarkers are indicated as is the precipitated Gi=.
labelling visualized with Texas red-streptavidin (Amersham, U.K.). Immunostaining with anti-Gi= after 4 h showed the accumulation of immunoreactive material on both the proximal and distal sides (Fig. 2A) of the ligature with the majority of the accumulation being on the distal side. The staining for Gi= was not present immediately after ligation (Fig. 2B), was barely visible after 2 h, increased up to 8 h and persisted for at least 24 h after ligation. The immunofluorescence was blocked by previous incubation of the antibodies with the specific Gi= decapeptide (Fig. 2C). Other peptides, including one specific for Gz ( G Q N N L K Y I G L C , single letter amino acid code) [8] failed to block the fluorescence, further demonstrating the specificity of the anti-Gi=. Taken together these results demonstrate that Gi= is not only transported to the terminal via anterograde transport after its synthesis in the cell body, but also undergoes a
retrograde axonal transport returning it from the periphery to the cell body. The specificity of the antisera and the presence of Gi= in the nerve was confirmed by immunoprecipitation. Sciatic nerve was dissected from the mid thigh region of a mouse, cut into 1-mm pieces, and placed into an octylglucoside-containing buffer [5]. The tissue was homogenized with a teflon pestle and the homogenate centrifuged for 30 min at 4°C at 16,000 g. The supernatant was used for immunoprecipitation of Gi= as previously described [5] either with untreated anti-Gi= (-) or with anti-Gi= blocked by 75 pg/ml of the immunizing peptide ( + ) . It can be seen that a protein of molecular weight corresponding to Gi= is precipitated by the antibodies and this precipitation alone is inhibited by the Gi= peptide (Fig. 3). Gi= may be transported as a membrane-bound species in coated vesicles, such as occurs for the retrograde transport of N G F [18]. Such vesicles are suggested by the confocal microscopic studies (Fig. 1B). Alternatively, Gi= may be modified so as to leave the membrane as occurs on activation of platelets [7] and 3T3 cells [5]. The modification of Gi= that must occur to initiate translocation may be due to phosphorylation [3, 16, 17], the removal of a membrane anchoring molecule such as myristate [4, 5, 19] or attachment of other molecules. The nature of this modification of Gi~ is unknown at present, This study suggests there may be at least two classes of target tissue-derived neurotrophic factors. The first, typified by the N G F family of molecules, are internalized at the nerve terminal together with their receptor, and the factor-receptor complex is then transported to the cell body [14]. As intracellular injection of N G F does not mimic its neurotrophic effect [15], it is thought that N G F itself is not the neurotrophic message but it must act via receptor-generated messengers. We postulate that messengers for this class of neurotrophic factors are relatively labile and need to be produced close to the nucleus. Molecules such as a F G F constitute a second class which is not internalized, perhaps due to their binding to the extracellular matrix [1, 20]. These can interact with receptors on the neuronal terminal to generate a stable second messenger which then acts as the retrograde signal transported via the axon to the cell body to act on the nucleus. Gi= appears to fulfil the criteria necessary for such a receptor-activated signal. We wish to thank Mrs. M. Preston and Mrs. K. Heydon for skillful technical assistance. l Baird,A. and Ling,N., Fibroblastgrowth factorsare presentin the extracellulairmatrixproducedby endothelialcellsin vitro: implications for a role of heparinase-like enzymesin the neovascularresponse, Biochem.Biophys,Res. Commun., 142 (1987) 428-435.
32 2 Bonyhady, R.E., Hendry, 1.A., Hill, C.E. and McLennan, I.S., Characterisation of a cardiac muslce factor required for the survival of cultured parasympathetic neurones, Neurosci. Lett., 18 (1980) 197-201. 3 Bushfield, M., Pyne, N.J. and Houslay, M.D., Changes in the phosphorylation state of the inhibitory guanine-nucleotide-binding protein Gi-2 in hepatocytes from lean (Fa/Fa) and obese (fa/fa) Zucker rats, Eur. J. Biochem., 192 (1990) 537-542. 4 Buss, J.E., Mumby, S.M., Casey, P.J., Gilman, A.G. and Sefton, B.M., Myristoylated alpha subunits of guanine nucleotide-binding regulatory proteins, Proc. Natl. Acad. Sci. U.S.A., 84 (1987) 7493 7497. 5 Crouch, M.F., Growth Factor induced cell division is paralleled by translocation of Gi, to the nucleus, FASEB J., 5 (1991) 200-206. 6 Crouch, M.F., Belford, D.A., Milburn, P.J. and Hendry, I.A., Pertussis toxin inhibits EGF-, phorbol ester- and insulin-stimulated DNA synthesis in BALB/c3T3 cells: evidence for post-receptor activation of Gi,, Biochem. Biophys. Res. Commun., 167 (1990) 1369-1376. 7 Crouch, M.F., Winegar, D.A. and Lapetina, E.G., Epinephrine induces changes in the subcellular distribution of the inhibitory GTPbinding protein Gi~.2 and a 38-kDa phosphorylated protein in the human platelet, Proc. Natl. Acad. Sci. U.S.A., 86 (1989) 17761780. 8 Dreyfus, C.F., Friedman, W.J., Markey, K.A. and Black, I.B., Depolarizing stimuli increase tyrosine hydroxylase in the mouse locus coeruleus in culture, Brain Res., 379 (1986) 216-222. 9 Ferguson, I.A., Schweitzer, J.B. and Johnson, E.M., Basic fibroblast growth factor: receptor-mediated internalization, metabolism, and anterograde axonal transport in retinal ganglion cells, J. Neurosci., 10 (1990) 2176-2189. 10 Goldsmith, P., Gierschik, P., Milligan, G., Unson, C.G., Vinitsky, R., Malech, H.L. and Spiegel, A.M., Antibodies directed against synthetic peptides distinguish between GTP-binding proteins in neutrophil and brain, J. Biol. Chem., 262 (1987) 14683-14688.
11 Hamburger, V., The effects of wing bud extirpation on the development of the central nervous system in chick embryos, J. Exp. Zool., 68 (1934) 449-494. 12 Hendry, I.A. and Belford, D.A., Lack of retrograde axonal transport of the heparin binding growth factors by chick ciliary neurones, Int. J. Dev. Neurosci., 9 (1991) 243-250. 13 Hendry, I.A. and Hill, C.E., Retrograde axonal transport of target tissue-derived macromolecules, Nature, 287 (1980) 647-649. 14 Hendry, I.A., Strckel, K., Thoenen, H. and Iversen, L.L., The retrograde axonal transport of nerve growth factor, Brain Res., 68 (1974) 103-12l. 15 Heumann, R., Schwab, M. and Thoenen, H., A second messenger required for nerve growth factor biological activity?, Nature, 292 (1981) 838-840. 16 Houslay, M.D., Bushfield, M. and Milligan, G., Phosphorylation of Gi in intact cells, Trends Biol. So., 15 (1990) 13. 17 Rosen, A., Keenan, K.F., Thelen, M., Nairn, A.C. and Aderem, A., Activation of protein kinase C results in the displacement of its myristoylated, alanine-rich substrate from punctate structures in macrophage filopodia, J. Exp. Med., 172 (1990) 1211-1215. 18 Schwab, M.E., Ultrastructural localization of a nerve growth factor horseradish peroxidase (NGF-HRP) coupling product after retrograde axonal transport in adrenergic neurons, Brain Res., 130 (1977) 190-196. 19 Utsumi, T., Yoshinaga, K., Koga, D., Ide, A., Nobori, K., Okimasu, E., Terada, S. and Utsumi, K., Association of a myristoylated protein with a biological membrane and its increased phosphorylation by protein kinase C, FEBS Lett., 238 (1988) 13-16. 20 Weiner, H.L. and Swain, J.L., Acidic fibroblast growth factor mRNA is expressed by cardiac myocytes in culture and the protein is localized to the extracellular matrix, Proc. Natl. Acad. Sci. U.S.A., 86 (1989) 2683-2687.
