JOURNAL OF CELLULAR PHYSIOLOGY 150:647-656 (1992)
Cyclic AMP Potentiates bFGF-Induced Neurite Outgrowth in PC12 Cells PAUL0 LEE HO* AND ISAIAS RAW Centro de Biotecnologia, lnstituto Butantan, CP65, 05504, Sao Paulo, Brazil We report here that basic fibroblast growth factor (bFGF)-elicited neurite outgrowth in PC12 cells is potentiated by dibutyryl cyclic adenosine monophosphate (dbcAMP) or forskolin. This property was also described for nerve growth factor (NGF), suggesting that both NGF and bFGF may share common intracellular events leading to neurite outgrowth and synergism with dbcAMP and forskolin. The synergistic effect of dbcAMP and forskolin i s specific, since treatment of PC12 cells with bFGF and dibutyryl cyclic guanosine monophosphate (dbcGMP) or phorbol ester did not change the neurite outgrowth response of cells treated with bFGF alone. Furthermore, neurite outgrowth depends on cellular adhesion. Increasing adhesion by plate treatment with poly-d-lysine increases the neurite outgrowth elicited by bFGF alone or bFGF plus dbcAMP. O n the other hand, decreasing cellular adhesiveness by plating PC12 cells in semi-solid agarose renders the cells unable to develop neuritic processes. In addition, 3H-methylthymidine incorporation studies showed that bFGF-treated PC12 cells cease growth only when they become fully differentiated after 3-5 days of treatment. In contrast, dbcAMP, which i s a poor differentiation factor, i s able to block cellular growth after 24 hour treatment. These results suggest that when PC12 cells become differentiated, they stop growing. However, growth inhibition does not necessarily lead to differentiation.
Basic fibroblast growth factor (bFGF) is a well known mitogen which has also been demonstrated to possess neurotrophic effects in vitro and in vivo (Burgess and Maciag, 1989). The PC12 clonal line of rat pheochromocytoma cells has become a useful model system to study the mechanisms of action of neurotrophic factors, particularly, nerve growth factor (NGF) (Guroff, 1985). This cell line also differentiates in the presence of FGFrelated proteins (Togari et al., 1983; Delli-Bovi et al., 1988; Burgess and Maciag, 1989), eliciting neurite outgrowth, expression of early specific mRNA like c-fos (Greenberg et al., 19851, expression of neural specific mRNA and enzymatic activities like acetylcholinesterase and ornithine decarboxylase (Togari et al., 1983; Rydel and Greene, 1987). The mechanisms by which these changes occur are still unknown, but some of them are probably mediated by tyrosine kinase activity displayed by bFGF receptor upon ligand binding (Burgess and Maciag, 1989; Lee et al., 1989; Ullrich and Schlessinger, 1990). It has been shown that bFGF action can be modulated negatively or positively depending on cellular phenotype. For instance, mitogenesis of endothelial cells by bFGF is inhibited in vitro by phorbol esters (Doctrow and Folkman, 1987; Hoshi e t al., 19881, tumor necrosis factor (TNF) (Schweigerer et al., 1987; FraterSchroder et al., 19871, Interleukin 1 A and B (Norioka et al., 1987), and transforming growth factor B (TGF B) (Frater-Schroder et al., 1986; Baird and Durkin, 1986; Saksela et al., 1987), whereas the mesoderm induction by bFGF in amphibian embryos is enhanced by TGF B (Kimelman and Kirschner, 1987). There is also a coop0 1992 WILEY-LISS, INC
eration between bFGF and platelet-derived growth factor (PDGF) in promoting division and differentiation of oligodendrocyte type 2-astrocyte (0-2A) progenitor cells (Bogler et al., 1990). Such modulation should be important in regulating cellular proliferation and differentiation. Using PC12 cells, we show here that the neurotrophic effects of bFGF can be enhanced by factors that directly activate CAMP-dependent protein kinases like dibutyryl cyclic adenosine monophosphate (dbcAMP) or indirectly through adenylate cyclase activation by forskolin. We also show that these responses depend on cellular adhesiveness, suggesting a n important role for attachment factors in modulating neurite outgrowth induced by trophic factors.
MATERIALS AND METHODS Materials. Dibutyryladenosine 3'5'-cyclic monophosphate (dbcAMP), dibutyrylguanosine 3'5'-cyclic monophosphate (dbcGMP), forskolin, phorbol 12myristate 13-acetate (PMA), poly-d-lysine and agarose were purchased from Sigma Chemical Co. (St. Louis, MO). A mixture of bovine pituitary bFGF (1-146) and (11-146) were purified according to Ho e t al. (1988). Placental basic FGF purification. Human placental bFGF was purified from placental cellular mass after processing for albumin production. Briefly,
Received January 8,1991; accepted October 16,1991.
*To whom reprint requestdcorrespondence should be addressed.
HO AND RAW
648
a
b
kDa
689 434 29.e
18.9 14.3
Fig. 1. Western-blotting of bovine pituitary bFGF and human placental bFGF. Human placental bFGFs (a)were purified from the cellular mass previously used for albumin production as described in Materials and Methods. The purified form has a higher molecular weight than bovine pituitary bFGF (1-146)(b).Some immunoreactive components of 2 S 3 0 kDa may represent copurified proteases (Ho et al., 1990). The blot was developed using peroxidase conjugates and metal ion enhancement (Harlow and Lane, 1988).
healthy human placentas were collected from Hospital Universitario (Universidade de Sao Paulo, SP) and stored frozen (-40°C) until use. Frozen placentas were thawed overnight in a cold room and ground to 2 4 cm3 pieces. These tissues were then extracted with 10% (v/v) ethanol in 38 mM acetate buffer pH 6.8 for four hours, The supernatant represents the fraction used for albumin production and the pellet represents the cellular mass used for growth factor purification. Cellular mass from 10 Kg of frozen placentas were extracted with 50 mM phosphate buffer pH 7.0 in a Waring blender and centrifuged. The crude extract (supernatant) was chromatographed in a S-Sepharose column (Pharmacia, Uppsala, Sweden). The active fractions eluted with 0.8 M NaC1, were pooled, diluted to 0.15 M NaCl and pumped into a Heparin-Sepharose column. The adsorbed materials were eluted stepwise with 0.8 M, 1.2 M and 2.0 M NaC1. The bFGF was eluted in the last wash. The purified placental bFGF (Fig. 1)has a molecular weight higher than bovine pituitary bFGF (1-146) and probably represents the bFGF (-11-146) form already described (Sommer et al., 1987). The antisera anti-bFGF was kindly provided by Dr. D. Rifkin (New York University School of Medicine, NY). Cell culture. PC12 cells were a generous gift of Dr. A. Howllet (Saint Louis University, St. Louis, MO). Cells were grown in Dulbecco’s Modified Eagle’s Medium (DME) supplemented with 10% fetal calf serum (FCS) a t 37°C in a humidified atmosphere of 5% CO,. Subcultures were done in a 1:4 split when cells became confluent. Quantitation of neurite outgrowth. Cells (2.5 x lo4 cells/ml) were plated directly a 12-well microplate (Corning, NY) in DME plus 10% FCS or in poly-d-lysine-(10 p,g/ml) treated 12-well microplates. Twelve hours later, the culture was maintained or
changed to DME plus 2.5% FCS supplemented or not with bovine pituitary bFGF, human placental bFGF, dbcAMP, dbcGMP, forskolin or PMA, as indicated in the figure legends. No difference in neurite outgrowth response was observed whether we performed the assays in 10% or 2.5% FCS. In all cases, medium was changed every two days with the addition of fresh supplements. A neurite was identified as a process whose length was equal or greater than one cell body length. Isolated cells or groups containing no more than 10 clearly distinguishable cells were considered for scoring neurite outgrowth. In cells presenting two or more neurites, only the longest one was scored. One to two hundred cells per well were randomly observed under phase-contrast microscope and the percentage of cells with neurites was calculated. Distribution of the cellular population according to neurite length. Cells were separated into seven classes according to neurite length: cells without neurites (0);cells with neurites smaller than one cell body length (- 1); cells presenting neurites corresponding to one (11, two (21, three (31, four (41, and more than four (+4) times the cell body length. The conditions were the same as described above. Plating in semi-solid agarose. Cells (1-2 x lo4 cells/well) were plated directly in untreated 24-well microplates (Corning, NY) in DME plus 10% FCS. On the next day, the medium was removed and melted 0.6% agarose was added over the cells. On the solidified agarose, we seeded a second cell layer (1-2 x lo4 cellslwell) and again, another layer of agarose. DME plus 10% FCS supplemented or not with human placental bFGF andlor dbcAMP was added over the second layer of agarose. Media and supplements were changed every two days. We followed the differentiation of the cells plated on a substrate (first cell layer) or without a substrate (second cell layer in semi-solid agarose) on the different microscope focus planes. Cellular proliferation assay. Cells (2 x lo4 cells/ well) were plated in untreated 24-well microplates in DME plus 10% FCS. Twelve hours later, we added bFGF and/or dbcAMP (time 0 of treatment). The media and supplements were changed every two days. Cellular proliferation was followed by 3H-methyl-thymidine uptake into DNA. Briefly, 3H-methyl-thymidine (0.5 pCi/ml, 1 x M) was added 24 hours before collecting the cells. The cells were treated with cold trichloroacetic acid (TCA) and lysed with sodium hydroxide which was further adsorbed to a filter paper, washed with cold TCA, then ethanol, dried and counted by scintillography .
RESULTS Potentiation of neurite o u t g r o w t h b y d b c A M P Based on previous data showing that bFGF and NGF share some common responses in promoting differentiation of PC12 cells (Wagner and D’Amore, 1986; Rydel and Greene, 1987) and that CAMP analogs are able to increase neurite outgrowth elicited by NGF in this cell line (Richter-Landsberg and Jastorff, 1986),we investigated whether this last property is also shared by bFGF. For this purpose, we used a human placental bFGF purified from the cellular mass previously used for albumin production (Fig. 1).
NEUROTROPHIC EFFECTS OF CAMPAND bFGF IN PC12 CELLS
Fig. 2. Effect of dbcAMP on bFGF-induced neurite outgrowth. PC12 cells in 2.5% FCS were exposed to placental bFGF (5ng/ml), dbcAMP (1mM) or placental bFGF (5 ngiml) plus dbcAMP (1mM) for 24 hours (a,c,e,g)or 96 hours (b,d,f,h).(a,b) untreated control; (c,d) dbcAMP; (e,f)placental bFGF; (g,h) placental bFGF plus dbcAMP. Bar i n a = 15 km, and applies to b-h.
649
HO AND RAW
650
1 I
1
50
5
3 I
I
DAYS
I
h 0
ll
4:
S
V
I 1
I
I
I
,
,
2
3
4
5
6
DAYS
Fig. 3. Kinetics of neurite formation. PC12 cells in 2.5%FCS were exposed to (0-0) 5 ng/ml placental bFGF; (-0) 1 mM dbcAMP; or ( A-A f 5 ng/ml placental bFGF plus 1 mM dbcAMP. Untreated control cells did not develop neurites at all (not shown). When not indicated, the error was less than 1%.
v)
n
-I
0
+
A 3 w 0 1
:b: 5
s 0 -1 1 2 3 4 + 4
0 -1 1 2 3 4 + 4
0 -1 1 2 3 4 4 INDEXES
Figure 2 shows that simultaneous treatment of PC12 4. Distribution of the cellular population according to neurite cells with bFGF and dbcAMP induces higher neurite Fig. length. PC12 cells in 2.5% FCS were exposed to 5 ngiml placental outgrowth density than bFGF or dbcAMP alone. This bFGF, 1mM dbcAMP, or 5 ngiml placental bFGF plus 1mM dbcAMP. effect can be detected either after 24 (Fig. 2g) or 96 (Fig. Cells in days 1, 3, 5 were classified according to neurite length (see 2h) hours of treatment. In the first 24 hours, the treat- Materials and Methods) and expressed as percentage of the total numment with dbcAMP or bFGF alone is equally ineffi- ber of cells analysed. Untreated control cells did not develop neurites cient, inducing few processes. However, after long term at all (not shown). When not indicated, the error was less than 1%. incubation, bFGF induces a higher number of neurites than dbcAMP (Figs. 2d,f and 5b,c). Quantitative analysis of cells bearing neurites show are present. These results show that PC12 cells present the synergistic effects of dbcAMP and bFGF (Fig. 3). different kinetics of neurite outgrowth in response to The percentage of cells with neurites in those treated dbcAMP, bFGF, and dbcAMP plus bFGF. with dbcAMP and bFGF in days 1 , 2 , and 3 are higher Effects of forskolin than the sum of the percentage of cells treated alone with each factor. Figure 3 also shows that the percentdbcAMP has to be metabolically activated before age of cells bearing neurites is low in dbcAMP treated binding to cAMP receptor proteins (Kaulkel and Hilz, cells, with plateau values of 20-25%. 1972). To exclude non-related side effects caused by Analysis of neurite length on days 1,3, and 5 clearly metabolites, such as butyric acid, generated during this shows different cellular distribution according to the activation, we used forskolin to confirm the synergistic treatment used (Fig. 4). Cells treated for 24 hours with effects of cAMP and bFGF in eliciting neurite outbFGF show high percentage of cells without neurites or growth. Forskolin is a known diterpene that activates with processes smaller than one cell body length. How- adenylate cyclase (Seamon and Daly, 1983).Treatment ever, with increasing incubation time, the percentage of PC 12 cells with forskolin raises the intracellular of cells without neurites diminishes and there is an cAMP level (Rabe et al., 1982) and has little effects on increase in the percentage of cells with neurites equal neurite outgrowth in cells continuously incubated for or greater than one cell body length. In contrast, the 144 hours (Fig. 5e). However, when bFGF and forskolin pattern of distribution of cells treated with dbcAMP is were added simultaneously, a potentiation of bFGFquite different; they did not change significantly with elicited neurite outgrowth was found (Fig. 50, confirmincubation time, being similar on days 1 and 5. The ing the involvement of intracellular cAMP in this repercentage of cells bearing no neurites or with neurites sponse. Although bFGF is able to induce neurite less than one cell body length remains high in all the outgrowth, the neurites are often short and linear, days examined. The synergistic effect of dbcAMP plus whereas in cells incubated with both bFGF and bFGF can also be seen in Figure 4. The cellular distri- dbcAMP or forskolin, the neurit.es are often longer and bution pattern changed during continuous treatment. branched and cell bodies are hyperthrophied at the inOn day 1there are cells bearing neurites with 1 , 2 , and cubation times analysed (Fig. 5). 3 cell body length, although a high percentage of cells Effects of dbcGMP and PMA with indexes 0 and -1 are still found. However, the percentage of cells with indexes 0 and -1 on the fifth Since dbcAMP and forskolin increase intracellular day is insignificant, and high percentage of cells with CAMP,the potentiation effects of these drugs on bFGFneurites with 2,3,4 and greater than 4 cell body length induced neurite outgrowth is probably mediated
NEUROTROPHIC EFFECTS OF CAMPAND bFGF IN PC12 CELLS
651
Fig. 5. Branching effects of dbcAMP or forskolin on bFGF-induced neurite outgrowth. PC12 cells i n 2.5%FCS were exposed to placental bFGF (5 ngiml), dbcAMP (1 mM), forskolin (10 pM),placental bFGF (5 ngiml) plus dbcAMP (1 mM) or placental bFGF (5 ngiml) plus forskolin (10 pM) for 144 hours. (a) untreated control; (b) placental bFGF; (c) dbcAMP; (d) placental bFGF plus dbcAMP ( e ) forskolin; (0 placental bFGF plus forskolin. Bar in a = 15 pm, and applies to b-f.
through CAMP-dependent protein kinase (Roesler et al., 1988; Taylor e t al., 1990). In order to verify ifthis effect could also be promoted by cGMP-dependent protein kinases or protein kinase C (PKC), we treated PC12 cells with 1mM dbcGMP or 10 nM PMA, alone or with bFGF. Neither dbcGMP nor PMA altered the neurite outgrowth elicited by bFGF (results not shown). Thus, we can conclude that this feature is specific to intracellular increase of CAMP, probably through the activation of protein kinase A.
Neurite outgrowth depends on cell adhesion The neurotrophic effects of bFGF in PC12 cells plated on plastic substrate can be easily followed and the neu-
rites quantitatively scored in short term incubation (6-8 days). However, the increasing clamping effects combined with low adhesion of PC12 cells on plastic after long term incubation (beyond 8 days) made the experiments difficult to follow and control. Culture plate treatment with poly-d-lysine allowed us to better control long term incubations and examine the effect of increased adhesion in bFGF-elicited neurites in these cells. Cells grown on poly-d-lysine treated plates show increased response to the neurotrophic effects of bFGF when compared to control cells on untreated plastic plates (Fig. 6e,f). dbcAMP alone is able to induce neurite outgrowth (Fig. 6c,d). The synergistic effects of bFGF and dbcAMP is also seen in Figure 6g,h. The long
652
HO AND RAW
Fig, 6. Effects of increased adhesiveness by plating PC12 cells on poly-d-lysine. PC12 cells in 10% FCS were exposed to placental bFGF (5 ngiml), dbcAMP (ImM) or placental bFGF (5 ngiml) plus dbcAMP (1 mM) for 96 hours in untreated (a,c , e, g) or poly-d-lysine treated culture plates (b,d, f, h). (a,b) untreated control; (c,d) dbcAMP; (e,f)placental bFGF; (g,h) placental bFGF plus dbcAMP. Bar in a = 15 km, and applies to b-h.
NEUROTROPHIC EFFECTS OF CAMPAND bFGF IN PC12 CELLS
term incubation effects of the supplements were examined under these conditions. Surprisingly, on day 10, we start to notice significant cell death in bFGF-treated cells. Fifty percent of the cellular population treated with bFGF remained on day 16. Neither dbcAMP alone nor combined with bFGF showed this feature. It seems that dbcAMP prevents cellular mortality induced by long term incubation with bFGF (results not shown). On the other hand, we also decreased adhesiveness by plating PC12 cells in semi-solid agarose, which renders the cells unresponsive to bFGF and/or dbcAMP (Fig. 71, in contrast to control cells in the same well plated on untreated plastic substrate.
Differentiation versus proliferation In order to evaluate if the promotion of neuronal differentiation of PC12 cells by bFGF is concomitant with growth inhibition, 3H-methyl-thymidine uptake studies were performed, as shown in Figure 8. Treatment with bFGF did not alter cellular proliferation during the first 4 days. After this time, cell growth was inhibited. These results parallel those showing that bFGF did not promote neurite outgrowth of PC12 cells in short term treatment (1-3 days), but only after 3-5 days of continuing incubation (Figs. 2-5). Furthermore, the combination of dbcAMP and bFGF, which induces neuritic processes a s soon a s day 1 of treatment (Figs. 2-4), inhibited the 3H-methyl-thymidine uptake of the cells in this condition. These data suggest that when the cells become differentiated, their growth is inhibited. However, dbcAMP, which alone is a poor differentiation factor in this system (Figs. 2-5), is able to stop cell division within 24 hours of treatment (Fig. 8). DISCUSSION Using PC12 cells as a model system for the study of neuronal differentiation, we investigated whether neurite outgrowth elicited by bFGF can be positively modulated by dbcAMP or forskolin, a drug that increases intracellular cAMP through adenylate cyclase activation. Our results show that bFGF-elicited neurite outgrowth is potentiated by dbcAMP or forskolin. This property was also described for NGF (Richter-Landsberg and Jastorff, 1986), suggesting that NGF and bFGF may share common intracellular events leading to neurite outgrowth. Since simultaneous addition of bFGF and dbcAMP or forskolin potentiate neurite outgrowth in PC12 cells synergistically, we can conclude that they probably act through different mechanisms. This observation is supported by recent reports showing that NGF, aFGF, and bFGF regulate neurite outgrowth in PC12 cells via both PKC and CAMP-independent mechanisms (Reinhold and Neet, 1989; Damon et al., 1990; Sigmund e t al., 1990). However, other responses elicited by FGFs and NGF can be totally or partially dependent on PKC and CAMP-dependent protein kinase. For instance, ornithine decarboxylase gene expression in bFGF and aFGF but not NGF treated cells is dependent on both PKC and CAMP-dependent protein kinase systems (Reinhold and Neet, 1989; Damon et al., 1990), whereas the expression of d5 gene is independent and the expression of d2 gene is partially dependent on PKC and CAMP-dependent protein kinase pathways (Damon et al., 1990). The expression of other genes like SCGlO was also shown to be indepen-
653
dent of PKC, while the expression of c-fos mRNA by NGF but not bFGF is partially mediated by PKC (Sigmund et al., 1990). These results show that some cellular responses are differentially mediated by bFGF and NGF and may reflect differences in receptor signalling. The synergistic effect of dbcAMP or forskolin is specific, since treatment of PC12 cells with bFGF and dbcGMP or PMA did not change the neurite outgrowth response of cells treated with bFGF alone (results not shown). In contrast, PMA is able to inhibit the mitogenic effects of bFGF in endothelial cells (Doctrow and Folkman, 1987; Hoshi et al., 19881, showing that the same trophic factors may have different influences in different cells. It is also worth considering the possibility that trophic factors like neurotransmiters or peptides, which increase intracellular cAMP levels, may exert regulatory roles in promoting neural differentiation together with bFGF and NGF. It was shown that cAMP analogs by themselves are sufficient to promote survival and neurite outgrowth in cultures of rat sympathetic and sensory neurons (Rydel and Greene, 19881, supporting the idea of a n important role for cAMP in neuronal differentiation. The experiments designed to increase or decrease adhesiveness (Figs. 6,7) showed the anchorage dependence of PC12 cells to a substrate in response to bFGF andlor dbcAMP, suggesting a role for attachment factors in modulating the action of neurotrophic factors. Dependence of culture substrate for cell survival and neurite outgrowth was also seen in other types of bFGF-treated nervous cells, like hippocampal and cerebellar granule neurons (Walicke et al., 1986; Hatten et al., 1988). Recently, it was shown that EGF receptors activated by ligand binding stimulate tyrosine phosphorylation of phospholipase CY 1 (PLCY1). In vitro, both phosphorylated and unphosphorylated PLCY 1are able to hydrolyse inositol phospholipids to inositol phosphate and diacylglicerol. However, the enzymatic activity of the non-phosphorylated form is inhibited by profilin, a n actin binding protein, whereas the phosphorylated form is not (Goldschmidt-Clermont et al., 1991; Rhee, 1991). It was also shown that another actin-binding protein, tensin, possesses a SH2 domain responsible for phosphotyrosine interaction (Davis et al., 1991, Cantley et al., 1991). These results make a connection between cytoskeleton and receptor signalling. Like EGF and PDGF receptors, FGF receptors have intrinsic tyrosine kinase activity and phosphorylate several protein targets, among them, PLCY 1(Burgess et al., 1990). So, i t is conceivable that changing adhesiveness of PC12 cells also changes the composition, structure and dynamics of cytoskeleton which, in turn, modify the interactions of cytoplasmic signalling proteins activated by FGF receptor andtor protein kinase A, resulting in decrease or increase of bFGF- and CAMP-induced response. The DNA synthesis of bFGF-treated PC12 cells appears to continue unaltered for about 4 days (Fig. 4). Then, the synthesis of DNA shows a progressive decrease. This was also described for NGF with similar kinetics. The decrease in DNA synthesis parallels with cell division decrease (Gunning et al., 1981). It was further shown t h a t incubation of PC12 cells with NGF progressively down-regulates EGF receptors, diminish-
654
HO AND RAW
Fig. 7. Effects of decreased adhesiveness by plating PC12 cells in semi-solid agarose. PC12 cells in 10% FCS were plated on a substrate (a, c, e, g) or in semi-solid agarose (b, d, f, h) and exposed to placental bFGF (5 ngiml), dbcAMP (1mM) or placental bFGF (5ngiml) plus dbcAMP (1 mM) for 96 hours, according to Materials and Methods. (a,b) untreated control; (c,d) dbcAMP; (e,O placental bFGF; (g,h) placental bFGF plus dbcAMP. Bar in a = 15 Km, and applies to b-h.
NEUROTROPHIC EFFECTS OF CAMPAND bFGF IN PC12 CELLS
655
eration of PDGF-treated rat Schwann cells (Weinmaster and Lemke, 1990). The increase in intracellular cAMP induces PDGF receptor gene transcription and translation, which accounts, at least in part, for this synergism (Weinmaster and Lemke, 1990). Whether cAMP increases the bFGF receptors in PC12 cells, which would in part explain our results, is unknown. This possibility should be considered.
ACKNOWLEDGMENTS P.L.H. would like to thank Drs. E. Kimura, H.A. Armelin, A.G. Gambarini, M.C.S. Armelin, J.C.C. Maia, K.M. Rocha, L. Rameh, J.F. Dias (Universidade de Sao Paulo, SP, Brazil), Drs. A.M. Silva, L.C.C. Leite, M. Katz (Instituto Butantan, SP, Brazil), Dr. D.B. Rifkin (New York University, NY, USA) and Dr. A. Howllet (Saint Louis University, MO, USA) for their helpful discussions, suggestions or for providing reagents. This work was supported by FAPESP (901 0168-5) and CNPq (410061/90-5) Brazilian grants to P.L.H. and FINEP to I.R.
LITERATURE CITED
2
3
4
5
6 DAYS
Fig. 8. Growth response of PC12 cells exposed to placental bFGF and/or dbcAMP, measured by 3H-methyl-thymidine uptake into DNA. PC12 cells in 10% FCS were exposed to (U 5 ng/ml ) of placental bFGF; (0-0) 1mM dbcAMP or (M 5) ngiml of placental bFGF represents untreated control. The results plus ImM dbcAMP.1-( were expressed by cpmiwell (a)or as % inhibition related to untreated control (b).
ing EGF binding as soon as 12 hours, being most of the binding capacity inhibited within 4 days. It was suggested that this heterologous down-regulation is part of the mechanism by which differentiating cells become insensitive to mitogens (EGF) present in serum (Lazarovici et al., 1987).This mechanism probably occurs in bFGF-treated PC12 cells. dbcAMP, which is a poor differentiation factor (Figs. 2-45), is able to decrease DNA synthesis (Fig. 8) as soon as 24 hours after incubation, in contrast to bFGF-treated cells. This inhibition is maintained over the days examined, although no significant neurite extension could be seen. The inhibition of DNA synthesis is greater in cells supplemented with both bFGF and dbcAMP. These results, taken together with those showing neurite outgrowth induced by bFGF and/or dbcAMP (Figs. 2-5), led us to conclude that when PC12 cells become fully differentiated, they stop growing; however, when their growth is inhibited, they are not necessarily differentiated. It is likely that growth inhibition facilitates the action of neurotrophic factors. The mechanism by which cAMP increases the neurite outgrowth promoted by bFGF is unknown. It was shown that cAMP synergistically stimulates the prolif-
Baird, A., and Durkin, T. (1986) Inhibition of endothelial cell proliferation by type B transforming growth factor: Interaction with acidic and basic fibroblast growth factor. Biochem. Biophys. Res. Commun., 138:47&482. Bogler, O., Wren, D., Barnett, S.C., Land, H., and Noble, M. (1990) Cooperation between two growth factors promotes extended self renewal and inhibits differentiation of oligodendrocyte-type 2 astrocyte (0-2A) progenitor cells. Proc. Natl. Acad. Sci. USA, 87r63686372. Burgess, W.H., and Maciag, T. (1989)The heparin-binding (fibroblast) growth factor family of proteins. Annu. Rev. Biochem., 58r575-606. Burgess, W.H., Dionne, C.A., Kaplow, J., Mudd, R., Friesel, R., Zilberstain, A,, Schlessinger, J.,and Jaye, M. 11990)Characterization and cDNA cloning of phospholipase C-Y, a major substrate for heparinbinding growth factor 1 (acidic fibroblast growth factorbactivated tyrosine kinase. Moll. Cell. Biol., 10r4770-4777. Cantley, L.C., Auger, K.R., Carpenter, C., Duckworth, B., Graziani, A,, Kapeller, R., and Soltoff, S. (1991) Oncogenes and signal transduction. Cell, 64r281-302. Damon, D.H., DAmore, P.A., and Wagner, J.A. (1990) Nerve growth factor and fibroblast growth factor regulate neurite outgrowth and gene expression in PC12 cells via both protein kinase C- and CAMPindependent mechanisms. J. Cell. Biol., lIOr1333-1339. Davis, S., Lu, M.L., Lo, S.H., Lin, S., Butler, J.A., Druker, B.J., Roberts, T.M., An, &., Chen, L.B. (1991) Presence of an SH2 domain in the actin-binding protein tensin. Science, 252r712-715. Delli-Bovi, P., Curatola, A.M., Newman, K.M., Sato, Y., Moscatelli, D., Hewick, R.M., Rifkin, D.B., and Basilico, C. (1988) Processing, secretion, and biological properties of a novel growth factor of the fibroblast growth factor family with oncogenic potential. Mol. Cell. Biol., 829334941. Doctrow, S.R., and Folkman, J. (1987) Protein kinase C activators suppress stimulation of capillary endothelial cell growth by angiogenic endothelial mitogen. J . Cell Biol., 104t679-687. Frater-Schroder,M.,Muller,G.,Birchmeier, W., andBohlen,P. (1986) Transforming growth factor B inhibits endothelial cell proliferation. Biochem. Biophys. Res. Commun., 137.295-302. Frater-Schroder, M., Risau, W., Halmman, R., Gautschi, P., and Bohlen, P. (1987) Tumor necrosis factor a, a potent inhibitor of endothelial cells in uitro.. is aneioeenic in uiuo. Proc. Natl. Acad. Sci. USA., 845277-5281. Goldschmidt-Clermont, P.J., Kim, J.W., Machesky, L.M., Rhee, S.G., and Pollard, T.D. (1991) Regulation of phospholipase C-yl by profilin and tyrosine phosphorylation. Science, 251 tlZ31-1233. Greenberg, M.E., Greene, L.A., and Ziff, E.G. (1985) Nerve growth factor and epidermal growth factor induce rapid transient changes in protooncogene transcription in PC12 cells. J. Biol. Chem., 260r14101-14110. Gunning, P.W., Landreth, G.E., Layer, P., Ignatius, M., and Shooter, E.M. (1981) Nerve growth factor induced differentiation of PC12 Y
l
656
HO AND RAW
cells: Evaluation of changes in RNA and DNA metabolism. J . Neurosci., 1t368-379. Guroff, G. (1985) PC12 cells as a model of neuronal differentiation. In: Cell Culture in the Neuroscience. J.E. Bottenstein, and G. Sato, eds. Plenum Publishing Corporation, New York, pp. 245-272. Harlow, E., and Lane, D. (1988) Antibodies-A Laboratory Manual. Cold Spring Harbor Laboratory, New York. Hatten, M.E., Lynch, M. Rydel, R.E., Sanchez, J . , Joseph-Silverstein, J., Moscatelli, D., and Rifkin, D.B. (1988) In vitro neurite extension by granule neurons is dependent upon astroglial-derived fibroblast growth factor. Dev. Biol., 125:280-289. Ho, P.L., Jakes, R., Northrop, F.D., and Gambarini, A.G. (1988) Pituitary fibroblast growth factors: Immunocharacterization of an acidic component and N-terminal sequence analysis of a truncated basic component. Biochem. Int., 17:973-980. Ho, P.L., Carpenter, M.R., Smillie, L.B., and Gambarini, A.G. (1990) Co-purification of proteases with basic fibroblast growth factor (FGF).Biochem. Biophys. Res. Commun., 170:769-774. Hoshi, H., Kan, M., Mioh, H., Chen, J.K., and McKeehan, W.L. (1988) Phorbol ester reduces number of heparin-binding growth factor receptor in human adult endothelial cells. FASEB J., 2:2797-2800. Kaukel, E., and Hilz, H. (1972) Permeation of dibutyryl cAMP into Hela cells and its conversion to monobutyryl CAMP.Biochem. Biophys. Res. Commun., 46:lOll-1018. Kimelman, D., and Kirschner, M. (1987) Synergistic induction of mesoderm by FGF and TGF B and the identification of an mRNA coding for FGF in the early Xenopus embryo. Cell, 51 2369-877. Lazarovici, P., Dickens, G., Kuzuya, H., and Guroff, G. (1987) Longterm, heterologous down-regulation of the epidermal growth factor receptor in PC12 cells by nerve growth factor. J. Cell Biol., 104:1611-1621. Lee, P.L., Johnson, D.E., Cousens, L.S., Fried, V.A., and Williams, L.T. (1989) Purification and complementary DNA cloning of a receptor for basic fibroblast growth factor. Science, 2 4 5 5 7 4 0 . Norioka, K., Hara, M., Kitami, A,, Hirose, T., Hirose, W., Harigai, M., Suzuki, K., Kawakami, M., Tabata, H., Kawagoe, M., and Nakamura, H. (1987)Inhibitory effect ofrecombinant interleukin 1 a and b on growth of human vascular endothelial cells. Biochem. Biophys. Res. Commun., 145t969-975. Rabe, C.S., Schneider, J . , and McGee Jr., R. (1982) Enhancement of depolarization-dependentneurosecretion from PC12 cells by forskolin-induced elevation of cyclic AMP. J. Cyclic Nucleotide Res., 8:371-384. Reinhold, D.S., and Neet, K.E. (1989) The lack of a role for protein kinase C in neurite extension and in the induction of ornithine decarboxylase by nerve growth factor in PC12 cells. J . Biol. Chem., 264:3538-3544. Rhee, S.G. (1991) Inositol phospholipid-specific phospholipase C: Interaction of the Y, isoform with tyrosine kinase. TIBS, 16t297-301.
Richter-Landsberg, C., and Jastorff, B. (1986) The role of cAMP in nerve growth factor promoted neurite outgrowth in PC12 cells. J . Cell Biol., 102r821-825. Roesler, W.J., Vandenbark, G.R., and Hanson, R.W. (1988) Cyclic AMP and the induction of eukariotic gene transcription. J. Biol. Chem., 263:9063-9066. Rydel, R.E., and Greene, L.A. (1987) Acidic and basic fibroblast growth factor promote stable neurite outgrowth and neuronal differentiation in cultures of PC12 cells. J. Neurosci., 7t3639-3653. Rydel, R.E., and Greene, L.A. (1988) cAMP analogs permit survival and neurite outgrowth in cultures of rat sympathetic and sensory neurons independently of nerve growth factor. Proc. Natl. Acad. Sci. USA, 85r1257-1261. Saksela, O., Moscatelli, D., and Rifkin, D.B. (1987). The opposing effects of basic fibroblast growth factor and transforming growth factor B on the regulation of plasminogen activator activity in capillary endothelial cells. J. Cell Biol., 105:957-963. Schweigerer, L., Malerstein, B., and Gospodarowicz, D. (1987) Tumor necrosis factor inhibits the proliferation of cultured capillary endothelial cells. Biochem. Biophys. Res. Commun., 143t997-1004. Seamon, K.B., and Daly, J.W. (1983) Forskolin, cyclic AMP and cellular physiology. Trends Pharmacol. Sci., 4t120-123. Sigmund, O.L.N., Anderson, D.J., and Stern, R. (1990) Effect of nerve growth factor and fibroblast growth factor on SCGlO and c-fos expression and neurite outgrowth in protein kinase C-depleted PC12 cells. J . Biol. Chem., 265,2257-2261. Sommer, A,, Brewer, M.T., Thompson, R.C., Moscatelli, D., Presta, M., and Rifkin, D.B. (1987) A form of human basic fibroblast growth factor with an extended amino terminus. Biochem. Biophys. Res. Commun., 144r543-550. Taylor, S.S., Buechler, J.A., and Yonemoto, W. (1990) CAMP-dependent protein kinase: Framework for a diverse family of regulatory enzymes. Annu. Rev. Biochem., 59:971-1005. Togari, A,, Baker, D., Dickens, G., and Guroff, G. (1983)The neuritepromoting effect of fibroblast growth factor on PC12 cells. Biochem. Biophys. Res. Commun., 114:1189-1193. Ullrich, A,, and Schlessinger, J . (1990) Signal transduction by receptor with tyrosine kinase activity. Cell, 61:203-212. Wagner, J.A., and DAmore, P.A. (1986) Neurite outgrowth induced by an endothelial cell mitogen isolated from retina. J . Cell. Biol., 103:1363-1367. Walicke, P., Cowan, W.M., Ueno, N., Baird, A,, and Guillemin, R. (1986) Fibroblast growth factor promotes survival of dissociated hippocampal neurons and enhances neurite extension. Proc. Natl. Acad. Sci. USA., 83:3012-3016. Weinmaster, G., and Lemke, G. (1990) Cell-specific cyclic AMP mediated induction of the PDGF receptor. EMBO J., 9t915-920.
Cyclic AMP Potentiates bFGF-Induced Neurite Outgrowth in PC12 Cells PAUL0 LEE HO* AND ISAIAS RAW Centro de Biotecnologia, lnstituto Butantan, CP65, 05504, Sao Paulo, Brazil We report here that basic fibroblast growth factor (bFGF)-elicited neurite outgrowth in PC12 cells is potentiated by dibutyryl cyclic adenosine monophosphate (dbcAMP) or forskolin. This property was also described for nerve growth factor (NGF), suggesting that both NGF and bFGF may share common intracellular events leading to neurite outgrowth and synergism with dbcAMP and forskolin. The synergistic effect of dbcAMP and forskolin i s specific, since treatment of PC12 cells with bFGF and dibutyryl cyclic guanosine monophosphate (dbcGMP) or phorbol ester did not change the neurite outgrowth response of cells treated with bFGF alone. Furthermore, neurite outgrowth depends on cellular adhesion. Increasing adhesion by plate treatment with poly-d-lysine increases the neurite outgrowth elicited by bFGF alone or bFGF plus dbcAMP. O n the other hand, decreasing cellular adhesiveness by plating PC12 cells in semi-solid agarose renders the cells unable to develop neuritic processes. In addition, 3H-methylthymidine incorporation studies showed that bFGF-treated PC12 cells cease growth only when they become fully differentiated after 3-5 days of treatment. In contrast, dbcAMP, which i s a poor differentiation factor, i s able to block cellular growth after 24 hour treatment. These results suggest that when PC12 cells become differentiated, they stop growing. However, growth inhibition does not necessarily lead to differentiation.
Basic fibroblast growth factor (bFGF) is a well known mitogen which has also been demonstrated to possess neurotrophic effects in vitro and in vivo (Burgess and Maciag, 1989). The PC12 clonal line of rat pheochromocytoma cells has become a useful model system to study the mechanisms of action of neurotrophic factors, particularly, nerve growth factor (NGF) (Guroff, 1985). This cell line also differentiates in the presence of FGFrelated proteins (Togari et al., 1983; Delli-Bovi et al., 1988; Burgess and Maciag, 1989), eliciting neurite outgrowth, expression of early specific mRNA like c-fos (Greenberg et al., 19851, expression of neural specific mRNA and enzymatic activities like acetylcholinesterase and ornithine decarboxylase (Togari et al., 1983; Rydel and Greene, 1987). The mechanisms by which these changes occur are still unknown, but some of them are probably mediated by tyrosine kinase activity displayed by bFGF receptor upon ligand binding (Burgess and Maciag, 1989; Lee et al., 1989; Ullrich and Schlessinger, 1990). It has been shown that bFGF action can be modulated negatively or positively depending on cellular phenotype. For instance, mitogenesis of endothelial cells by bFGF is inhibited in vitro by phorbol esters (Doctrow and Folkman, 1987; Hoshi e t al., 19881, tumor necrosis factor (TNF) (Schweigerer et al., 1987; FraterSchroder et al., 19871, Interleukin 1 A and B (Norioka et al., 1987), and transforming growth factor B (TGF B) (Frater-Schroder et al., 1986; Baird and Durkin, 1986; Saksela et al., 1987), whereas the mesoderm induction by bFGF in amphibian embryos is enhanced by TGF B (Kimelman and Kirschner, 1987). There is also a coop0 1992 WILEY-LISS, INC
eration between bFGF and platelet-derived growth factor (PDGF) in promoting division and differentiation of oligodendrocyte type 2-astrocyte (0-2A) progenitor cells (Bogler et al., 1990). Such modulation should be important in regulating cellular proliferation and differentiation. Using PC12 cells, we show here that the neurotrophic effects of bFGF can be enhanced by factors that directly activate CAMP-dependent protein kinases like dibutyryl cyclic adenosine monophosphate (dbcAMP) or indirectly through adenylate cyclase activation by forskolin. We also show that these responses depend on cellular adhesiveness, suggesting a n important role for attachment factors in modulating neurite outgrowth induced by trophic factors.
MATERIALS AND METHODS Materials. Dibutyryladenosine 3'5'-cyclic monophosphate (dbcAMP), dibutyrylguanosine 3'5'-cyclic monophosphate (dbcGMP), forskolin, phorbol 12myristate 13-acetate (PMA), poly-d-lysine and agarose were purchased from Sigma Chemical Co. (St. Louis, MO). A mixture of bovine pituitary bFGF (1-146) and (11-146) were purified according to Ho e t al. (1988). Placental basic FGF purification. Human placental bFGF was purified from placental cellular mass after processing for albumin production. Briefly,
Received January 8,1991; accepted October 16,1991.
*To whom reprint requestdcorrespondence should be addressed.
HO AND RAW
648
a
b
kDa
689 434 29.e
18.9 14.3
Fig. 1. Western-blotting of bovine pituitary bFGF and human placental bFGF. Human placental bFGFs (a)were purified from the cellular mass previously used for albumin production as described in Materials and Methods. The purified form has a higher molecular weight than bovine pituitary bFGF (1-146)(b).Some immunoreactive components of 2 S 3 0 kDa may represent copurified proteases (Ho et al., 1990). The blot was developed using peroxidase conjugates and metal ion enhancement (Harlow and Lane, 1988).
healthy human placentas were collected from Hospital Universitario (Universidade de Sao Paulo, SP) and stored frozen (-40°C) until use. Frozen placentas were thawed overnight in a cold room and ground to 2 4 cm3 pieces. These tissues were then extracted with 10% (v/v) ethanol in 38 mM acetate buffer pH 6.8 for four hours, The supernatant represents the fraction used for albumin production and the pellet represents the cellular mass used for growth factor purification. Cellular mass from 10 Kg of frozen placentas were extracted with 50 mM phosphate buffer pH 7.0 in a Waring blender and centrifuged. The crude extract (supernatant) was chromatographed in a S-Sepharose column (Pharmacia, Uppsala, Sweden). The active fractions eluted with 0.8 M NaC1, were pooled, diluted to 0.15 M NaCl and pumped into a Heparin-Sepharose column. The adsorbed materials were eluted stepwise with 0.8 M, 1.2 M and 2.0 M NaC1. The bFGF was eluted in the last wash. The purified placental bFGF (Fig. 1)has a molecular weight higher than bovine pituitary bFGF (1-146) and probably represents the bFGF (-11-146) form already described (Sommer et al., 1987). The antisera anti-bFGF was kindly provided by Dr. D. Rifkin (New York University School of Medicine, NY). Cell culture. PC12 cells were a generous gift of Dr. A. Howllet (Saint Louis University, St. Louis, MO). Cells were grown in Dulbecco’s Modified Eagle’s Medium (DME) supplemented with 10% fetal calf serum (FCS) a t 37°C in a humidified atmosphere of 5% CO,. Subcultures were done in a 1:4 split when cells became confluent. Quantitation of neurite outgrowth. Cells (2.5 x lo4 cells/ml) were plated directly a 12-well microplate (Corning, NY) in DME plus 10% FCS or in poly-d-lysine-(10 p,g/ml) treated 12-well microplates. Twelve hours later, the culture was maintained or
changed to DME plus 2.5% FCS supplemented or not with bovine pituitary bFGF, human placental bFGF, dbcAMP, dbcGMP, forskolin or PMA, as indicated in the figure legends. No difference in neurite outgrowth response was observed whether we performed the assays in 10% or 2.5% FCS. In all cases, medium was changed every two days with the addition of fresh supplements. A neurite was identified as a process whose length was equal or greater than one cell body length. Isolated cells or groups containing no more than 10 clearly distinguishable cells were considered for scoring neurite outgrowth. In cells presenting two or more neurites, only the longest one was scored. One to two hundred cells per well were randomly observed under phase-contrast microscope and the percentage of cells with neurites was calculated. Distribution of the cellular population according to neurite length. Cells were separated into seven classes according to neurite length: cells without neurites (0);cells with neurites smaller than one cell body length (- 1); cells presenting neurites corresponding to one (11, two (21, three (31, four (41, and more than four (+4) times the cell body length. The conditions were the same as described above. Plating in semi-solid agarose. Cells (1-2 x lo4 cells/well) were plated directly in untreated 24-well microplates (Corning, NY) in DME plus 10% FCS. On the next day, the medium was removed and melted 0.6% agarose was added over the cells. On the solidified agarose, we seeded a second cell layer (1-2 x lo4 cellslwell) and again, another layer of agarose. DME plus 10% FCS supplemented or not with human placental bFGF andlor dbcAMP was added over the second layer of agarose. Media and supplements were changed every two days. We followed the differentiation of the cells plated on a substrate (first cell layer) or without a substrate (second cell layer in semi-solid agarose) on the different microscope focus planes. Cellular proliferation assay. Cells (2 x lo4 cells/ well) were plated in untreated 24-well microplates in DME plus 10% FCS. Twelve hours later, we added bFGF and/or dbcAMP (time 0 of treatment). The media and supplements were changed every two days. Cellular proliferation was followed by 3H-methyl-thymidine uptake into DNA. Briefly, 3H-methyl-thymidine (0.5 pCi/ml, 1 x M) was added 24 hours before collecting the cells. The cells were treated with cold trichloroacetic acid (TCA) and lysed with sodium hydroxide which was further adsorbed to a filter paper, washed with cold TCA, then ethanol, dried and counted by scintillography .
RESULTS Potentiation of neurite o u t g r o w t h b y d b c A M P Based on previous data showing that bFGF and NGF share some common responses in promoting differentiation of PC12 cells (Wagner and D’Amore, 1986; Rydel and Greene, 1987) and that CAMP analogs are able to increase neurite outgrowth elicited by NGF in this cell line (Richter-Landsberg and Jastorff, 1986),we investigated whether this last property is also shared by bFGF. For this purpose, we used a human placental bFGF purified from the cellular mass previously used for albumin production (Fig. 1).
NEUROTROPHIC EFFECTS OF CAMPAND bFGF IN PC12 CELLS
Fig. 2. Effect of dbcAMP on bFGF-induced neurite outgrowth. PC12 cells in 2.5% FCS were exposed to placental bFGF (5ng/ml), dbcAMP (1mM) or placental bFGF (5 ngiml) plus dbcAMP (1mM) for 24 hours (a,c,e,g)or 96 hours (b,d,f,h).(a,b) untreated control; (c,d) dbcAMP; (e,f)placental bFGF; (g,h) placental bFGF plus dbcAMP. Bar i n a = 15 km, and applies to b-h.
649
HO AND RAW
650
1 I
1
50
5
3 I
I
DAYS
I
h 0
ll
4:
S
V
I 1
I
I
I
,
,
2
3
4
5
6
DAYS
Fig. 3. Kinetics of neurite formation. PC12 cells in 2.5%FCS were exposed to (0-0) 5 ng/ml placental bFGF; (-0) 1 mM dbcAMP; or ( A-A f 5 ng/ml placental bFGF plus 1 mM dbcAMP. Untreated control cells did not develop neurites at all (not shown). When not indicated, the error was less than 1%.
v)
n
-I
0
+
A 3 w 0 1
:b: 5
s 0 -1 1 2 3 4 + 4
0 -1 1 2 3 4 + 4
0 -1 1 2 3 4 4 INDEXES
Figure 2 shows that simultaneous treatment of PC12 4. Distribution of the cellular population according to neurite cells with bFGF and dbcAMP induces higher neurite Fig. length. PC12 cells in 2.5% FCS were exposed to 5 ngiml placental outgrowth density than bFGF or dbcAMP alone. This bFGF, 1mM dbcAMP, or 5 ngiml placental bFGF plus 1mM dbcAMP. effect can be detected either after 24 (Fig. 2g) or 96 (Fig. Cells in days 1, 3, 5 were classified according to neurite length (see 2h) hours of treatment. In the first 24 hours, the treat- Materials and Methods) and expressed as percentage of the total numment with dbcAMP or bFGF alone is equally ineffi- ber of cells analysed. Untreated control cells did not develop neurites cient, inducing few processes. However, after long term at all (not shown). When not indicated, the error was less than 1%. incubation, bFGF induces a higher number of neurites than dbcAMP (Figs. 2d,f and 5b,c). Quantitative analysis of cells bearing neurites show are present. These results show that PC12 cells present the synergistic effects of dbcAMP and bFGF (Fig. 3). different kinetics of neurite outgrowth in response to The percentage of cells with neurites in those treated dbcAMP, bFGF, and dbcAMP plus bFGF. with dbcAMP and bFGF in days 1 , 2 , and 3 are higher Effects of forskolin than the sum of the percentage of cells treated alone with each factor. Figure 3 also shows that the percentdbcAMP has to be metabolically activated before age of cells bearing neurites is low in dbcAMP treated binding to cAMP receptor proteins (Kaulkel and Hilz, cells, with plateau values of 20-25%. 1972). To exclude non-related side effects caused by Analysis of neurite length on days 1,3, and 5 clearly metabolites, such as butyric acid, generated during this shows different cellular distribution according to the activation, we used forskolin to confirm the synergistic treatment used (Fig. 4). Cells treated for 24 hours with effects of cAMP and bFGF in eliciting neurite outbFGF show high percentage of cells without neurites or growth. Forskolin is a known diterpene that activates with processes smaller than one cell body length. How- adenylate cyclase (Seamon and Daly, 1983).Treatment ever, with increasing incubation time, the percentage of PC 12 cells with forskolin raises the intracellular of cells without neurites diminishes and there is an cAMP level (Rabe et al., 1982) and has little effects on increase in the percentage of cells with neurites equal neurite outgrowth in cells continuously incubated for or greater than one cell body length. In contrast, the 144 hours (Fig. 5e). However, when bFGF and forskolin pattern of distribution of cells treated with dbcAMP is were added simultaneously, a potentiation of bFGFquite different; they did not change significantly with elicited neurite outgrowth was found (Fig. 50, confirmincubation time, being similar on days 1 and 5. The ing the involvement of intracellular cAMP in this repercentage of cells bearing no neurites or with neurites sponse. Although bFGF is able to induce neurite less than one cell body length remains high in all the outgrowth, the neurites are often short and linear, days examined. The synergistic effect of dbcAMP plus whereas in cells incubated with both bFGF and bFGF can also be seen in Figure 4. The cellular distri- dbcAMP or forskolin, the neurit.es are often longer and bution pattern changed during continuous treatment. branched and cell bodies are hyperthrophied at the inOn day 1there are cells bearing neurites with 1 , 2 , and cubation times analysed (Fig. 5). 3 cell body length, although a high percentage of cells Effects of dbcGMP and PMA with indexes 0 and -1 are still found. However, the percentage of cells with indexes 0 and -1 on the fifth Since dbcAMP and forskolin increase intracellular day is insignificant, and high percentage of cells with CAMP,the potentiation effects of these drugs on bFGFneurites with 2,3,4 and greater than 4 cell body length induced neurite outgrowth is probably mediated
NEUROTROPHIC EFFECTS OF CAMPAND bFGF IN PC12 CELLS
651
Fig. 5. Branching effects of dbcAMP or forskolin on bFGF-induced neurite outgrowth. PC12 cells i n 2.5%FCS were exposed to placental bFGF (5 ngiml), dbcAMP (1 mM), forskolin (10 pM),placental bFGF (5 ngiml) plus dbcAMP (1 mM) or placental bFGF (5 ngiml) plus forskolin (10 pM) for 144 hours. (a) untreated control; (b) placental bFGF; (c) dbcAMP; (d) placental bFGF plus dbcAMP ( e ) forskolin; (0 placental bFGF plus forskolin. Bar in a = 15 pm, and applies to b-f.
through CAMP-dependent protein kinase (Roesler et al., 1988; Taylor e t al., 1990). In order to verify ifthis effect could also be promoted by cGMP-dependent protein kinases or protein kinase C (PKC), we treated PC12 cells with 1mM dbcGMP or 10 nM PMA, alone or with bFGF. Neither dbcGMP nor PMA altered the neurite outgrowth elicited by bFGF (results not shown). Thus, we can conclude that this feature is specific to intracellular increase of CAMP, probably through the activation of protein kinase A.
Neurite outgrowth depends on cell adhesion The neurotrophic effects of bFGF in PC12 cells plated on plastic substrate can be easily followed and the neu-
rites quantitatively scored in short term incubation (6-8 days). However, the increasing clamping effects combined with low adhesion of PC12 cells on plastic after long term incubation (beyond 8 days) made the experiments difficult to follow and control. Culture plate treatment with poly-d-lysine allowed us to better control long term incubations and examine the effect of increased adhesion in bFGF-elicited neurites in these cells. Cells grown on poly-d-lysine treated plates show increased response to the neurotrophic effects of bFGF when compared to control cells on untreated plastic plates (Fig. 6e,f). dbcAMP alone is able to induce neurite outgrowth (Fig. 6c,d). The synergistic effects of bFGF and dbcAMP is also seen in Figure 6g,h. The long
652
HO AND RAW
Fig, 6. Effects of increased adhesiveness by plating PC12 cells on poly-d-lysine. PC12 cells in 10% FCS were exposed to placental bFGF (5 ngiml), dbcAMP (ImM) or placental bFGF (5 ngiml) plus dbcAMP (1 mM) for 96 hours in untreated (a,c , e, g) or poly-d-lysine treated culture plates (b,d, f, h). (a,b) untreated control; (c,d) dbcAMP; (e,f)placental bFGF; (g,h) placental bFGF plus dbcAMP. Bar in a = 15 km, and applies to b-h.
NEUROTROPHIC EFFECTS OF CAMPAND bFGF IN PC12 CELLS
term incubation effects of the supplements were examined under these conditions. Surprisingly, on day 10, we start to notice significant cell death in bFGF-treated cells. Fifty percent of the cellular population treated with bFGF remained on day 16. Neither dbcAMP alone nor combined with bFGF showed this feature. It seems that dbcAMP prevents cellular mortality induced by long term incubation with bFGF (results not shown). On the other hand, we also decreased adhesiveness by plating PC12 cells in semi-solid agarose, which renders the cells unresponsive to bFGF and/or dbcAMP (Fig. 71, in contrast to control cells in the same well plated on untreated plastic substrate.
Differentiation versus proliferation In order to evaluate if the promotion of neuronal differentiation of PC12 cells by bFGF is concomitant with growth inhibition, 3H-methyl-thymidine uptake studies were performed, as shown in Figure 8. Treatment with bFGF did not alter cellular proliferation during the first 4 days. After this time, cell growth was inhibited. These results parallel those showing that bFGF did not promote neurite outgrowth of PC12 cells in short term treatment (1-3 days), but only after 3-5 days of continuing incubation (Figs. 2-5). Furthermore, the combination of dbcAMP and bFGF, which induces neuritic processes a s soon a s day 1 of treatment (Figs. 2-4), inhibited the 3H-methyl-thymidine uptake of the cells in this condition. These data suggest that when the cells become differentiated, their growth is inhibited. However, dbcAMP, which alone is a poor differentiation factor in this system (Figs. 2-5), is able to stop cell division within 24 hours of treatment (Fig. 8). DISCUSSION Using PC12 cells as a model system for the study of neuronal differentiation, we investigated whether neurite outgrowth elicited by bFGF can be positively modulated by dbcAMP or forskolin, a drug that increases intracellular cAMP through adenylate cyclase activation. Our results show that bFGF-elicited neurite outgrowth is potentiated by dbcAMP or forskolin. This property was also described for NGF (Richter-Landsberg and Jastorff, 1986), suggesting that NGF and bFGF may share common intracellular events leading to neurite outgrowth. Since simultaneous addition of bFGF and dbcAMP or forskolin potentiate neurite outgrowth in PC12 cells synergistically, we can conclude that they probably act through different mechanisms. This observation is supported by recent reports showing that NGF, aFGF, and bFGF regulate neurite outgrowth in PC12 cells via both PKC and CAMP-independent mechanisms (Reinhold and Neet, 1989; Damon et al., 1990; Sigmund e t al., 1990). However, other responses elicited by FGFs and NGF can be totally or partially dependent on PKC and CAMP-dependent protein kinase. For instance, ornithine decarboxylase gene expression in bFGF and aFGF but not NGF treated cells is dependent on both PKC and CAMP-dependent protein kinase systems (Reinhold and Neet, 1989; Damon et al., 1990), whereas the expression of d5 gene is independent and the expression of d2 gene is partially dependent on PKC and CAMP-dependent protein kinase pathways (Damon et al., 1990). The expression of other genes like SCGlO was also shown to be indepen-
653
dent of PKC, while the expression of c-fos mRNA by NGF but not bFGF is partially mediated by PKC (Sigmund et al., 1990). These results show that some cellular responses are differentially mediated by bFGF and NGF and may reflect differences in receptor signalling. The synergistic effect of dbcAMP or forskolin is specific, since treatment of PC12 cells with bFGF and dbcGMP or PMA did not change the neurite outgrowth response of cells treated with bFGF alone (results not shown). In contrast, PMA is able to inhibit the mitogenic effects of bFGF in endothelial cells (Doctrow and Folkman, 1987; Hoshi et al., 19881, showing that the same trophic factors may have different influences in different cells. It is also worth considering the possibility that trophic factors like neurotransmiters or peptides, which increase intracellular cAMP levels, may exert regulatory roles in promoting neural differentiation together with bFGF and NGF. It was shown that cAMP analogs by themselves are sufficient to promote survival and neurite outgrowth in cultures of rat sympathetic and sensory neurons (Rydel and Greene, 19881, supporting the idea of a n important role for cAMP in neuronal differentiation. The experiments designed to increase or decrease adhesiveness (Figs. 6,7) showed the anchorage dependence of PC12 cells to a substrate in response to bFGF andlor dbcAMP, suggesting a role for attachment factors in modulating the action of neurotrophic factors. Dependence of culture substrate for cell survival and neurite outgrowth was also seen in other types of bFGF-treated nervous cells, like hippocampal and cerebellar granule neurons (Walicke et al., 1986; Hatten et al., 1988). Recently, it was shown that EGF receptors activated by ligand binding stimulate tyrosine phosphorylation of phospholipase CY 1 (PLCY1). In vitro, both phosphorylated and unphosphorylated PLCY 1are able to hydrolyse inositol phospholipids to inositol phosphate and diacylglicerol. However, the enzymatic activity of the non-phosphorylated form is inhibited by profilin, a n actin binding protein, whereas the phosphorylated form is not (Goldschmidt-Clermont et al., 1991; Rhee, 1991). It was also shown that another actin-binding protein, tensin, possesses a SH2 domain responsible for phosphotyrosine interaction (Davis et al., 1991, Cantley et al., 1991). These results make a connection between cytoskeleton and receptor signalling. Like EGF and PDGF receptors, FGF receptors have intrinsic tyrosine kinase activity and phosphorylate several protein targets, among them, PLCY 1(Burgess et al., 1990). So, i t is conceivable that changing adhesiveness of PC12 cells also changes the composition, structure and dynamics of cytoskeleton which, in turn, modify the interactions of cytoplasmic signalling proteins activated by FGF receptor andtor protein kinase A, resulting in decrease or increase of bFGF- and CAMP-induced response. The DNA synthesis of bFGF-treated PC12 cells appears to continue unaltered for about 4 days (Fig. 4). Then, the synthesis of DNA shows a progressive decrease. This was also described for NGF with similar kinetics. The decrease in DNA synthesis parallels with cell division decrease (Gunning et al., 1981). It was further shown t h a t incubation of PC12 cells with NGF progressively down-regulates EGF receptors, diminish-
654
HO AND RAW
Fig. 7. Effects of decreased adhesiveness by plating PC12 cells in semi-solid agarose. PC12 cells in 10% FCS were plated on a substrate (a, c, e, g) or in semi-solid agarose (b, d, f, h) and exposed to placental bFGF (5 ngiml), dbcAMP (1mM) or placental bFGF (5ngiml) plus dbcAMP (1 mM) for 96 hours, according to Materials and Methods. (a,b) untreated control; (c,d) dbcAMP; (e,O placental bFGF; (g,h) placental bFGF plus dbcAMP. Bar in a = 15 Km, and applies to b-h.
NEUROTROPHIC EFFECTS OF CAMPAND bFGF IN PC12 CELLS
655
eration of PDGF-treated rat Schwann cells (Weinmaster and Lemke, 1990). The increase in intracellular cAMP induces PDGF receptor gene transcription and translation, which accounts, at least in part, for this synergism (Weinmaster and Lemke, 1990). Whether cAMP increases the bFGF receptors in PC12 cells, which would in part explain our results, is unknown. This possibility should be considered.
ACKNOWLEDGMENTS P.L.H. would like to thank Drs. E. Kimura, H.A. Armelin, A.G. Gambarini, M.C.S. Armelin, J.C.C. Maia, K.M. Rocha, L. Rameh, J.F. Dias (Universidade de Sao Paulo, SP, Brazil), Drs. A.M. Silva, L.C.C. Leite, M. Katz (Instituto Butantan, SP, Brazil), Dr. D.B. Rifkin (New York University, NY, USA) and Dr. A. Howllet (Saint Louis University, MO, USA) for their helpful discussions, suggestions or for providing reagents. This work was supported by FAPESP (901 0168-5) and CNPq (410061/90-5) Brazilian grants to P.L.H. and FINEP to I.R.
LITERATURE CITED
2
3
4
5
6 DAYS
Fig. 8. Growth response of PC12 cells exposed to placental bFGF and/or dbcAMP, measured by 3H-methyl-thymidine uptake into DNA. PC12 cells in 10% FCS were exposed to (U 5 ng/ml ) of placental bFGF; (0-0) 1mM dbcAMP or (M 5) ngiml of placental bFGF represents untreated control. The results plus ImM dbcAMP.1-( were expressed by cpmiwell (a)or as % inhibition related to untreated control (b).
ing EGF binding as soon as 12 hours, being most of the binding capacity inhibited within 4 days. It was suggested that this heterologous down-regulation is part of the mechanism by which differentiating cells become insensitive to mitogens (EGF) present in serum (Lazarovici et al., 1987).This mechanism probably occurs in bFGF-treated PC12 cells. dbcAMP, which is a poor differentiation factor (Figs. 2-45), is able to decrease DNA synthesis (Fig. 8) as soon as 24 hours after incubation, in contrast to bFGF-treated cells. This inhibition is maintained over the days examined, although no significant neurite extension could be seen. The inhibition of DNA synthesis is greater in cells supplemented with both bFGF and dbcAMP. These results, taken together with those showing neurite outgrowth induced by bFGF and/or dbcAMP (Figs. 2-5), led us to conclude that when PC12 cells become fully differentiated, they stop growing; however, when their growth is inhibited, they are not necessarily differentiated. It is likely that growth inhibition facilitates the action of neurotrophic factors. The mechanism by which cAMP increases the neurite outgrowth promoted by bFGF is unknown. It was shown that cAMP synergistically stimulates the prolif-
Baird, A., and Durkin, T. (1986) Inhibition of endothelial cell proliferation by type B transforming growth factor: Interaction with acidic and basic fibroblast growth factor. Biochem. Biophys. Res. Commun., 138:47&482. Bogler, O., Wren, D., Barnett, S.C., Land, H., and Noble, M. (1990) Cooperation between two growth factors promotes extended self renewal and inhibits differentiation of oligodendrocyte-type 2 astrocyte (0-2A) progenitor cells. Proc. Natl. Acad. Sci. USA, 87r63686372. Burgess, W.H., and Maciag, T. (1989)The heparin-binding (fibroblast) growth factor family of proteins. Annu. Rev. Biochem., 58r575-606. Burgess, W.H., Dionne, C.A., Kaplow, J., Mudd, R., Friesel, R., Zilberstain, A,, Schlessinger, J.,and Jaye, M. 11990)Characterization and cDNA cloning of phospholipase C-Y, a major substrate for heparinbinding growth factor 1 (acidic fibroblast growth factorbactivated tyrosine kinase. Moll. Cell. Biol., 10r4770-4777. Cantley, L.C., Auger, K.R., Carpenter, C., Duckworth, B., Graziani, A,, Kapeller, R., and Soltoff, S. (1991) Oncogenes and signal transduction. Cell, 64r281-302. Damon, D.H., DAmore, P.A., and Wagner, J.A. (1990) Nerve growth factor and fibroblast growth factor regulate neurite outgrowth and gene expression in PC12 cells via both protein kinase C- and CAMPindependent mechanisms. J. Cell. Biol., lIOr1333-1339. Davis, S., Lu, M.L., Lo, S.H., Lin, S., Butler, J.A., Druker, B.J., Roberts, T.M., An, &., Chen, L.B. (1991) Presence of an SH2 domain in the actin-binding protein tensin. Science, 252r712-715. Delli-Bovi, P., Curatola, A.M., Newman, K.M., Sato, Y., Moscatelli, D., Hewick, R.M., Rifkin, D.B., and Basilico, C. (1988) Processing, secretion, and biological properties of a novel growth factor of the fibroblast growth factor family with oncogenic potential. Mol. Cell. Biol., 829334941. Doctrow, S.R., and Folkman, J. (1987) Protein kinase C activators suppress stimulation of capillary endothelial cell growth by angiogenic endothelial mitogen. J . Cell Biol., 104t679-687. Frater-Schroder,M.,Muller,G.,Birchmeier, W., andBohlen,P. (1986) Transforming growth factor B inhibits endothelial cell proliferation. Biochem. Biophys. Res. Commun., 137.295-302. Frater-Schroder, M., Risau, W., Halmman, R., Gautschi, P., and Bohlen, P. (1987) Tumor necrosis factor a, a potent inhibitor of endothelial cells in uitro.. is aneioeenic in uiuo. Proc. Natl. Acad. Sci. USA., 845277-5281. Goldschmidt-Clermont, P.J., Kim, J.W., Machesky, L.M., Rhee, S.G., and Pollard, T.D. (1991) Regulation of phospholipase C-yl by profilin and tyrosine phosphorylation. Science, 251 tlZ31-1233. Greenberg, M.E., Greene, L.A., and Ziff, E.G. (1985) Nerve growth factor and epidermal growth factor induce rapid transient changes in protooncogene transcription in PC12 cells. J. Biol. Chem., 260r14101-14110. Gunning, P.W., Landreth, G.E., Layer, P., Ignatius, M., and Shooter, E.M. (1981) Nerve growth factor induced differentiation of PC12 Y
l
656
HO AND RAW
cells: Evaluation of changes in RNA and DNA metabolism. J . Neurosci., 1t368-379. Guroff, G. (1985) PC12 cells as a model of neuronal differentiation. In: Cell Culture in the Neuroscience. J.E. Bottenstein, and G. Sato, eds. Plenum Publishing Corporation, New York, pp. 245-272. Harlow, E., and Lane, D. (1988) Antibodies-A Laboratory Manual. Cold Spring Harbor Laboratory, New York. Hatten, M.E., Lynch, M. Rydel, R.E., Sanchez, J . , Joseph-Silverstein, J., Moscatelli, D., and Rifkin, D.B. (1988) In vitro neurite extension by granule neurons is dependent upon astroglial-derived fibroblast growth factor. Dev. Biol., 125:280-289. Ho, P.L., Jakes, R., Northrop, F.D., and Gambarini, A.G. (1988) Pituitary fibroblast growth factors: Immunocharacterization of an acidic component and N-terminal sequence analysis of a truncated basic component. Biochem. Int., 17:973-980. Ho, P.L., Carpenter, M.R., Smillie, L.B., and Gambarini, A.G. (1990) Co-purification of proteases with basic fibroblast growth factor (FGF).Biochem. Biophys. Res. Commun., 170:769-774. Hoshi, H., Kan, M., Mioh, H., Chen, J.K., and McKeehan, W.L. (1988) Phorbol ester reduces number of heparin-binding growth factor receptor in human adult endothelial cells. FASEB J., 2:2797-2800. Kaukel, E., and Hilz, H. (1972) Permeation of dibutyryl cAMP into Hela cells and its conversion to monobutyryl CAMP.Biochem. Biophys. Res. Commun., 46:lOll-1018. Kimelman, D., and Kirschner, M. (1987) Synergistic induction of mesoderm by FGF and TGF B and the identification of an mRNA coding for FGF in the early Xenopus embryo. Cell, 51 2369-877. Lazarovici, P., Dickens, G., Kuzuya, H., and Guroff, G. (1987) Longterm, heterologous down-regulation of the epidermal growth factor receptor in PC12 cells by nerve growth factor. J. Cell Biol., 104:1611-1621. Lee, P.L., Johnson, D.E., Cousens, L.S., Fried, V.A., and Williams, L.T. (1989) Purification and complementary DNA cloning of a receptor for basic fibroblast growth factor. Science, 2 4 5 5 7 4 0 . Norioka, K., Hara, M., Kitami, A,, Hirose, T., Hirose, W., Harigai, M., Suzuki, K., Kawakami, M., Tabata, H., Kawagoe, M., and Nakamura, H. (1987)Inhibitory effect ofrecombinant interleukin 1 a and b on growth of human vascular endothelial cells. Biochem. Biophys. Res. Commun., 145t969-975. Rabe, C.S., Schneider, J . , and McGee Jr., R. (1982) Enhancement of depolarization-dependentneurosecretion from PC12 cells by forskolin-induced elevation of cyclic AMP. J. Cyclic Nucleotide Res., 8:371-384. Reinhold, D.S., and Neet, K.E. (1989) The lack of a role for protein kinase C in neurite extension and in the induction of ornithine decarboxylase by nerve growth factor in PC12 cells. J . Biol. Chem., 264:3538-3544. Rhee, S.G. (1991) Inositol phospholipid-specific phospholipase C: Interaction of the Y, isoform with tyrosine kinase. TIBS, 16t297-301.
Richter-Landsberg, C., and Jastorff, B. (1986) The role of cAMP in nerve growth factor promoted neurite outgrowth in PC12 cells. J . Cell Biol., 102r821-825. Roesler, W.J., Vandenbark, G.R., and Hanson, R.W. (1988) Cyclic AMP and the induction of eukariotic gene transcription. J. Biol. Chem., 263:9063-9066. Rydel, R.E., and Greene, L.A. (1987) Acidic and basic fibroblast growth factor promote stable neurite outgrowth and neuronal differentiation in cultures of PC12 cells. J. Neurosci., 7t3639-3653. Rydel, R.E., and Greene, L.A. (1988) cAMP analogs permit survival and neurite outgrowth in cultures of rat sympathetic and sensory neurons independently of nerve growth factor. Proc. Natl. Acad. Sci. USA, 85r1257-1261. Saksela, O., Moscatelli, D., and Rifkin, D.B. (1987). The opposing effects of basic fibroblast growth factor and transforming growth factor B on the regulation of plasminogen activator activity in capillary endothelial cells. J. Cell Biol., 105:957-963. Schweigerer, L., Malerstein, B., and Gospodarowicz, D. (1987) Tumor necrosis factor inhibits the proliferation of cultured capillary endothelial cells. Biochem. Biophys. Res. Commun., 143t997-1004. Seamon, K.B., and Daly, J.W. (1983) Forskolin, cyclic AMP and cellular physiology. Trends Pharmacol. Sci., 4t120-123. Sigmund, O.L.N., Anderson, D.J., and Stern, R. (1990) Effect of nerve growth factor and fibroblast growth factor on SCGlO and c-fos expression and neurite outgrowth in protein kinase C-depleted PC12 cells. J . Biol. Chem., 265,2257-2261. Sommer, A,, Brewer, M.T., Thompson, R.C., Moscatelli, D., Presta, M., and Rifkin, D.B. (1987) A form of human basic fibroblast growth factor with an extended amino terminus. Biochem. Biophys. Res. Commun., 144r543-550. Taylor, S.S., Buechler, J.A., and Yonemoto, W. (1990) CAMP-dependent protein kinase: Framework for a diverse family of regulatory enzymes. Annu. Rev. Biochem., 59:971-1005. Togari, A,, Baker, D., Dickens, G., and Guroff, G. (1983)The neuritepromoting effect of fibroblast growth factor on PC12 cells. Biochem. Biophys. Res. Commun., 114:1189-1193. Ullrich, A,, and Schlessinger, J . (1990) Signal transduction by receptor with tyrosine kinase activity. Cell, 61:203-212. Wagner, J.A., and DAmore, P.A. (1986) Neurite outgrowth induced by an endothelial cell mitogen isolated from retina. J . Cell. Biol., 103:1363-1367. Walicke, P., Cowan, W.M., Ueno, N., Baird, A,, and Guillemin, R. (1986) Fibroblast growth factor promotes survival of dissociated hippocampal neurons and enhances neurite extension. Proc. Natl. Acad. Sci. USA., 83:3012-3016. Weinmaster, G., and Lemke, G. (1990) Cell-specific cyclic AMP mediated induction of the PDGF receptor. EMBO J., 9t915-920.
