Basic fibroblast growth factor rescues CNS neurons from cell death caused by high oxygen atmosphere in culture

261

Brain Research, 599 (1992) 261-271 © 1992 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.00

BRES 18379

Basic fibroblast growth factor rescues CNS neurons from cell death caused by high oxygen atmosphere in culture Yasushi Enokido a Yukio Akaneya a,c, Michio Niinobe b, Katsuhiko Mikoshiba b,d and Hiroshi Hatanaka a "Division of Protein Biosynthesis and b Dicision of Regulation of Macromolecular Function, Institute for Protein Research, Osaka Unicersity, Osaka (Japan), c Department of Neurology, Kinki University School of Medicine, Osaka (Japan) and d Institute of Medical Sciences, The Unicersity of Tokyo, Tokyo (Japan) (Accepted 28 July 1992)

Key words: Oxygen radical; Nerve growth factor; Basal forebrain; Neuronal differentiation; Neurotrophic factor; Cholinergic neuron

In the present study, we cultured rat CNS neurons and tested the neurotrophic support provided by basic fibroblast growth factor (bFGF) to prevent the oxygen-induced neuronal cell death. When rat basal forebrain (septum and vertical limb of diagonal band of Broca) cells of embryonic day 20 were cultured in a serum-free medium containing 5/~M cytosine arabinoside in a 50% oxygen atmosphere, the neuronal cells, which were immunostained by an anti-microtubule-associated protein 2 (MAP2) antibody, gradually died after 1 day in culture. After 3.5 days in culture, only 2-5% of neuronal cells survived. This oxygen-induced cell death of cultured basal forebrain neurons was reversed by the addition of bFGF at a concentration of 100 ng/ml. This cell-saving effect was dose-dependent, and the ED50 value was 12 ng/ml. Nerve growth factor (NGF) and insulin-like growth factor II could not prevent cell death. The activity of choline acetyltransferase was also maintained when bFGF was present in the basal forebrain culture. Viable astroglial cells, which were immunostained by an anti-glial fibrillary acidic protein, accounted for a few percent of the total number of cells after 3 days in culture both with and without 100 ng/ml of bFGF. The survival-enhancing effect of bFGF was observed not only in basal forebrain neurons but also in neocortical and hippocampal neurons. However, the sensitivity to oxygen toxicity of cultured neurons from the 3 CNS regions varied greatly. The neocortical neurons were the most sensitive to oxidative stress, while the hippocampal neurons were the most resistant. These results suggest that bFGF plays an important role in saving neuronal cells from oxidative stress during their long life without division.

INTRODUCTION

The brain is one of the highest energy-consuming organs in the mammalian body. It is exclusively dependent on the aerobic energy metabolism of oxygen and glucose. This characteristic energy metabolism of the brain changes dramatically during its maturation 27. This change is deeply related to the construction of neuronal architecture and the development of neuronal activity. This, in turn, means that neuronal cells can be constantly exposed to oxidative stress during their long, non-dividing life. It is also known that the brain readily undergoes oxidative damage as a result of cerebrovascular injury and aging 7'21. In previous studies, when the rat embryonic CNS neurons and PC12h pheochromocytoma cells were cultured under a high oxygen atmo-

sphere, the cells could not s u r v i v e 13'30. Furthermore, we demonstrated that the tolerance for oxidative stress of neurons in vitro increases during maturation, and that nerve growth factor (NGF) and basic fibroblast growth factor (bFGF) could prevent the death of PC12h cells following oxidative events. These results suggest that the culture system using PC12h cells and CNS neurons were useful to study the mechanisms not only of the neuronal cell death induced by oxidative stress but also of survival-promoting effects of neurotrophic factors. FGFs are potent mitogens that promote the proliferation, migration and differentiation of a wide variety of cell types 28. They are classified in two categories according to their different isoelectric points, that is acidic and basic FGF. Both aFGF and bFGF have

Correspondence: H. Hatanaka, Division of Protein Biosynthesis, Institute for Protein Research, Osaka University, 3-2 Yamadaoka, Suita, Osaka 565, Japan. Fax: (81) (6) 876-2533.

262 been purified from brain and are 55% homologous in terms of their total amino acid sequences ~s. Recently, it has been reported that a F G F and b F G F have neurotrophic effects and promote the differentiation and cell survival of many types of PNS 44'5° and C N S 15"22'36'49'50"52 n e u r o n s in vitro. Furthermore, it has been shown that these factors exist in PNS and CNS neurons 12'~6'4°'56 and their receptors are distributed to the similer neuronal populations in v i v o 24'54. FGFs have been found to have certain characteristics 6"s8. First, FGFs are not secreted via the normal Golgi pathway because they have no signal sequence. Second, FGFs bind to heparan sulfate proteoglycans in the extracellular matrix and then form the high affinity receptor complexes for FGFs. Third, b F G F is thought to be transferred to cellular nuclei. From these results, although the trophic effect of b F G F have been well known, it is difficult to speculate how cytosolic proteins, FGFs, function as an extracellular growth factor. Moreover, as it is also known that FGFs stimulate the proliferation and maturation of glial cells 9A4, their trophic support has to be considered. In the present study, we demonstrated the cell-saving effect of b F G F on CNS neurons damaged by oxidative stress. These results suggest that b F G F plays an important role in preventing cell death from oxidative stress in CNS neurons. MATERIALS

AND METHODS

Chemicals and determination of enzyme activity N G F was prepared as 2.5S-form from male m o u s e submandibular glands according to the m e t h o d s of Bocchini and Angeletti 5, with some modifications by Suda et al. 47. H u m a n recombinant b F G F was a gift from Takeda Pharmaceutical Co. H u m a n recombinant insulinlike growth factor II (IGF II) was a gift from Daiichi Pharmaceutical Co. Choline acetyltransferase (CHAT) activity was determined by the method of F o n n u m ~7, with modifications described previously 2°. Protein was determined by the method of Lowry et al. 31 using bovine y-globulin as a standard.

Preparation of cultured cells Tissue fragments of the basal forebrain area (containing septum and vertical limb of diagonal band of Broca) as well as neocortical and hippocampal areas were dissected from embryonic E20 rat (Wistar ST, both sexes, Shizuoka) brains. Dissection was carried out on ice u n d e r a Nikon stereomicroscope with a fiberoptic light source. T h e location of these areas was determined by using a developmental rat atlas 45. The day of vaginal plug was designated fetal day 1. The m e t h o d s of preparing cultured cells have been described previously z°. In brief, tissue fragments were added to 10 ml of freshly prepared Ca2+,Mg2+-free phosphate-buffered saline (PBS) containing papain (Worthington, 90 units), DNase I (Sigma, 2,000 units), D,L-cysteine-HC1 (Sigma, 2 mg), bovine serum albumin (recrystallized, Armour, 2 rag) and glucose (50 mg), and then incubated twice for 15 min at a constant rotation of 200 rpm in a 37°C incubator (New Brunswick Sci., G24 Incubator Shaker). After papain digestion, the tissue fragments were resuspended in a culture medium consisting of 5% (v/v) precolostrum newborn calf serum (lot M N J 82, Mitsubishi Kasei), 5% (v/v) heat-inactivated horse serum (lot 32N3176, Gibco) and 90% of a I:1 mixture of

P02 50%02 300

_ e.o

e--i

200 -l.,,,...O

ol

Z

POz 20%02 :~'0

E E

100

II

0 8.0

-

75

°" --

0 I 0

Pcoz

I

2

,io --

pH

_

I H ~ "1 I

4

Time

24

( hr

96

)

Fig. 1. Time-dependent changes in the tension values of 0 2 and CO 2 in the culture medium, and pH values at different oxygen concentrations. After the transfer of culture plates from a 20% to 50% oxygen atmosphere, the 0 2 (e), CO 2 (11) and pH ( A ) values were measured. The corresponding control values ( 0 , [] and A) in a culture of 20% oxygen are also were shown. Values were measured with a blood gas analyzer.

Dulbecco's modified Eagle's and H a m ' s F12 m e d i u m ( D F m e d i u m , both Gibco) containing 15 m M HEPES buffer, pH 7.4. 30 nM selenium, 1.9 m g / m l sodium bicarbonate, 50 u n i t s / m l penicilin G and 0.1 m g / m l streptomycin sulfate. Viable cells that excluded nigrosin dye were then counted in a hemocytometer. T h e viability of the resulting dissociated cells was about 95%. The dissociated cells were directly plated at a density of 6 - 8 x 105 c e l l s / c m 2 on a polyethyleneimine-coated surface in 48-well plates (0.66 cm 2 of culture surface area, Sumilon 48F multi-well plate). After 1 day of culture in a humidified CO 2 incubator with 5% CO 2 and 95% air, the medium was changed to a serum-free T I P / D F medium that contained 5 p , g / m l of h u m a n transfferin (Sigma), 5 , ~ g / m l of bovine insulin (Colab. Res.) and 20 nM progesterone in the D F medium. Cytosine arabinoside (Ara C, Sigma) was added to a final concentration of 5 /xM. Cells were transferred and continued to be cultivated in the chambers with different amounts of oxygen. That is. cultures were carried out in 9, 20 and 50% {v/v) 0 2 and 86, 75 and 45% (v/v) N2, respectively, and a constant 5% ( v / v ) CO 2 atmosphere in a N2-O 2CO 2 incubator (TABAI, BNP-I10M). The concentrations of 0 2 and CO 2 were monitored by corresponding sensors in the incubator. Fig. 1 presents time-dependent changes in the tension values of O 2 and CO2, and pH values in the medium, after the culture plates were transferred into incubators with atmospheres of 20% and 50% oxygen. The oxygen tension in the 50% 0 2 culture increased rapidly and reached a plateau level of ~ 300 m m H g after 2 - 3 h. In contrast. there was no change in the 20% 0 2 atmosphere. T h e values of CO 2 and pH were stable in both the 20% and 50% 0 2 cultures. Each value was measured by a blood gas analyzer (Shimadzu. BGAI01).

Immunohistochemistry The cultured neurons were stained with a polyclonal anti-MAP2 antibody 37. Cells were fixed in 4% paraformaldehyde at room temperature for 20 min. After incubation with PBS containing 5% ( v / v ) normal goat serum and 0.3% (v/v) Triton X-100, for 1 h at room temperature, cells were incubated with anti-MAP2 antibody (1 : 10,000) overnight at 4°C. Cells were then incubated with biotiny-

263 A B C kit as described above, except for cold 100% methanol fixation, and use of biotinylated rabbit anti-mouse antibody as the second antibody.

RESULTS

Oxygen-induced neuronal cell death in serum-free medium We have previously shown that the cell death of PC12h pheochromocytoma cells and rat CNS neurons were observed when they cultured in high oxygen atmosphere 13. In the present study, we cultured embryonic day 20 (just prior to birth) rat CNS neurons in a serum-free T I P / D F medium with Ara C at a final concentration of 5/~M, in order to suppress the proliferation of non-neuronal cells. Fig. 2 shows the survival of MAP2-positive neurons from E20 rat basal forebrain cultured in 9, 20 (the same 0 2 concentration as in air) and 50% oxygen after 3 days in culture. As the concentration of oxygen increased, the survival of neurons was suppressed. Viable neurons in the 9% oxygen atmosphere were about 1.5-fold more numerous than in 20% oxygen and less than 5% of neurons were able to survive in 50% oxygen.

I

I

I

I

100 20~02

•.;

=8o m

m

50

i m e i

Fig. 2. Immunohistochemical staining by anti-MAP2 antibody of cultured basal forebrain neurons u n d e r different oxygen concentrations. Cells were prepared from E20 rat basal forebrain and plated on polyethyleneimine-coated 48-well plates at a density of 6 × 105 c e l l s / c m 2. After 3 days in culture in T I P / D F m e d i u m containing 5 /zM A r a C at 9% (A), 20% (B) and 50% (C) O 2 concentrations, cells were fixed by 4% paraformaldehyde and immunostained by antiM A P 2 antibody as described in the text. Bar = 100/~m. lated goat anti-rabbit antibody (1 : 200) for 45 min at room temperature. Visualization was carried out with a Vectastain A B C kit (Vector) under exposure to 0.02% ( w / v ) 3,3'-diamino-benzidine 4-HC1 and 0.1% ( w / v ) (NH4)2Ni(SO4) 2 dissolved in 0.05 M Tris HC1 buffer, pH 7.6, containing 0.01% (v/v) H 2 0 2. The n u m b e r of immunoreactive neurons was determined u n d e r microscopic observation and also by examining their microphotographs. The MAP2stained neurons in 5 randomly selected microscopic fields per well were counted and averaged. O n e field was about 1.3% of the surface area of the well. Astroglial cells were stained with a monoclonal anti-glial fibrillary acidic protein (GFAP) antibody ( A m e r s h a m ) using t h e Vectastain

40

02

A,A, none 0 , 0 +bFC

0

I

0

I

I

T~

1 2 3 days in culture

1

Fig. 3. Time course of b F G F - e n h a n c e d survival of cultured basal forebrain neurons under different oxygen concentrations. Cells were prepared from E20 rat basal forebrain and cultured in T I P / D F medium containing 5 #,M AraC at 20% ( r , , ©) and 50% ( A , e) 0 2. b F G F was added at a concentration of 100 n g / m l (e, ©). Results for the cells in medium without b F G F are also shown ( A, zx). After 3.5 days of culture, cells were immunostained with anti-MAP2 antibody, as described in the text. The survival rate of the cultured neurons was expressed as a percentage of the initial cell number. Values are means+_ S.D. (n = 4).

264

Fig. 4. Phase-contrast and bright-field micrographs of basal forebrain neurons cultured with and without b F G F under different oxygen concentrations. Cells were cultured for 3.5 days at 20% (A,B) and 50% ( C - F ) 0 2 in T I P / D F medium containing 5 / z M AraC. as described in the legend of Fig. 3. bFGF was added at a concentration of 100 n g / m l (13,D,F). Control was without b F G F (A,C,E). A - D are phase-contrast micrographs. E and F are bright-field micrographs of the same field of C and D, respectively. E and F show the neurons immunostained by anti-MAP2 antibody. Bar = I00 p.m.

265 T i m e - d e p e n d e n t cell death in the 50% oxygen atmosphere is shown in Fig. 3. Cell death was observed even after 1 day of culture, and only 2 - 5 % of the initial number of cells survived after 3 days in culture. The rate of cell death observed in the present serum-free culture was higher than that in the previously reported serum-containing culture. The same result was observed in the culture of PC12h cells (data not shown). It has been known that serum contains some unknown factors which support the cell survival 19. For these reasons, we used in the present study this serum-free culture condition to investigate the cell-saving effects of various neurotrophic factors.

! 2

bFGF prevents cell death caused by oxygen toxicity in serum-free medium We have already reported that N G F and b F G F could save the cell death of PC12h cells in the highly oxidized culture condition 13. In the present study, we tested whether the cell-saving effect of b F G F to the cultured CNS neurons is also observed on the neuronal cell death caused by high oxygen atmosphere. The cell death of cultured E20 rat basal forebrain neurons observed in 50% 0 2 was markedly alleviated by the addition of b F G F to the culture medium at a final concentration of 100 n g / m l . From 55 to 65% of the initial number of dissociated cells survived for 3.5 days in the presence of b F G F as shown in Fig. 3. No morphological change of survival neurons occurred in b F G F - p r e s e n c e culture, while in the culture without b F G F the massive damage to somata and neurites were observed (Fig. 4A). The cell-saving effect of b F G F was not observed clearly in the case of a 20% O 2 culture (Fig. 3). No evident morphological differences between the cultured neurons without and with b F G F were observed (Fig. 4A,B). It has been known that b F G F promotes not only the survival of neuronal cells but also the proliferation and differentiation of non-neuronal cells, for example astroglial cells. To confirm whether the cell-saving effects of b F G F on the cultured neurons were direct or indirect, a n t i - G F A P antibody staining of astroglial cells in culture was done (Fig. 5). U n d e r the same culture condition, the number of viable GFAP-positive astroglial cells were only a few percent. No significant difference between the cultures without and with b F G F was found. That is, viable GFAP-positive astroglial cells after 3 days in 20% O 2 culture without and with b F G F at the concentration of 100 n g / m l were 1.9% _+ 0.2 and 1.3% _+ 0.3 (means + S.D., n = 4) of the corresponding MAP2-positive neurons, respectively. Fig. 5 shows the morphological changes in GFAP-positive cells in culture without and with bFGF. Although

Fig. 5. Immunocytochemicalstaining of astroglial cells by anti-GFAP antibody in culture with and without bFGF. Cells were cultured in TIP/DF medium containing 5 ~M AraC for 3.5 days in 20% oxygen atmosphere as described in the legend to Fig. 3. Cultures were done without (A) and with (B) bFGF at a concentration of 100 ng/ml. Cells were immunostained by anti-GFAP antibody as described in the text. Note that the addition of bFGF increased GFAP-positive astroglial fibers. Bar = 100/zm.

b F G F did not affect the number of viable astroglial cells, b F G F affected the increase in glial fibers (Fig. 5B). The same p h e n o m e n a were observed in 50% 0 2 culture (data not shown). The cell-saving effect of b F G F in 50% 02 was dose-dependent on the concentration of b F G F (Fig. 6). The ED50 value was 12 n g / m l and the maximal effect was obtained at 100 n g / m l of bFGF. In 20% oxygen,

266

I 100

.;

I

I

I

I

-

Cell-saving effects of other neurotrophic factors on oxygen-damaged neuronal cells

-

_

•

m

the d o s e - d e p e n d e n t effect of b F G F was not clcarly observed.

I

I I

I

5o%0oT_/

.

-I I

bFGF(ng/ml)

Fig. 6. Dose-dependent survival curves of basal forebrain neurons in 2 days culture with bFGF under different oxygen concentrations. Cells were cultured at 20% (©) and 50% (e) 0 2 in TIP/DF medium containing 5 ~M AraC, as described in the legend of Fig. 3. Values are means _+S.D. (n = 4).

N G F a n d I G F - I I at c o n c e n t r a t i o n s o f 100 n g / m l a n d 500 n g / m l , respectively, could not p r e v e n t t h e cell d e a t h in 5 0 % 0 2 culture, as shown in Fig. 7. Previous studies have shown that N G F a n d b F G F c o u l d p r e v e n t t h e d e a t h of PC12h cells. In the p r e s e n t study, differe n t from t h e results o f PC12h cells, only b F G F could effectively p r o m o t e t h e survival of t h e n e u r o n a l cells in a 50% oxygen a t m o s p h e r e . In CNS, it has b e e n k n o w n t h a t N G F i n d u c e s the d i f f e r e n t i a t i o n a n d p r o m o t e s t h e suvival o f t h e basal f o r e b r a i n c h o l i n e r g i c n e u r o n s 2°'48. T o investigate the effect o f N G F for t h e survival o f c h o l i n e r g i c n e u r o n s in this area, we m e a s u r e d the C h A T activity in the culture at 50% O~, as c o m p a r e d to the C h A T activity in 20% oxygen. A s shown in Fig. 8, t h e C h A T activity was c o m p a r a b l e to t h a t with 20% oxygen, only w h e n b F G F was p r e s e n t , b u t N G F h a d little effect.

Fig. 7. Immunocytoehemical staining of cultured basal forebrain neurons with different growth factors at 50% oxygen concentration. Cells were prepared and cultured in TIP/DF medium containing 5#M AraC for 3.5 days. The following factors were added: control (A). 1O0 ng/ml of NGF (B), 100 ng/ml of bFGF (C) and 500 ng/ml of IGF II (D). Cultured neurons were immunostained by anti-MAP2 antibody as described in the legend to Fig. 3. Bar = 100 txm.

267

ChAT activity ( pmol/min/well ) 0

5

10

=

none

6

h

0

50 % 02

5 I

nor're

10

15

20

i

20%02

NGF FGF NGF+r-.GF Fig. 8. Effects of NGF and bFGF on ChAT activity of basal forebrain neurons cultured under different oxygen concentrations. Cells were cultured at 20% (lower figure) and 50% (upper figure) 0 2 in TIP/DF medium containing 5/xM AraC, as described in the legend to Fig. 3. After 3 days in culture without or with NGF or bFGF at the concentration of 100 ng/ml, ChAT activity was determined as described in the text. Values are means + S.D. (n = 4).

Cell-saving effect of bFGF on neurons from uarious CNS areas b F G F have been known to promote the survival of variety parts of CNS neurons in vitro 52. In the present study, we investigated the effect of b F G F to prevent the oxygen-induced cell death on various CNS neurons, as shown in Fig. 9. When the neurons from neocortical and hippocampal regions were cultured in 50% 0 2, the cell death occurred as same as that in the basal forebrain culture (Fig. 9C,E). Interestingly, the tolerance for oxygen toxity was markedly different between the 3 cultures obtained from basal forebrain, neocortex and hippocampus. As shown in Fig. 10, after 1.5 days of culture in 50% 0 2, less than 5% of the initial number of neocortical neurons were viable. Then, all of the neurons died after 3.5 days of culture. In contrast, the hippocampal neurons in the same kind of culture could survive still at the extent from 10 to 20% of the initial number of neurons, after 3.5 days of culture. These results m e a n that the neocortical neurons were the most sensitive to oxidative stress, while the hippocampal neurons were the most resistant among the 3 types in culture. The effect of b F G F in promoting the survival of neocortical and hippocampal neurons was also observed (Figs. 9 and 10). However, neocortical neurons were most damageable even in the presence of bFGF. DISCUSSION In this study, we investigated a clue to know the physiological functions of b F G F , especially the effects

to support the survival and recovery of injured neuronal cells. In conclusion, the neuronal cell death induced by a high oxygen atmosphere in culture was markedly prevented by the addition of bFGF. These results suggest that b F G F plays a role in preventing the neuronal injury induced by oxidative damage. Recently, the physiological roles of b F G F on CNS neurons have been examined in vivo and in vitro 49'5°'52'53, and found b F G F promoting the differentiation and survival of various neuronal populations. However, by reason of the low specificity of its effects on neuronal cells, including its broad distribution in the brain, new concepts for its role as neurotrophic factor have been proposed, one that is different from that of so-called target-derived neurotrophic factors, like N G F and its family proteins. Especially, since its ability to bind to heparan sulfate proteoglycans in the extracellular matrix and its lack of signal sequences, it has been proposed that b F G F is released from damaged cells and restores injured cells. In the present study, we used cultured rat CNS neurons and supplied excess oxygen to cause CNS neuronal cell death. Furthermore, we examined the effect of neurotrophic factors on the survival of injured neurons. This was done to limit the conditions eliciting cell death and to clarify the role of neurotrophic support. In previous studies, we demonstrated that during development neurons acquire tolerance for the oxidative stress induced by a high oxygen atmosphere 3°. This suggests that not only aerobic energy metabolism but also a mechanism for protecting the oxygen stress are obtained in this stage. In addition, we have shown that various neurotrophic factors could prevent the death of PC12h cells ]3. This result suggests that neurotrophic factors are closely involved in the mechanisms that prevent oxidative injury. A similar result was reported, namely that N G F and retinoic acid could rescue the cell death induced by hydrogen peroxide 26. In this study, we used cultured CNS neurons and showed the effectiveness of b F G F to prevent the oxygen-induced neuronal cell death. We consider this effect to be the result of the direct action of b F G F on neuronal ceils. As we used serumfree T I P / D F medium containing 5 tzM A r a C for culture medium, the surviving cells were mainly non-dividing neurons. Thus, it is uhlikely that b F G F gave rise to contaminated proliferating ceils, for example, astroglial cells. In fact, the number of GFAP-positive astroglial cells was very low in this culture, and there was no difference between cultures with and without bFGF. Furthermore, the cell-saving effect of b F G F was observed early in culture, suggesting that b F G F directly prevents oxidative damage to neurons. As shown in

268

Fig. 9. Immunocytochemical staining of neurons from different CNS areas cultured at a 50% oxygen atmosphere. Cells were prepared from E20 rat basal forebrain (A,B), hippocampus (C,D) and neocortex (E,F) as described in the text. Cell culture was carried out in T I P / D F medium containing 5 izM AraC for 1.5 days (neocortex) and 3.5 days (basal forebrain and hippocampus) in 50f~: 0 2. Cultures were carricd out without (left column) and with (right column) 100 n g / m l of bFGF. Plating cell densities and culture conditions were the same. Cultured neurons were immunostained by anti-MAP2 antibody as described in the legend to Fig. 2. Bar = 100 tzm

269 Neuron Survival (% of Initial Number) 0 I

20

40

i

i

60

80

I

I

1.5

]

-bFGF

]

+bFGF

A 3.5 ¢-

~" 3.5 121

3.5

I

C

Fig. 10. Effect of b F G F on the survival of neurons from different CNS areas cultured at a 50% oxygen atmosphere. Cells from neocortex (A), basal forebrain (B) and hippocampus (C) were prepared and cultured in T I P / D F medium conteining 5/~M Ara C for 1.5 (neocortex) or 3.5 (all 3 areas) days in a 50% oxygen atmosphere. Surviving neurons were immunostained with anti-MAP2 antibody and counted as described in Fig. 3. b F G F was added at a concentration of 100 n g / m l b F G F (open columns). Control was without b F G F (closed columns). Values are means + S.D. (n = 4).

Fig. 6, the increase in astroglial fibers caused by b F G F may suggest that changes in astroglial cells occurred with the addition of b F G F in culture. Therefore, we can not completely exclude the possibility that b F G F had some indirect effect on neuronal survival. Recently, experiments that examined the distribution of b F G F and its receptor using immuno-histochemical staining and in situ hybridization 4'16'24'4°'54 demonstrated that b F G F and its receptor are present in neuronal populations. These results suggest a possibility that b F G F protects neurons against oxidative stress in a physiological condition. Interestingly, it has been reported that b F G F partially prevents the deleterious chemical and morphological consequences of 1-methyl-4-phenyl-l,2,3,6-tetrahydropyridine (MPTP)-mediated nigrostriatal lesions 38 and MPTP-induced neuronal cell death closely related to oxidative events in mitochondria 35'5~. It can also be speculated that when the brain is damaged by surgical injury, macrophages, which are activated by the inflammatory response, release oxygen radicals 39. Therefore, b F G F may be released to protect nondamaged neurons in the area. In fact, it is known that activated macrophages could release FGFs 2. In addition, in cerebrovascular injury, it is hypothesized that excess oxygen is supplied in the ischemia-reperfusion and this causes neuronal injury. Recently, it has been reported that some neurotrophic factors protect the cell death after ischemia 46"57.

As shown in Fig. 4, the dose of b F G F needed to prevent neuronal cell death was higher than that needed to promote the survival of cultured neurons reported previously 15'36'5°'52. The reason is considered to be as follows: first, as the oxygen concentration in the medium was very high, it is thought that the added b F G F might be oxidized and inactivated; second, as the cell density was high, the non-specific binding of b F G F made its concentration in the medium much lower than the added concentration; third, the survival of basal forebrain neurons may require a high dose of bFGF. In contrast to the results in PC12h cells, N G F could not effectively promote the survival of neurons including the cholinergic neurons. A possibility is that the cholinergic neurons that respond to N G F are few in this area 23, and even if N G F promoted their survival, the main populations of non-responsive neurons would die. As a result, the cell density would be low and this might cause the death of cholinergic neurons. In the present study, the survival-promoting effects of b F G F were observed not only on basal forebrain neurons but also on neocortical and hippocampal neurons. In addition, a difference in sensitivity to oxygen was observed. These results, as mentioned above, may reflect the survival-promoting effect of b F G F on various parts of CNS neurons damaged by a high oxygen toxicity atmosphere. Interestingly, hippocampal neurons, which are among the most easily damaged by ischemia and glutamate neurotoxicity, were most resistant to the high oxygen concentration. Since b F G F 16'4°, and other neurotrophic factors such as a F G F 56, N G F ~, brain-derived n e u r o t r o p h i c factor a5'41 and neurotrophin-3 3z'41'55 are synthesized in this area, it is interesting to study whether these factors may protect against cell death induced by a high oxygen atmosphere. Furthermore, it is necessary to consider the effect of neuronal network between hippocampus and other areas in vivo. In contrast, neocortical neurons were barely able to survive in a serum-free medium, so it is thought that the addition of other neurotrophic factors is necessary for cell survival. In summary, the present studies have shown that b F G F promotes the survival of CNS neurons injured by oxygen-induced cell death. This may help in understanding the mechanism of neuronal cell survival induced by b F G F and that of cell death induced by oxidative events. Recently, conditions that induce neuronal cell death have been reported in vivo and in vitro, for example, neurotrophic factor deprivation a9"33, serum deprivation 3'19'43, glutamate neurotoxicity 8,34,42 and glucose deprivation 9,1°. To compare the oxygen-induced neuronal cell death with these phenomena may

270 help solving not only the mechanism of neuronal cell death but also that of neuronal cell survival. Especially, it has been reported that growth factors can stabilize neuronal calcium homeostasis in CNS neurons and thereby protect them against neuronal injuries. Thus, it will be required to study whether the disorder of neuronal calcium homeostasis correlates oxygen-induced neuronal cell death. In our preliminary study, we observed that cycloheximide and actinomycin D greatly reduced the celt death induced by a high oxygen concentration. This may be a key point in answering whether the de novo protein syntesis and RNA synthesis are involved in the mechanisms of oxygen-induced neuronal cell death. Acknowledgements. This work was supported in part by a Grand-inAid for Scientific Research on Priority Areas, Ministry of Education, Science and Culture, in part by a Research Grant for Neurons and Mental Disorders from the Ministry of Health and Welfare, and in part by a grant from Nissan Science Foundation, Japan. We thank Takeda and Daiichi Pharmaceutical Co. for their kind gifts of human recombinant bFGF and IGF II, respectively.

REFERENCES 1 Ayer-LeLievre, C., Olson, L., Ebendal, T., Seiger, A. and Pers-

son, H., Expression of the beta-nerve growth factor gene in hippocampal neurons, Science, 240 (1988) 1339-1341. 2 Baird, A., Morm~de, P. and B6hlen, P., Immunoreactive fibroblast growth factor in cells of peritoneal exudate suggests its identity with macrophage-derived growth factor, Biochem. Biophys. Res. Commun., 126 (1985) 358-364. 3 Batistatou, A. and Greene, L.A., Aurintricarboxylic acid rescues PC12 cells and sympathetic neurons from cell death caused by nerve growth factor deprivation: correlation with suppression of endonuclease activity, J. Cell Biol., 115 (1991)461-471. 4 Bean, A.J., Elde, R., Cao, Y., Oellig, C., Tamminga, C., Goldstein, M., Pettersson, R.F. and H6kfelt, T., Expression of acidic and basic fibroblast growth factors in the substantia nigra of rat, monkey, and human, Proc. Natl. Acad. Sci. USA, 88 (1991) 10237-10241. 5 Bocchini, V. and Angeleni, P.U., The nerve growth factor: purification as a 30,000-molecular weight protein, Proc. Natl. Acad. Sci. USA, 64 (1969) 787-794. 6 Burgess, W.H. and Maciag, T., The heparin-binding (fibroblast) growth factor family of proteins, Annu. Rev. Biochem., 58 (1989) 575-6O6. 7 Cao, W., Carney, J.M., Duchon, A., Floyd, R.A. and Chevion, M., Oxygen free radical involvement in ischemia and reperfusion injury to brain, Neurosci. Lett., 88 (1988) 233-238. 8 Choi, D.W., Glutamate neurotoxicity and diseases of the nervous system, Neuron, 1 (1988)623-634. 9 Cheng, B. and Mattson, M.P., NGF and bFGF protect rat hippocampal and human cortical neurons against hypoglycemic damage by stabilizing calcium homeostasis, Neuron, 7 (1991) 10311041. 10 Cheng, B. and Mattson, M.P., IGF-I and IGF-II protect cultured hippocampal and septal neurons against calcium-mediated hypoglycemic damage, 3. Neurosci., 12 (1992) 1558-1566. 11 Eccleston, P.A. and Silberberg, D.H., Fibroblast growth factor is a mitogen for oligodendtocytes in vitro, Brain Res., 353 (1985) 315-318. 12 Elde, R., Cao, Y., Cintra, A., Brelje, T.C., Pelto-Huikko, M., Junttila, T., Fuxe, K., Pettersson, R.F. and H6kfelt, T., Prominent expression of acidic fibroblast growth factor in motor and sensory neurons, Neuron, 7 (1991) 349-364.

13 Enokido, Y. and ttatanaka, If.. lligh oxyg,:n almosphe~c i,, neuronal celt culture with nerve growth faclor 11. Survival ~lnd growth of clonal rat pheochromocytoma PC12h cells, Brain Re~. 536 {1990) 23-29. 14 Fcrrara, N., Ousley, F. and Gospodarowicz, D., Bovine hi-am astrocytes express basic fibroblast growth factor, a neurotrophic and angiogenic mitogen, Brain Res., 462 (1988) 223:232. 15 Ferrari, G., Minozzi, M.-C., Toffano, G.. ],con, A. and Skapcr, S.D., Basic fibroblast growth factor promotes the survival and development of mesencephalic neurons in culture, Dev. Biol., 133 (1989) 14(l- 147. 16 Finklestein, S.P., Apostolides, PJ., Caday, C.G_ Prosser, J.. Philips, M.F. and Klagsbrun, M., Increased basic fibroblast growth factor (bFGF) immunoreactivity at the site of tocal brain wounds, Brain Res., 460 (1988) 253-259. 17 Fonnum, F., A rapid radiochemical method for the determination of choline acetyltransferase, Z Neurochem.. 24 (1975) 407-409. 18 Goldfarb, M., The fibroblast growth factor family, Cell Growth DrillS, 1 (1990)439-445. 19 Greene, L.A., Nerve growth factor prevents the death and stimulates neuronal differentiation of clonal PCI2 pheochromocytoma cells in serum-free medium, J. Cell Biol., 78 (1978) 747-755. 20 Hatanaka, H., Tsukui, H. and Nihonmatsu. l., Developmental change in the nerve growth factor action from induction of choline acetyltransferase to promotion of cell survival in cultured basal forebrain cholinergic neurons from postnatal rats, Det,. Brain Res., 39 (1988) 85-95. 21 Harman, D., The aging process, Proc. Natl. Acad. Sci. USA, 78 (1981) 7124-7128. 22 Hatten, M.E., Lynch, M., Rydel, R.E., Sanchez, J., Joseph-Silverstein, J., Moscatelli, D. and Rifkin, D.B., In vitro neurite extension by granule neurons is dependent upon astroglial-derived fibroblast growth factor, Dev. Biol., 125 (1988) 280 289. 23 Hefti, F., Hartikka, J., Eckenstein, F., Gnahn, H., Heumann, R. and Schwab, M., Nerve growth factor increases choline acetyltransferase but not survival or fiber outgrowth of cultured fetal septal cholinergic neurons, Neuroscience, 14 (1985) 55-68. 24 Heuer, J.G., yon Bartheld, C.S., Kinoshita, Y,, Evers, P.C. and Bothwell, M., Alternating phases of FGF receptor and NGF receptor expression in the developing chicken nervous system, Neuron, 5 (1990) 283-296. 25 Hofer, M., Pagliusi, S.R., Hohn, A., Leibrock, J. and Barde, Y.-A., Regional distribution of brain-derived neurotrophic factor mRNA in the adalt mouse brain, EMBO J., 3 (1990) 3183-3189. 26 Jackson, G.R., Morgan, B.C., Werrbach-Perez, K. and Perez-Polo, J.R., Antioxidant effect of retinoic acid on PC12 rat pheochromocytoma, Int. Z Det'. Neurosci., 9 (1991) 161-170. 27 Kawai, S., Yonetani, M., Nakamura, H. and Okada, Y., Effects of deprivation of oxygen and glucose on the neural activity and the level of high energy phosphates in the hippocampal slices of immature and adult rats, Det,. Brain Res.. 48 (1989) 11-18. 28 Klagsbrun, M., The fibroblast growth factor family; structural and biological properties, Prog. Growth Factor Res., 1 (1989)207-235. 29 Koike, T., Martin, D.P. and Johnson Jr., E.M., Rote of Ca :+ channels in the ability of membrane depolarization to prevent neuronal death induced by trophic-factor deprivation: evidence that levels of internal Ca 2+ determine nerve growth factor dependence of sympathetic ganglion cells, Proc. Natl. A e a d Sci. USA, 86 (1989) 6421-6425. 30 Kushima, Y., Tsukui, H., Enokido, Y., Nishio, C. and Hatanaka, H., High oxygen atmosphere for neuronal cell culture with nerve growth factor. I. Primary culture of basal forebrain cholinergic neurons from fetal and postnatal rats, Brain Res., 536 (1990) 16-22. 31 Lowry, O.H., Rosebrough, N,J., Farr, A.L. and Randell, R,J., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 265-275. 32 Maisonpierre, P.C., Belluseio, L., Friedman, B., Atderson, R.F., Wiegand, S.J., Furth, M.E., Linsay, R.M. and Yancopoul0s, G.D., NT-3, BDNF, and NGF in the developping rat nervous system: parallel as well as reciprocal patterns of expression, Neuron, 5 (1990) 5{/1-500.

271 33 Martin, D.P., Schmidt, R.E., DiStefano, P.S., Lowry, O.H., Carter, J.G. and Johnson Jr., E.M., Inhibitors of protein synthesis and RNA synthesis prevent neuronal death caused by nerve growth factor deprivation, J. Cell Biol., 106 (1988) 829-844. 34 Mattson, M.P., Murrain, M., Guthrie, P.B. and Kater, S.B., Fibroblast growth factor and glutamate: opposing roles in the generation of hippocampal neuroarchitecture, J. Neurosci., 9 (1989) 3728-3740. 35 Mizuno, Y., Sone, N. and Saitoh, T., Effects of 1-methyl-4phenyl-l,2,3,6-tetrahydropyridinium ion on activities of the enzymes in the electron transport system in mouse brain, J. Neurochem. 48 (1987) 1787-1793. 36 Morrison, R.S., Sharma, A., DeVellis, J. and Bradshow, R.A., Basic fibroblast growth factor supports the survival of cerebral cortical neurons in primary culture, Proc. Natl. Acad. Sci. USA, 83 (1986) 7537-7541. 37 Niinobe, M., Maeda, N., Ino, H. and Mikoshiba, K., Characterization of microtubule-associated protein 2 from mouse brain and its localization in the cerebellar cortex, J. Neurochem., 51 (1988) 1132-1139. 38 Otto, D. and Unsicker, K., Basic FGF reverses chemical and morphological deficits in the nigrostriatal system of MPTP-treated mice, J. Neurosci., 10 (1990) 1912-1921. 39 Pabst, M.J. and Johnston Jr., R.B., Increased production of superoxide anion by macrophages exposed in vitro to muramyl dipeptide or lipopolysaccharide, J. Exp. Med., 151 (1980) 101-114. 40 Pettmann, B., Labourdette, G., Weibel, M. and Sensenbrenner, M., The brain fibroblast growth factor (FGF) is localized in neurons, Neurosci. Lett., 68 (1986) 175-180. 41 Phillips, H.S., Hains, J.M., Laramee, G.R., Rosenthal, A. and Winslow, J.W., Widespread expression of BDNF but not NT-3 by target areas of basal forebrain cholinergic neurons, Science, 250 (1990) 290-294. 42 Rothman, S.M., Thurston, J.H. and Hauhart, R.E., Delayed neurotoxity of excitatory amino acids in vitro, Neuroscience, 22 (1987) 471-480. 43 Rydel, R.E. and Greene, L.A., Acidic and basic fibroblast growth factors promote stable neurite outgrowth and neuronal differentiation in cultures of PC12 cells, J. Neurosci., 7 (1987) 3639-3653. 44 Schubert, D., Ling, N. and Baird, A., Multiple influences of a heparin-biding growth factor on neuronal development, J. Cell. Biol., 104 (1987) 635-643. 45 Sherwood, N.M. and Timiras, P.S., A Stereotaxic Atlas of the DevelopingRat Brain, University California Press, Berkeley, 1970. 46 Shigeno, T., Mima, T., Takakura, K., Graham, D., Kato, G., Hashimoto, Y. and Furukawa, S., Amelioration of delayed neu-

ronal death in the hippocampus by nerve growth factor, J. Neurosci., 11 (1991) 2914-2919. 47 Suda, K., Barde, Y.-A. and Thoenen, H., Nerve growth factor in mouse and rat serum: Correlation between bioassay and radioimmunoassay determinations, Proc. Natl. Acad. Sci. USA, 75 (1978) 4042-4046. 48 Thoenen, H., Bandtlow, C. and Heumann, R., The physiological function of nerve growth factor in the central nervous system: comparison with the periphery, Rev. Physiol. Biochem. Pharmacol., 109 (1987) 145-178. 49 Torres-Aleman. I., Naftolin, F. and Robbins, R.J., Trophic effects of basic fibroblast growth factor on fetal rat hypothalamic cells: interactions with insulin-like growth factor I, Dev. Brain Res., 52 (1990) 253-257. 50 Unsicker, K., Reichert-Preibsch, H., Schmidt, R., Pettmann, B., Labourdette, G. and Sensenbrenner, M., Astroglial and fibroblast growth factor have neurotrophic functions for cultured peripheral and central nervous system neurons, Proc. Natl. Acad. Sci. USA, 84 (1987) 5459-5463. 51 Vyas, I., Heikkila, R.E. and Nicklas, W.J., Studies of the neurotoxicity of MPTP: Inhibition of NADH-linked substrate oxidation by its metabolite, MPP ÷, J. Neurochern., 46 (1986) 1501-1507. 52 Walicke, P.A., Basic and acidic fibroblast growth factors have trophic effects of neurons from multiple CNS regions, 3. Neurosci., 8 (1988) 2618-2627. 53 Walicke, P.A. and Baird, A., Neurotrophic effects of basic and acidic fibroblast growth factors are not mediated through glial cells, Dev. Brain Res., 40 (1988) 711-719. 54 Wanaka, A., Johnson Jr., E.M. and Milbrandt, J., Localization of FGF receptor mRNA in the adult rat central nervous system by in situ hibridization, Neuron, 5 (1990) 267-281. 55 Wetmore, C., Ernfors, P., Persson, H. and Olson, L., Localization of brain-derived neurotrophic factor mRNA to neurons in the brain by in situ hybridization, Exp. Neurol., 109 (1990) 141-152. 56 Wilcox, B.J. and Unnerstall, J.R., Expression of acidic fibroblast growth factor mRNA in the developing and adult rat brain, Neuron, 6 (1991) 397-409. 57 Yamada, K., Kinosita, A., Kohmura, E., Sakaguchi, T., Taguchi, J., Kataoka, K. and Hayakawa, T., Basic fibroblast growth factor prevents thalamic degeneration after cortical infarction, J. Cereb. Blood Flow Metab., 11 (1991) 472-478. 58 Yayon, A., Klagsbrun, M., Esko, J.D., Leder, P. and Ornitz, D.M., Cell surface, heparin-like molecules are required for binding of basic fibroblast growth factor to its high affinity receptor, Cell, 64 (1991) 841-848.