- Email: [email protected]
Basic fibroblast growth factor high and low affinity binding sites in developing mouse brain, hippocampus and cerebellum
Biol Cell (1992) 76, 1-13
1
© Elsevier, Paris
Original article
Basic fibroblast growth factor high and low affinity binding sites in developing mouse brain, hippocampus and cerebellum Nicole-Adeline Fayein, Yves Courtois, Jean-Claude Jeanny * Unitd de Recherches G~rontologiques, INSERM Ul18, CNRS UA630, Ddveloppement et S~nescence Cellulaire, Association Claude Bernard, 29, rue Wilhem, 75016 Paris, France (Received 17 July 1992; accepted 19 October 1992)
Summary - Fibroblast growth factors (FGFs), first extracted from brain and retina, are potent neurotrophic factors. They stimulate neuroblast proliferation and neuron differentiation and survival. In order to study the spatial and temporal distribution of the target cells in the mouse brain we studied by autoradiography and quantified by image analysis ~251-bFGFbinding sites as a function of development. We have revealed the presence of two types of specific bFGF receptors. One is heparitinase sensitive and is co-localized with heparan sulfate proteoglycans of the basement membranes (meninges, choroid plexus and blood vessels). It is not developmentally regulated and corresponds to the low affinity receptors. It may be a storage form. The second type is heparitinase resistant and is modified during development, matching, in the adult, layering of the hippocampus and cerebellum. At 13 days of embryonic development there is a preferential distribution of silver grains on the ecto- and neuroectodermal tissues. In the adult, the labeling is localized on the neural process layers. It likely corresponds to the specific binding to cell high affinity receptors. Binding patterns according to the developmental stages of the brain can be correlated with mitotic, migration and differentiation phases of the neuronal cells. basic fibroblast growth factor / high and low affinity receptors / mouse brain / hippocampus / cerebellum
Introduction Basic and acidic fibroblast growth factors (bFGF a n d aFGF) are mitogenic and neurotrophic. They have been isolated from a wide range of tissues, including brain and retina [2]. bFGF is a mitogen for a variety of mesenchymal cell types [15], for neuroectodermal-derived cells, glial cells [4, 8, 24] as well as neuronal precursors [13, 38, 51]. Its main property, however, seems to be neurotrophic for various neurons of multiple CNS regions [35, 56, 58, 59, 61], including retina [29, 52] and photoreceptors [18]. bFGF is acting through specific cell surface receptors which have been characterized in membranes prepared from various tissues of the CNS: brain [27, 34], retina [31] or neural cells in culture [40, 62]. The localisation of these receptors during retinal embryogenesis and development has been demonstrated in mice [I0, 22] and chicken [6] ocular tissues by autoradiographic techniques. By this technique two types of binding sites could be determined, depending on their resistance to heparitinase digestion and sensitivity to heparin or N-glycanase treatment. Thef first two treatments revealed the low affinity sites that were correlated with the presence of heparan sulfate in basement membranes or at the cell surface, the latter with the high affinity receptor sites for which the N-glycan moiety is necessary for its activity [11].
* Correspondence and reprints
The FGF receptor (FGF-R) is a transmembrane protein that contains three extracellular immunoglobulin-like domains with an unusual acidic region, and an intracellular tyrosine kinase domain [28, 43]. There are three different families of related FGF receptors encoded by three different genes. Complementary DNA clones were isolated with oligonucleotidic probes corresponding to determined amino acid sequences of fragments of the purified receptor. Two clones were isolated from mouse (bFGF-R and bek) [25, 48]. Such clones have been used by in situ hybridization to study the localization and the expression of the genes for FGF receptors during chick and mouse embryonic development [17, 64]. The cell surface proteoglycan, syndecan [50], that is likely to interact with bFGF, has also been recently cloned [23]. In spite of these results we do not know yet the distribution in the brain of the different types of proteins coded by these genes nor their activity in situ. In this study we have analysed by autoradiography, after incubation with iodinated bFGF, the binding sites and their specificity in mouse brain and in more detail in hippocampus and cerebellum, during development. This experimental approach does not allow to discriminate between the various classes of high and low affinity FGF receptors. It reflects their global spatio-temporal localization at various times of development and potentially their accessibility towards the various members of FGF family. We have found the distribution of bFGF receptors (high and low affinity) to be correlated with the stage of differentiation in various neuronal layers.
2
NA Fayein et al
Materials and methods
Chemicals and reagents The following reagents, were used: heparin-Sepharose, PD 10 columns (Pharmacia, St Quentin en Yvelines, France), t2SI-Na (Amersham, Les Ulis, France), optimal cutting temperature medium (OCT) and heparitinase (Miles Scientific, Puteaux, France) hematoxylin and eosin (Gurr). Other chemicals were obtained from Sigma Chemical Co (La Verpilliere, France), Prolabo (Paris, France) and Bio-Rad (Paris, France) and were reagent grade.
were washed in PBS and incubated with iodinated bFGF as above.
Analysis system The analysis system [7] consisted of a camera linked to monitors; data input was directly into a computer system. The whole was controlled by a series of Imagenia R (Biocom, Les Ulis, France) software. After adjusting the image of the monitor, image focusing and image treatment, the computer analyzed automatically the labeling intensity of different structures from brain.
Data acquisition Growth factor preparation and iodination Human recombinant basic FGF was a gift from Carlo Erba (Farmitalia). It was labeled with ~25I according to the chloramine-T method [5] as already described [10]. The specific activity of the 125I-bFGF purified on heparin-Sepharose was 1500-2000 Ci/ mmol. When this iodinated material was subjected to electrophoresis on sodium dodecyl'sulfate (SDS)-polyacrylamide gels [26] a single band migrating in the position of human recombinant bFGF was observed by autoradiography. Full biological activity of bFGF as tested by [3H]thymidine incorporation in bovine epithelial lens cells (BEL) [55], was recovered after iodination.
Preparation of mouse brain frozen sections Pregnant mice (OF1) were obtained from IFFA Credo (Domalne des Oncins, France). Animals were killed by inhalation of carbon dioxide from dry ice. PCD (Purkinje cell degeneration) [37] and TBL (Tambaleante) [65] mice'were a gift from Dr JL Gu6net (Institut Pasteur, Paris, France). Brains of 10- to 18-day-old embryos and 2- to 4-week-old postnatal animals were dissected, embedded in Tissue-Tek OCT compound (embedding medium for frozen tissue specimens) on a metal block, and frozen with liquid nitrogen. Longitudinal 10-/~m sections were cut at - 2 0 ° C in a cryostat (SLEE, London or OTF/AS, Bright, DIS, Blanc Mesnil, France), collected on gelatin-coated glass slides, and stored at - 2 0 ° C until use.
Incubation with iodinated bFGF and autoradiographic techniques Non-fixed sections, dried for 5 min at room temperature after OCT extraction in PBS, were incubated for 1 h at ambient temperature in a water-saturated atmosphere with 20/zl of 1251b F G F diluted in PBS (20 ng/ml, specific activity: 1500:-2000 Ci/mmol). Following the incubation, the slides were washed six times, 5 min each, with PBS. They were then fixed with 3% paraformaldehyde in PBS for 30 min, washed, dehydrated and dried before dipping. They were dipped in photoemulsion LM1 (Amersham). The slides were exposed for 2 - 4 weeks at 4°C, developed in Kodak DI9, and fixed in Hypare Ilford fix. The photographs were taken under the same conditions, using PAN-F film (Ilford) on a Zeiss photomicroscope (Oberkochen, Germany). All experiments were run in triplicate, and each treatment to test the specificity and nature of the binding was performed on a minimum of three slides.
Competition experiment Unlabeled purified bFGF (5/~g/ml) was mixed with its respective iodinated form (20 ng/ml) and deposited on the sections. After 1 h of incubation the slides were treated as above.
Heparitinase treatment of the sections Unfixed sections were treated with various concentrations of heparitinase (25, 50, 100 mU/ml), in the presence of Ca 2÷ in a final volume of 20 tzl, for I h at 37°C. After treatment slides
Images were acquired on film negatives by a charge couple device (CCD) camera, 512 x 512 pixels and 100 grey levels (expressed from 0 to 255). The negatives from each figure were placed on a light box and held flat by a weighted glass slide. The analogue video images were digitized in the computer where the grey levels were calculated. Ten measurements were taken from each hippocampal structure (Stratum oriens, St pyramidale, St radiatum, St lacunosum moleculare, St moleculare, St granulosum, Hilus fasciae dentatae) and also from the background (out of the sections). Data were presented as the mean + standard deviation. Student's t-test was carried out between different hippocampal structures and at two stages of post-natal development. The comparisons were done on sections of the same experiment developed after the same exposure time.
Histological staining After OCT extraction in PBS the sections were stained with hematoxylin and eosin. Photographs were taken as previously described.
Results
Binding o f iodinated bFGF on brain sections during embryonic development At 1 3 - 1 4 days o f embryonic development (ED), the telencephalon and diencephalon are growing a r o u n d the lateral and third ventricles and start to differentiate. There is a differential distribution o f 125I-bFGF binding sites on horizontal head sections. The basement membranes are the highest labeled structures with 125I-bFGF (fig la) for instance below the ectoderm, nasal cavities, tongue, blood vessels and pial surfaces. Strong labeling remains in the developing telencephalon and diencephalon, but with a heterogeneous distribution. To confirm the specificity o f 125I-bFGF binding the sections were treated with 500-fold excess o f unlabeled bFGF. At this concentration (fig lb) there is a m a j o r chase o f the 125I-bFGF as we already described in previous studies in retina [10, 22]. A l t h o u g h the general labeling was decreased, the pattern remained similar. Within the telencephalon, there are two distinct, densely, labeled layers, one periventricular, next to the ventricles, the second in an intermediate zone under the cortical surface. These two layers coincide with the highest cell density (fig l c). At 1 6 - 1 7 days there is no change in the labeling o f the basement membranes beside a strong labeling on the newly-formed choroid plexus. The diencephalon is unif o r m l y labeled f r o m the ventricle to the external surface (fig 2a). The telencephalon displays a distinct pattern, different f r o m the one observed 2 days earlier (fig 2a, b). The labeling is n o w localized in the ventricular and subventricular layers, rather than in the intermediate zone as
bFGF binding in mouse brain
•
.
.
.
.
.
---~..~
.
_
=
.
-
..,-?.~,~,-..,
..-
~t.,.~.
'
~
~
.
~,~.a~,'
,
:~ .....
,.
...
-
.''~'i~
.
.
;:
.,~
~ '~,-
...... •
3
.
',_~r
.¢
,.,
Fig 1. Brain sections from 13-day-old mouse embryo incubated with iodinated bFGF (a) in the presence of unlabeled bFGF (b) or stained with hematoxylin-eosin (c) (bar = 300 tzm). The basement membranes are the strongest labeled structures, for instance below the ectoderm, nasal cavities (nc), tongue (t), blood vessels (bv) and pial surface (ps). The labeling is heterogeneously distributed on the telencephalon (Te) in two zones (periventricular: p; intermediate: i). Competition with native growth factor decreases the initial binding on all labeled structures, v: lateral ventricles.
4
NA Fayein et al
Fig 2. Brain sections from 16-day-old mouse embryo incubated with iodinated bFGF (a, b, d) or stained with hematoxylin-eosin (c, e) (bar = 300/~m). There is strong labeling on the newly-formed choroid plexus (ch pl) and mast cells (mc). The diencephalon (Di) is uniformly labeled from ventricle (v) to meninges (m) whereas the telencephalon (Te) displays a stronger sub- and periventricular labeling. Iv: lateral ventricle.
observed at 13-14 days. To match this pattern with the cellular pattern of this tissue, a neighbouring section has been stained with hematoxylin-eosin (fig 2c). It is apparent that the intense labeling is only colocalized with the periventricular cell layer, and not with the cell layer in the external zone. The dense patches of silver grains under the head ecto-
derm correspond to the labeling on mast cells as already identified in mouse, after metachromatic staining with toluidine blue. At the same stage in the periventricular zone of the telencephalon the labeling is observed more intensely on the external side of the lateral ventricle (fig 2d), progressing to the internal side corresponding to the presumptive zone
bFGF binding in mouse brain
a plexu$ chor.
',.
,,
b .
..,L~,'~.. - ~
•
~
7"4.,:;~
.'-,~
~
•
,"'~'--t":?
~-
Fig 3. Brain sections through the hippocampus from l-month-old mouse incubated with iodinated bFGF (a) or stained with hematoxylineosin (c) (bar = 300 ~m). h. Schematic representation of the hippocampus organization. The labeling is distributed on the hilus (hil) of the area dentata, the stratum radiatum (str rad hip) and with a decreasing intensity on the stratum lacunosum and moleculare (str lac mol hip). There is also a weaker labeling on the stratum oriens (str or hip). Layers of cell bodies are weakly labeled or not labeled: stratum pyramidale (str pyr) and stratum granulare (str gran).
6
NA Fayein et al
of the hippocampus. The staining of a neighbouring section indicates again that the labeling corresponds to the cell population at the border of the ventricle (fig 2e). The other external zones are not associated with a high density of silver grains. Binding o f J2sI bFGF on hippocampus
As the hippocampus has been a favorite structure in the brain for which a physiological role for FGF has been proposed, we focused our study on the developing hippocampus from the 16-17-day embryo to the adult stage. At ED 18, the pattern of labeling matchs precisely the hippocampic morphogenesis. It is localized on the pyramidai cell layer of the subiculum and hippocampus (data not shown). At 1 month post-natal (PN), the hippocampus reaches its final differential stage and is formed by several stratified layers with sharply defined neurons. At the level of the ammonic pyramid.al cell layer, there was no binding while a very dense binding is observed on the stratum radiatum, with a decreasing intensity to the stratum lacunosum and moleculare (fig 3). There was also a weaker labeling on the stratum oriens. At the level of the area dentata the binding is strongly localized in the hilus surrounded by the stratum granulosum which shows an apparent absence of labeling as already observed in the ammonic pyramidal layer (CA1 and CA3). The juxtaposition of the hematoxylin-eosin stained sections with the autoradiographic sections shows clearly that the poorly labeled areas correspond to the nuclear cell layers for both the hippocampus and the dentate gyrus. The high density of labeling is localized on cell processes but with different gradation. The other structures on which the labeling is intense correspond to the blood vessel walls and to the choroid plexus as observed at early stages. To quantify the binding to the different structures the grey levels were analysed by image analysis (table I). The data show a highly significant difference of binding between the pyramidal cell layers and the areas of dendritic ramifications (stratum oriens, radiatum and moleculare). Even within the zone of dendritic ramifications, there is a highly significant differential labeling between the radiatum layer and the molecular layer. Similar results are obtained in the dentate gyrus where the binding level is different in the granular cell layer and the hilus. The specificity of bFGF binding on the different regions of the
adult hippocampus is also confirmed by a chase experiment (fig 4). After heparitinase treatment, the labeling on the capillaries, blood vessels, meninges and choroid plexus completely disappears (fig 5). It corresponds to the binding site on one of the mean basement membrane components: the heparan sulfate proteoglycans (HSP). The specific cellular binding pattern of the hippocampus is conserved. During hippocampus maturation (from birth to 15 days PN) the pyramidal ammonic cells have aChieved their maturation and the labeling of the ammonic horn is clearly distributed on the cellular processes and not on the cell bodies (fig 6). Comparison of the labeling in the stratum radiatum shows that there is no significant difference between the 15 day PN versus 1-month-old (fig 6a, c). Conversely, in the dentate gyrus, at 15 days PN (where few granular cells have achieved their differentiation), there is a significant difference in the labeling of the dentate gyrus between the cell granular layer and the hilus (fig 6a, c). At 1 month the mossy fiber zone displays a mature pattern. The comparison of the labeling in the hilus at 15 days versus 1-month-old showed a highly significant difference with more intense labeling in the latter (table II) and especially in the mossy fiber layer at the level of the CA2, CA3, CA4 regions of the Ammon's horn. Binding o f 12SI-bFGF on choroid plexus
The choroid plexus occupies a large area in the lumen of the ventricles, especially during embryonic development. They are formed by an epithelial cell layer. There is a strong labeling on the basolateral face corresponding to the low affinity sites on the basement membrane, whereas the high affinity receptors likely correspond to the weaker labeling on the digitated apical face and the surface of the epithelial cells (fig 7a, c). The dense patches of silver grains around the ventricular surface correspond to the intensely labeled cells of the ependymal layer (fig 7d, e). Binding o f leSI-bFGF on cerebellum
The cerebellum was studied because it displays a stratified pattern reminiscent of characteristic patterns seen in other neural tissues, ie retina and hippocampus. The binding is observed on the cerebellar cortex in three distinct regions (fig 8). A high density of bFGF binding sites is localized on the molecular layer, a region poor in neural cell bodies but which contains the dendrites of Purkinje cells (PCs).
Table I. In situ binding of iodinated bFGF (20 ng/ml) to different regions and principal cell layers of hippocampus. Mean grey levels determinated by image analysis. Cell layers
Stratum oriens Stratum pyramidale Stratum radiatum Stratum lacunosum moleculare
Hippocampal formation Ammons's horn
Dentate gyrus
166
Stratum moleculare
119
+ 179 141 110 + 100 +19J 135 _+ 18
Stratum granulosum
+ 147 15]-1 + ll/ 68 J +4
Hilus fasciae dentatae
Data are presented as the mean of 15 measurements on each region of slides corresponding to figure 4, + standard deviation. Statistical significance was assessed, between the mean values by Student's t-test, t I difference statistically significant at P < 0.001.
bFGF binding in mouse brain
~
7
Fig 4. Brain sections through the hippocampus from 1-month-old mouse incubated with iodinated b F G F (a) in the presence of unlabeled b F G F (h) (bar = 300/~m). The organization of these horizontal sections is described in figure 6d. In competition with native growth factor, the initial binding is significantly decreased on the hippocampus.
Fig S. Brain sections through the hippocampus from 1-month-old mouse incubated with iodinated b F G F (a) after heparitinase treatment (b, c) (bar = 300 t~m). b, e correspond at higher magnification to the two areas surrounded in (a). The organization of these horizontal sections is described in figure 6d. Binding on the choroid plexus (ch pl) and meninges (m) completely disappears after specific heparitinase digestion of the heparan sulfate proteoglycans. The cellular labeling pattern of the hippocampus is still present.
8
|
NA Fayein e t a l
~!,.~-~~k,.~,.
.
~-,~ ..... ~--~:~:~....
.
. - L
~--:;:-]rc~'{
k~,f~.,, ~'~,~-~
,,,
-'
~',~
a-'
z~-
i
t~.~ : - ~
. ' - - ~-~-" i . " :,~,~,
• ;
.~
~-~',
,..
""~ L
WB~#
J
~;~:'-i-~',
~ ~. ~' . ~¢ , . ~ '~ ~. , f ~ . : ! ~" : : : ' ~ , , , - . , 5 :. 9 . .~.. . . ,. •~ ".~.. " ~ . . . . . ..2~ ""....~,~';~.~i~'" ~ ' ~. . . . . ,.":, . .>~
!i~"
i~j~rv~.~
I ' ~ , ~ . ' :-.
~;,~,~,,:,....
• .~._
~..~_,~'¢~3.~ . : . : ,
j
~'.'-',~:~-~.;22.'~'-~:~:'~',~.~ " ~ ""
-
.- ..: ' ~ h 4 - ' - ~ : .v-.~%,~,.~,L:,
.
"
'~,,-~,,.
~
"
"
:~..:.
.
,-'-~ .-,..., .., ,. ,..,.
:"~.~~.~.. , ~~, : • "'~'~" '
~k.
!
•
.
..
~'I"
X-, • ~.-~_~ ,,,-
. . . . "m,¢ k~"
"
~i~.~.-
- ~ ~ . " : . ~ - ~ ' - "
q . ,~., . - , : ,
~-','. . . . .
B
,~.
"
N
-. I . ' L - " ' i _ ".
cAa
I ! |
f t
¢ .... ¢_'...,," ihil.f.d.'',~Zff,,."-
.1
"
"
'
r
.
g
r
~
.,,-,
_o
R
~
\
\
X
•
•
d
Fig 6. Brain sections through the hippocampus at two different stages of PN development (15 days: a, b; 1 month: e) incubated with iodinated bFGF (a, c) (bar = 300/~m). d. Schematic representation of the hippocampus organization. At 15 days PN the layers of the cell bodies are clearly stained with hematoxylin-eosin (b). From 15 days to 1 month, the autoradiography shows a labeling progressively increased in the hilus of the dentate area and in the layer of mossy fibers (mr).
Table II. In situ binding of iodinated bFGF (20 ng/ml) to different cell layers of hippocampus during post natal development. Mean grey levels determinated by image analysis.
Hippocampal cell layers
15 days
One month I
Hilus fasciae dentatae
58.5-i + 2.8/
Stratum granulosum
-66.2/_~ +2.1
Stratum radiatum
59.7 _+ 2.8
I- . . . . . .
I
50.4 +__ 5.3
-1
58.5 + 2.8
Data are presented as the mean of 15 measurements on each region of slides corresponding to figure 7, +_ standard deviation. Statistical significance was assessed, between the mean values by Student's t-test., ~Difference statistically significant at P < 0.001. t - - - n Difference not statistically significant.
The granular layer, packed with different neural cell types, shows the lowest amount of b F G F binding sites. The third region, which corresponds to the fibers going to the cerebellar nucleus, shows an intermediate intensity of labeling. Thus, in cerebellum and hippocampus, bFGF binding
sites (ie high affinity receptors) are localized mainly on the areas rich in neuritic extensions. To confirm this observation and to define whether Purkinje cells were responsible for the high density binding observed in the molecular layer and the external granular layer, we incubated with ~25I-bFGF sections from two cerebellar mutant mice strains called P C D (Purkinje cell degeneration) [37] and TBL (tambaleante) [65], at a stage where PC degeneration is almost complete. In P C D (fig 8) as well as in TBL (results not shown) the pattern of labeling shows a major decrease in labeling intensity. This supports the view that the major part of the labeling was due to the dendritic extensions of PCs in the molecular layer. A residual labeling could be relevant to the other neural cell types. The structure of this molecular layer is conserved as shown by the histological staining.
Discussion
The results reported in this work demonstrate the presence of two different types of F G F receptors on mouse brain during embryonic and post-natal development. Competition experiments with unlabeled b F G F attest to the specificity of the binding. These specific receptors correspond to the two types previously described. The first one is localized on the basement membranes corresponding to the meninges, the choroid plexus and the blood vessels. This
bFGF binding in mouse brain
9
Fig 7. Sections through the choroid plexus from mouse brain incubated with iodinated bFGF (a, b, d, e) or stained with hematoxylineosin (c) (bar = 100 ~m). a, c, d. Bright field, b, e. Phase contrast. There is a strong labeling on the basolateral (bl) face corresponding to the basement membrane and a weaker labeling on the plasma membrane of the cells (apical face: ap). The intensely labeled cells around the ventricular surface correspond to the ependymal cells (ep).
binding site is associated with the heparan sulfate proteoglycans localized in these basement membranes [57]. We have already reported [6, 10, 22, 53, 57] that lens capsule, inner limiting membrane of the retina, Bruch's and Descemet's membranes, vessel basement membranes, and EHS tumor contain such binding sites. Its K d value is approximately 20 nM. The binding of FGFs to heparan sulfate chains of proteoglycans appears to protect the growth factors from degradation. Moreover, proteoglycan binding has been thought to be important in providing a matrix-bound or cell surface-bound reservoir of FGFs [49]. More recently, Yayon et al [67] showed that the binding of bFGF to its receptor requires prior binding either to the heparan sulfate side chains of a membrane heparan sulfate proteoglycan or to free heparan sulfate (heparin) chains. This new information on the cooperativity between high and low affinity receptors ([41], Mascarelli and Courtois, unpublished results) to participate in the formation of the active FGF receptor allows us to interprete the 'results of the specificity experiments performed on frozen sections in this and previous studies. Firstly, the low affinity sites are not saturated with 125I-bFGF. Secondly, bFGF (as well as aFGF) can oligomerize very easily in presence of heparan sulfate proteoglycans or heparin. Thus in a chase experiment with 125I-bFGF an increasing amount of unlabeled FGF will begin by occupying all free sites, then
will oligomerize with bound 125I-FGF and will partially displace it only at extremely high doses. This explains why a residual labeling is still present (fig 1). Conversely, heparitinase will digest easily all the potential low affinity binding sites and will leave mostly'intact the high affinity receptors. This is illustrated by the high sensitivity of the basement membrane labeling toward the heparitinase pretreatment. The second site is essentially localized on the neuritic extensions of the neurons in hippocampus and cerebellum. Their relative insensitivity to heparitinase treatment makes them likely to be similar to the high affinity receptors with a K d of 2 0 - 2 0 0 pM observed in cultured cells or in purified membrane preparations. It is similar to the one described on fetal hippocampal neurons [62] and also on PC12 cells [40], plasma membrane preparations from rod outer segments [31], adult bovine Muller cells [32] and baby hamster kidney (BHK) cells [39]. Our data, however, do not permit us to distinguish between the different types or subtypes of receptors (RI: flg in human, bFGF-R in mouse, cek 1 in chick, R2, R3, etc). We tried to assess the specific bFGF labeling at the different stages of neuronal development in the brain. At 10-12 days ED [10] the neuroblasts are actively dividing and the labeling is homogeneously distributed on the neural tube as described previously. At 13-14 days ED, the telencephalon is the highest labeled structure and the hetero-
10
NA Fayein et al .
~
~
.
.
.
.
,
.
,
z
~
~
~
~,..
,
. . , .
......
Fig 8. Brain sections through the cerebellum from 1-month-old normal (a, b) and PCD (c, d) mice incubated with iodinated bFGF (a, e) or stained with hematoxylin-eosin (b, d) (bar = 300 mm). The binding sites are preferentially localized in the molecular layer (ml) of the cerebellar cortex and seem reduced in the granuler cell layer (gr). In PCD mice, a much less intense labeling remains in the molecular layer. It is heterogeneously distributed, by: blood vessels; c: capillaries; m: meninges.
geneous labeling may correspond to both stages of proliferation and migration [63]. Cortical neurons are produced near the lateral ventricle early in gestation and migrate radially toward the cortical surface. The cortical layers are generated in a defined, 'inside out' sequence, with neurons destined for the deepest cortical layers formed first and neurons destined for more superficial layers formed in sequence later. At the middle of the period of cortical neurogenesis (16-17 days ED) the labeling corresponds to the ventricular and subventricular layers. It can be correlated with both the migration of neurons to their final location, and to the differentiation as well as the establishment of synaptic contacts. The migration concerns both neurons and glial cells. Finishing their last mitosis, the neuroblasts move to their final position guided by radial cells, whose extensions go from the germinative layer to the cortical surface [47]. Thus within a few days, in this structure there is a considerably redistribution of the binding sites. The mast cells, intensely labeled at the same stage under the ectoderm, are known to synthesize heparin and probably bind FGF through a heparin binding site. The development of the hippocampus proceeds in graded fashion with the formation of the pyramidal cell layer before birth, and of the dentate gyrus which is mainly postnatal. The maturation of the pyramidal cells occurs be-
tween the 18th embryonic day and day 15th post-natal. A large periventricular zone of labeling is observed before birth during the period of neurogenesis and of neuroblast migration, whereas at the 15th day post-natal the binding sites are clearly compartmented in the extension cells layers of the CA1, CA2 and CA3 sectors of the Ammon's horn. These findings support the role of FGF in development at an early embryonic stage. This may include proliferation and migration of the neuroblasts from the ventricular surface according to the inside-out sequence and later, when the neurons have reached their final destination, the growth and differentiation of the pyramidal cell processes. In the dentate gyrus, it appears that the distribution and compartmentalization of the labeling in the molecular layer, hilus and mossy fibers layer parallel the cytoarchitectonic maturation of the granular cells. The labeling within the hilus that is weaker at 15 days PN develops a mature pattern at 1 month PN. At 15 days PN, the late forming cells had not yet completed their differentiation. The density of silver grains seems to correlate with the differentiation of granule cells, which give rise to the mossy fibers, develop a dense plexus of collaterals and the main axons continue in to the inferior region of Ammon's horn CA3, CA4, where they form prominent mossy fiber layer [12]. We observed at 1 month, but not at 15 days
bFGF binding in mouse brain PN, high density binding sites in the narrow band which correspond to the mossy fibers layer. These observations suggest that bFGF is involved in different functions in the development of the dentate gyrus, including differentiation of granule cells as well as elongation of mossy fibers and could also play a role in synapse formation between mossy fibers and the proximal dendrites of pyramidal cells. The presence of an intense labeling in the cerebellar molecular layer suggests that the rich dendritic arborisation of the PCs possess a major part of the receptors. Wanaka et al [64] did not observe a FGF-R1 mRNA expression in the rat PCs while Heuer et al [17], with a sequence encoding the extracellular domain of the chicken FGF-R, found a mRNA expression in the cerebellar granular cell layer and PCs. Purkinje cells could express a specific type of receptors [43]. In mouse cerebellum, by in situ hybridization, we observed a mRNA expression in PCs with a specific FGF-R2 probe derived from TK14 [19] (unpublished data). The pattern of iodinated bFGF binding correlated with specific developmental events suggests that FGFs play a significant role in both central nervous system development and neuronal viability in the adult brain. There are many studies supporting that view. Expression and localization of both FGF and FGF receptors in neuronal cells and biological effects of FGF on the same cells suggest a paracrine or autocrine action. Both acidic and basic FGFs are abundant in the central nervous system (CNS). They have been purified from bovine brain [14, 20, 44, 54] and rat brain [46]. They are present in all or almost all brain neurons [21, 45]. A regional analysis of the expression of bFGF mRNA in brain reveals that it is widely distributed in the cortex (frontal, parietal, and occipital), the hippocampus, hypothalamus and pons [9]. aFGF mRNA has been localized in the developing and adult rat brain using in situ hybridization histochemistry [66]. Prenatally, hybridization to aFGF mRNA was observed throughout the brain, with the strongest signal associated with cells of the developing cortical plate. Postnatally, labeling was localized to specific neuronal populations. In the hippocampus, labeling of the pyramidal cell layer and dentate granule cells was observed and became progres-" sively more intense with maturation. Labeling was also observed in both the external and internal granule cell layers of the developing cerebellum. In addition to FGF localization and expression in brain, a recent report has demonstrated the distribution of an mRNA for aFGF receptor in rat brain using in situ hybridization and an flg clone [64]. The most intense hybridization signal is observed in the hippocampus and in the positive cholinergic neurons. Cerebellar granule cells and spinal cord neurons are also positive for FGF-R mRNA. The biological effects of a- and bFGF on neuronal cells are multiple. Over the last 3 years, several in vitro studies demonstrated that a- and bFGFs can promote the survival of embryonic neurons of various regions of the brain and induce neurite outgrowth in a variety of cells (reviewed in [3]). a- and bFGF support initial survival and subsequent fiber outgrowth of dissociated rodent fetal neurons in culture [36, 61]. bFGF stimulates cell proliferation'[13] neurite outgrowth [60], and cell survival in PC12 cells [51] and neurons from various brain regions of postnatal and fetal rats, including the hippocampus, neocortex, striatum, septum and thalamus [33, 59]. bFGF also displays a potent neurotrophic action on rat cerebral cortical neurons [35] and hippocampal neurons [58], implying that these cells possess FGF receptors [40]. The stimulation of neurite
11
growth is strongly correlated with the amount of 125IbFGF bound by different glycosaminoglycans (heparin, heparan sulfate or hyaluronic acid) [60]. In addition to its role on neurite outgrowth and synaptogenesis aFGF seems to maintain neuronal viability and • connectivity in various experimental brain injuries [1, 16, 30, 42, 52]. In conclusion, this study provides evidence that the presence of bFGF receptors can be correlated with brain developmental events such as proliferation, migration, differentiation and synaptogenesis. It would be interesting to determine the type, number, affinity, level of expression of such receptors in degenerative or survival processes. Work in this area is in progress.
Acknowledgments The authors thank Dr D Mc Devitt for his review of the manuscript, Dr C Sotelo for his scientific advice about cerebellum experiments, Dr JL Gu~net for providing mutant mice, L Jonet for excellent technical assistance, H Coet for the photographs and N Breugnot and Y Maville for typing.
References 1 Anderson KJ, Dam D, Lee S, Cotman CW (1988) Basic fibroblast growth factor prevents death of lesioned cholinergic neurons in vivo. Nature 332, 360-361 2 Baird A, Waiicke PA (1989) Fibroblast growth factor. Br Med Bull 45,438-452 3 Barde YA (1989) Trophic factors and neuronal survival. Neuron 2, 1525-1534 4 Besnard F, Perraud F, Sensenbrenner M, Labourdette G (1989) Effects of acidic and basic fibroblast growth factors on proliferation and maturation of cultured rat oligodendrocytes. Int J Dev Neurosci 7, 401-409 5 Bikfalvi A, Dupuy E, Inyang AL, Fayein N, Leseche G, Courtois Y, Tobelem G (1989) Binding, internalization and degradation of basic fibroblast growth factor in human microvascular endothelial cells. Exp Cell Res 181, 75-84 6 Cirillo A, Arruti C, Courtois Y, Jeanny JC (1990) Localization of basic fibroblast growth factor binding sites in the chick embryonic neural retina. Differentiation 45, 161-167 7 Durand J, Karpousas I, Salvodelli M, Keller N, Pouliquen Y (1990) Corneal stroma quantimetric analysis: Data acquisition optimization. Cybernetica 2, 107-114 8 Eccleston PA, Silberberg DH (1985) Fibroblast growth factor is a mitogen for oligodendrocytes in vitro. Brain Res 353, 315-318 9 Emoto N, Gonzalez AM, Walicke PA, Wada E, Simmons DM, Shimasaki S, Baird A (1989) Basic fibroblast growth factor (FGF) in the central nervous system: identification of specific loci of basic FGF expression in the rat brain. Growth Factors 2, 21-29 10 Fayein NA, Courtois Y, Jeanny JC (1990) Ontogeny of basic fibroblast growth factor binding sites in mouse ocular tissues. Exp Cell Res 188, 75-88 11 Feige J J, Baird A (1988) Glycosylation of the basic fibroblast growth factor receptor. The contribution of carbohydrate to receptor function. J B i o l Chem 263, 14023-14029 12 Gaarskjaer FB (1985) The development of the dendrate area and the hippocampal mossy fiber projection of the rat. J Comp Neurol 241, 154-170 13 Gensburger C, Labourdette G, Sensenbrenner M (1987) Brain basic fibroblast growth factor stimulates the proliferation of rat neuronal precursor cells in vitro. FEBS Lett 217, I-5 14 Gospodarowicz D, Cheng J, Lui GM, Baird A, Bollen P (1984) Isolation of brain fibroblast growth factor by heparine-sepharose affinity chromatography: identity with
12
15
16
17
18 19
20
21 22
23
24 25 26 27
28
29
30 31 32 33 34
NA Fayein et al pituitary fibroblast growth factor. Proc Natl A cad Sci USA 81, 6963-6967 Gospodarowicz D, Neufeld G, Schweigerer L (1986) Molecular and biological characterization of fibroblast growth factor, an angiogenic factor which also controls the proliferation and differentiation of mesoderm and neuroectoderm derived cells. Cell Differ 19, 1-17 Hefti F, Hartikka J, Kniisel B (1989) Function of neurotrophic factors in the adult and aging brain and their possible use in the treatment of neurodegenerative diseases. Neurobiol Aging 10, 515-533 Heuer JG, Von Bartheld CS, Kinoshita Y, Evers PC, Bothwell M (1990) Alternating phases of FGF receptor and NGF receptor expression in the developing chicken nervous system. Neuron 5, 283-296 Hicks D, Courtois Y (1988) Acidic fibroblast growth factor stimulates opsin levels in retinal photoreceptor cells in vitro. FEBS Lett 234, 475-479 Houssaint E, Blanquet PR, Champion-Arnaud P, Gesnei MC, Torriglia A, Courtois Y, Breathnach R (1990) Related fibroblast growth factor receptor genes exist in the human genome. Proc Natl Acad Sci USA 87, 8180-8184 Jacquot Dourges MA, Gulino D, Courty J, Badet J, Barritault D, Jozefonvicz J (1990) Affinity chromatography of fibroblast growth factors on substituted polystyrene. J Chromatogr 526, g5-45 Janet T, Miehe M, Pettrnann B, Labourdette G, Sensenbrenner M (1987) Ultrastructural localization of fibroblast growth factor in neurons of rat brain. Neurosci Lett 80, 153-157 Jeanny JC, Fayein N, Moenner M, Chevallier B, Barritault D, Courtois Y (1987) Specific f'Lxationof bovine brain and retinal acidic and basic fibroblast growth factors to mouse embryortic eye basement membranes. Exp Cell Res 171, 63-75 Kiefer MC, Stephans JC, Crawford K, Onico K, Barr PJ (1990) Ligand-affinity cloning and structure of a cell surface heparan sulfate proteoglycan that binds basic fibroblast growth factor. Proc Natl A cad Sci USA 87, 6985- 6989 Kniss DA, Burry RW (1988) Serum and fibroblast growth factor stimulate quiescent astrocytes to re-enter the cell cycle. Brain Res 439, 281-288 Kornbluth S, Paulson KE, Hanafusa H (1988) Novel tyrosine kinase identified by phosphotyrosine antibody screening of cDNA libraries. Mol Cell Biol 8, 5541-5544 Laemmli VH (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685 Ledoux D, Mereau A, Pieri I, Barritault D, Courty J (1991) High affinity receptors to acidic and basic fibroblast growth factor (FGF) are detected mainly in adult brain membrane preparations but not in liver, kidney, intestine, lung or stomach. Growth Factors 5, 221-231 Lee PL, Johnson DE, Cousens LS, Fried VA, Williams LT (1989) Purification and complementary DNA cloning of a receptor for basic fibroblast growth factor. Science, 245, 57-60 Lipton SA, Wagner JA, Madison RD, D'Amore PA (1988) Acidic fibroblast growth factor enhances regeneration of processes by postnatal .mammalian retinal ganglion cells in culture. Proc Natl Acad Sci USA 85, 2388-2392 Logan A (1988) Elevation of acidic fibroblast growth factor mRNA in lesioned rat brain. Mol Cell Endocrinol 58, 275-278 Mascarelli F, Rauials D, Courtois Y (1989) Fibrohlast growth factor phosphorylation and receptors in rod outer segments. EMBO J 8, 2265-2273 Mascarelli F, Tassin J, Courtois Y (1991) Effect of FGFs on adult bovine Muller cells: proliferation, binding and internalization. Growth Factors 4, 81-95 Matsuda S, Salto H, Nishiyama N (1990) Effect of basic fibroblast growth factor on neurons cultured from various regions of postnatal rat brain. Brain Res 520, 310-316 Mereau A, Pieri I, Gamby C, Courty J, Barritault D (1989) Purification of basic fibroblast growth factor receptors from bovine brain. Biochimie 71, 865-871
35 Morrison RS, Sharma A, de Vellis J, Bradshaw A 0986) Basic fibroblast growth factor supports the survival of cerebral cortical neurons in primary culture. Proc NatlAcad Sci USA 83, 7537-7541 36 Morrison RS, Keating RF, Moskal JR (1988) Basic fibroblast growth factor and epidermal growth factor exert differential trophic effects on CNS neurons. JNeurosci Res 21, 71-79 37 Mullen R J, Eicher EM, Sidman RL (1976) Purkinje cell degeneration, a new neurological mutation in the mouse. Proc Natl Acad Sci USA 73,208-212 38 Murphy M, Drago J, Bartlett PF (1990) Fibroblast growth factor stimulates the proliferation and differentiation of neural precursor cells in vitro. J Neurosci Res 25, 463-475 39 Neufeld G, Godpodarowicz D (1985) The identification and partial characterization of the fibroblast growth factor receptor of baby hamster kidney cells. J Biol Chem 260, 13860-13868 40 Neufeld G, Gospodarowicz D, Dodge L, Fujii DK (1987) Heparin modulation of the neurotrophic effects of acidic and basic fibroblast growth factors and nerve growth factor on PC12 cells. J Cell Physiol 131, 131-140 41 Ornitz DM, Yayon A, Flanagan JG, Svahn CM, Levi E, Leder P (1992) Heparin is required for cell-free binding of basic fibroblast growth factor to a soluble receptor and for mitogenesis in whole cells. Mol Cell Biol 12, 240-247 42 Otto D, Unsicker K (1990) Basic FGF reverses chemical and morphological deficits in the nigrostriatal system of MPTPtreated mice. J Neurosci 10, 1912-1921 43 Partanen J, Vainikka S, Korhnen J, Armstrong E, Alitalo K (1992) Diverse receptors for fibroblast growth factors. Prog Growth Factor Res 4, 69-83 44 Pettmann B, Weibel M, Sensenbrenner M, Labourdette G (1985) Purification of two astroglial growth factors from bovine brain. FEBS Lett 189, 102-108 45 Pettmann B, Labourdette G, Weibel M, Sensenbrenner M (1986) The brain fibroblast growth factor (FGF) is localized in neurons. Neurosci Lett 68, 175-180 46 Presta M, Rusnati M, Maler JAM, Ragnotti G (1988) Purification of basic fibroblast growth factor from rat brain: identification of a M r 22000 immunoreactive form. Biochem Biophys Res Commun 155, 1161-1172 47 Rakic P (1986) Specification of cerebral cortical areas. Science 241, 170-176 48 Reid HH, Wilks AF, Bernard O (1990) Two forms of the basic fibroblast growth factor receptorqike mRNA are expressed in the developing mouse brain. Proc Natl Acad Sci USA 87, 1596-1600 49 Ruoslahti E, Yamaguchi Y (1991) Proteogiycans as modulators of growth factor activities. Cell 64, 867-869 50 Saunders S, Jalkanen M, O'Farrell S, Bernfield M (1989) Syndecan, an integral membrane proteoglycan. J Cell Biol 108, 1547-1556 51 Schubert D, Ling N, Baird A (1987) Multiple influences of a heparin-binding growth factor on neuronal development. J Cell Biol 104, 635-643 52 Sievers J, Hausmann B, Unsicker K, Berry M (1987) Fibroblast growth factors promote the survival of adult rat retinal ganglion cells after transection of the optic nerve. Neurosci Lett 76, 157-162 53 Soubrane G, Jerdan J, Karpouzas I, Fayein NA, Glaser B, Coscas G, Courtois Y, Jeanny JC (1990) Binding of basic fibroblast growth factor to normal and neovascularized rabbit cornea. Invest Ophthalmol Vis Sci 31, 323-333 54 Thomas KA, Rios Candelore M, Fitzpatrick S (1984) Purification and characterization of acidic fibroblast growth factor from bovine brain. ProcNatlAcadSci USA 81,357-361 55 Uhlrich S, Lagente O, Lenfant M, Courtois Y (1986) Effect of heparin on the stimulation of non-vascular cells by human acidic and basic FGF. Biochem Biophys Res Commun 137, 1205-1213 56 Unsicker K, Reichert-Preibsch H, Schmidt R, Pettmann B, Labourdette G, Sensenbrenner M (1987) Astroglial and fibroblast growth factors have neurotrophic functions for
bFGF binding in mouse brain cultured peripheral and central nervous system neurons. Proc Natl Acad Sci USA 84, 5459-5463
57 Vigny M, OUier-Hartmann MP, Lavigne M, Fayein N, Jeanny JC, Laurent M, Courtois Y (1988) Specific binding of basic fibroblast growth factor to basic fibroblast growth factor to basement membrane-like structures and to purified heparan sulfate proteoglycan of the EHS tumor. J Cell Physiol 137, 321-328 58 Walicke P, Cowan WM, Ueno N, Baird A, Guillemin R (1986) Fibroblast growth factor promotes survival of dissociated hippocampal neurons and enhances neurite extension. Proc Natl Acad Sci USA 83, 3012-3016 59 Walicke PA (1988) Basic and acidic fihroblast growth factors have trophic effects on neurons from multiple CNS regions. J Neurosci 8, 2618-2627 60 Walicke PA (1988) Interactions between basic fibroblast growth factor (FGF) and glycosaminoglycans in promoting neurite outgrowth. Exp Neurol 102, 144-148 61 Walicke PA, Baird A (1988) Neurotrophic effects of basic and acidic fibroblast growth factors are not mediated through glial cells. Dev Brain Res 40, 71-79
13
62 Walicke PA, Feige J J, Baird A (1989) Characterization of the neuronal receptor for basic fibroblast growth factor and comparison to receptors on mesenchymal cells. JBiol Chem 264, 4120-4126 63 Walsh C, Cepko CL (1988) Clonally related cortical cells show several migration patterns. Science 241, 1342-1345 • 64 Wanaka A, Johnson EM, Milbrandt J (1990) Localization of FGF receptor mRNA in the adult central nervous system by in situ hybridization. Neuron 5, 267-281 65 Wassef M, Sotelo C, Cholley B, Brehier A, Thomasset M 0987) Cerebellar mutations affecting the postnatal survival of Purkinje cells in the mouse disclose a longitudinal pattern of differentially sensitive cells. Dev Biol 124, 379-389 66 Wilcox BJ, Unnerstall JR (1991) Expression of acidic fibroblast growth factor mRNA in the developing and adult rat brain. Neuron 6, 397-409 67 Yayon A, Klagsbrun M, Esko JD, Leder P, Ornitz DM (1991) Cell surface, heparin-like molecules are required for binding of basic fibroblast growth factor to its high affinity receptor. Cell 64, 841-848
1
© Elsevier, Paris
Original article
Basic fibroblast growth factor high and low affinity binding sites in developing mouse brain, hippocampus and cerebellum Nicole-Adeline Fayein, Yves Courtois, Jean-Claude Jeanny * Unitd de Recherches G~rontologiques, INSERM Ul18, CNRS UA630, Ddveloppement et S~nescence Cellulaire, Association Claude Bernard, 29, rue Wilhem, 75016 Paris, France (Received 17 July 1992; accepted 19 October 1992)
Summary - Fibroblast growth factors (FGFs), first extracted from brain and retina, are potent neurotrophic factors. They stimulate neuroblast proliferation and neuron differentiation and survival. In order to study the spatial and temporal distribution of the target cells in the mouse brain we studied by autoradiography and quantified by image analysis ~251-bFGFbinding sites as a function of development. We have revealed the presence of two types of specific bFGF receptors. One is heparitinase sensitive and is co-localized with heparan sulfate proteoglycans of the basement membranes (meninges, choroid plexus and blood vessels). It is not developmentally regulated and corresponds to the low affinity receptors. It may be a storage form. The second type is heparitinase resistant and is modified during development, matching, in the adult, layering of the hippocampus and cerebellum. At 13 days of embryonic development there is a preferential distribution of silver grains on the ecto- and neuroectodermal tissues. In the adult, the labeling is localized on the neural process layers. It likely corresponds to the specific binding to cell high affinity receptors. Binding patterns according to the developmental stages of the brain can be correlated with mitotic, migration and differentiation phases of the neuronal cells. basic fibroblast growth factor / high and low affinity receptors / mouse brain / hippocampus / cerebellum
Introduction Basic and acidic fibroblast growth factors (bFGF a n d aFGF) are mitogenic and neurotrophic. They have been isolated from a wide range of tissues, including brain and retina [2]. bFGF is a mitogen for a variety of mesenchymal cell types [15], for neuroectodermal-derived cells, glial cells [4, 8, 24] as well as neuronal precursors [13, 38, 51]. Its main property, however, seems to be neurotrophic for various neurons of multiple CNS regions [35, 56, 58, 59, 61], including retina [29, 52] and photoreceptors [18]. bFGF is acting through specific cell surface receptors which have been characterized in membranes prepared from various tissues of the CNS: brain [27, 34], retina [31] or neural cells in culture [40, 62]. The localisation of these receptors during retinal embryogenesis and development has been demonstrated in mice [I0, 22] and chicken [6] ocular tissues by autoradiographic techniques. By this technique two types of binding sites could be determined, depending on their resistance to heparitinase digestion and sensitivity to heparin or N-glycanase treatment. Thef first two treatments revealed the low affinity sites that were correlated with the presence of heparan sulfate in basement membranes or at the cell surface, the latter with the high affinity receptor sites for which the N-glycan moiety is necessary for its activity [11].
* Correspondence and reprints
The FGF receptor (FGF-R) is a transmembrane protein that contains three extracellular immunoglobulin-like domains with an unusual acidic region, and an intracellular tyrosine kinase domain [28, 43]. There are three different families of related FGF receptors encoded by three different genes. Complementary DNA clones were isolated with oligonucleotidic probes corresponding to determined amino acid sequences of fragments of the purified receptor. Two clones were isolated from mouse (bFGF-R and bek) [25, 48]. Such clones have been used by in situ hybridization to study the localization and the expression of the genes for FGF receptors during chick and mouse embryonic development [17, 64]. The cell surface proteoglycan, syndecan [50], that is likely to interact with bFGF, has also been recently cloned [23]. In spite of these results we do not know yet the distribution in the brain of the different types of proteins coded by these genes nor their activity in situ. In this study we have analysed by autoradiography, after incubation with iodinated bFGF, the binding sites and their specificity in mouse brain and in more detail in hippocampus and cerebellum, during development. This experimental approach does not allow to discriminate between the various classes of high and low affinity FGF receptors. It reflects their global spatio-temporal localization at various times of development and potentially their accessibility towards the various members of FGF family. We have found the distribution of bFGF receptors (high and low affinity) to be correlated with the stage of differentiation in various neuronal layers.
2
NA Fayein et al
Materials and methods
Chemicals and reagents The following reagents, were used: heparin-Sepharose, PD 10 columns (Pharmacia, St Quentin en Yvelines, France), t2SI-Na (Amersham, Les Ulis, France), optimal cutting temperature medium (OCT) and heparitinase (Miles Scientific, Puteaux, France) hematoxylin and eosin (Gurr). Other chemicals were obtained from Sigma Chemical Co (La Verpilliere, France), Prolabo (Paris, France) and Bio-Rad (Paris, France) and were reagent grade.
were washed in PBS and incubated with iodinated bFGF as above.
Analysis system The analysis system [7] consisted of a camera linked to monitors; data input was directly into a computer system. The whole was controlled by a series of Imagenia R (Biocom, Les Ulis, France) software. After adjusting the image of the monitor, image focusing and image treatment, the computer analyzed automatically the labeling intensity of different structures from brain.
Data acquisition Growth factor preparation and iodination Human recombinant basic FGF was a gift from Carlo Erba (Farmitalia). It was labeled with ~25I according to the chloramine-T method [5] as already described [10]. The specific activity of the 125I-bFGF purified on heparin-Sepharose was 1500-2000 Ci/ mmol. When this iodinated material was subjected to electrophoresis on sodium dodecyl'sulfate (SDS)-polyacrylamide gels [26] a single band migrating in the position of human recombinant bFGF was observed by autoradiography. Full biological activity of bFGF as tested by [3H]thymidine incorporation in bovine epithelial lens cells (BEL) [55], was recovered after iodination.
Preparation of mouse brain frozen sections Pregnant mice (OF1) were obtained from IFFA Credo (Domalne des Oncins, France). Animals were killed by inhalation of carbon dioxide from dry ice. PCD (Purkinje cell degeneration) [37] and TBL (Tambaleante) [65] mice'were a gift from Dr JL Gu6net (Institut Pasteur, Paris, France). Brains of 10- to 18-day-old embryos and 2- to 4-week-old postnatal animals were dissected, embedded in Tissue-Tek OCT compound (embedding medium for frozen tissue specimens) on a metal block, and frozen with liquid nitrogen. Longitudinal 10-/~m sections were cut at - 2 0 ° C in a cryostat (SLEE, London or OTF/AS, Bright, DIS, Blanc Mesnil, France), collected on gelatin-coated glass slides, and stored at - 2 0 ° C until use.
Incubation with iodinated bFGF and autoradiographic techniques Non-fixed sections, dried for 5 min at room temperature after OCT extraction in PBS, were incubated for 1 h at ambient temperature in a water-saturated atmosphere with 20/zl of 1251b F G F diluted in PBS (20 ng/ml, specific activity: 1500:-2000 Ci/mmol). Following the incubation, the slides were washed six times, 5 min each, with PBS. They were then fixed with 3% paraformaldehyde in PBS for 30 min, washed, dehydrated and dried before dipping. They were dipped in photoemulsion LM1 (Amersham). The slides were exposed for 2 - 4 weeks at 4°C, developed in Kodak DI9, and fixed in Hypare Ilford fix. The photographs were taken under the same conditions, using PAN-F film (Ilford) on a Zeiss photomicroscope (Oberkochen, Germany). All experiments were run in triplicate, and each treatment to test the specificity and nature of the binding was performed on a minimum of three slides.
Competition experiment Unlabeled purified bFGF (5/~g/ml) was mixed with its respective iodinated form (20 ng/ml) and deposited on the sections. After 1 h of incubation the slides were treated as above.
Heparitinase treatment of the sections Unfixed sections were treated with various concentrations of heparitinase (25, 50, 100 mU/ml), in the presence of Ca 2÷ in a final volume of 20 tzl, for I h at 37°C. After treatment slides
Images were acquired on film negatives by a charge couple device (CCD) camera, 512 x 512 pixels and 100 grey levels (expressed from 0 to 255). The negatives from each figure were placed on a light box and held flat by a weighted glass slide. The analogue video images were digitized in the computer where the grey levels were calculated. Ten measurements were taken from each hippocampal structure (Stratum oriens, St pyramidale, St radiatum, St lacunosum moleculare, St moleculare, St granulosum, Hilus fasciae dentatae) and also from the background (out of the sections). Data were presented as the mean + standard deviation. Student's t-test was carried out between different hippocampal structures and at two stages of post-natal development. The comparisons were done on sections of the same experiment developed after the same exposure time.
Histological staining After OCT extraction in PBS the sections were stained with hematoxylin and eosin. Photographs were taken as previously described.
Results
Binding o f iodinated bFGF on brain sections during embryonic development At 1 3 - 1 4 days o f embryonic development (ED), the telencephalon and diencephalon are growing a r o u n d the lateral and third ventricles and start to differentiate. There is a differential distribution o f 125I-bFGF binding sites on horizontal head sections. The basement membranes are the highest labeled structures with 125I-bFGF (fig la) for instance below the ectoderm, nasal cavities, tongue, blood vessels and pial surfaces. Strong labeling remains in the developing telencephalon and diencephalon, but with a heterogeneous distribution. To confirm the specificity o f 125I-bFGF binding the sections were treated with 500-fold excess o f unlabeled bFGF. At this concentration (fig lb) there is a m a j o r chase o f the 125I-bFGF as we already described in previous studies in retina [10, 22]. A l t h o u g h the general labeling was decreased, the pattern remained similar. Within the telencephalon, there are two distinct, densely, labeled layers, one periventricular, next to the ventricles, the second in an intermediate zone under the cortical surface. These two layers coincide with the highest cell density (fig l c). At 1 6 - 1 7 days there is no change in the labeling o f the basement membranes beside a strong labeling on the newly-formed choroid plexus. The diencephalon is unif o r m l y labeled f r o m the ventricle to the external surface (fig 2a). The telencephalon displays a distinct pattern, different f r o m the one observed 2 days earlier (fig 2a, b). The labeling is n o w localized in the ventricular and subventricular layers, rather than in the intermediate zone as
bFGF binding in mouse brain
•
.
.
.
.
.
---~..~
.
_
=
.
-
..,-?.~,~,-..,
..-
~t.,.~.
'
~
~
.
~,~.a~,'
,
:~ .....
,.
...
-
.''~'i~
.
.
;:
.,~
~ '~,-
...... •
3
.
',_~r
.¢
,.,
Fig 1. Brain sections from 13-day-old mouse embryo incubated with iodinated bFGF (a) in the presence of unlabeled bFGF (b) or stained with hematoxylin-eosin (c) (bar = 300 tzm). The basement membranes are the strongest labeled structures, for instance below the ectoderm, nasal cavities (nc), tongue (t), blood vessels (bv) and pial surface (ps). The labeling is heterogeneously distributed on the telencephalon (Te) in two zones (periventricular: p; intermediate: i). Competition with native growth factor decreases the initial binding on all labeled structures, v: lateral ventricles.
4
NA Fayein et al
Fig 2. Brain sections from 16-day-old mouse embryo incubated with iodinated bFGF (a, b, d) or stained with hematoxylin-eosin (c, e) (bar = 300/~m). There is strong labeling on the newly-formed choroid plexus (ch pl) and mast cells (mc). The diencephalon (Di) is uniformly labeled from ventricle (v) to meninges (m) whereas the telencephalon (Te) displays a stronger sub- and periventricular labeling. Iv: lateral ventricle.
observed at 13-14 days. To match this pattern with the cellular pattern of this tissue, a neighbouring section has been stained with hematoxylin-eosin (fig 2c). It is apparent that the intense labeling is only colocalized with the periventricular cell layer, and not with the cell layer in the external zone. The dense patches of silver grains under the head ecto-
derm correspond to the labeling on mast cells as already identified in mouse, after metachromatic staining with toluidine blue. At the same stage in the periventricular zone of the telencephalon the labeling is observed more intensely on the external side of the lateral ventricle (fig 2d), progressing to the internal side corresponding to the presumptive zone
bFGF binding in mouse brain
a plexu$ chor.
',.
,,
b .
..,L~,'~.. - ~
•
~
7"4.,:;~
.'-,~
~
•
,"'~'--t":?
~-
Fig 3. Brain sections through the hippocampus from l-month-old mouse incubated with iodinated bFGF (a) or stained with hematoxylineosin (c) (bar = 300 ~m). h. Schematic representation of the hippocampus organization. The labeling is distributed on the hilus (hil) of the area dentata, the stratum radiatum (str rad hip) and with a decreasing intensity on the stratum lacunosum and moleculare (str lac mol hip). There is also a weaker labeling on the stratum oriens (str or hip). Layers of cell bodies are weakly labeled or not labeled: stratum pyramidale (str pyr) and stratum granulare (str gran).
6
NA Fayein et al
of the hippocampus. The staining of a neighbouring section indicates again that the labeling corresponds to the cell population at the border of the ventricle (fig 2e). The other external zones are not associated with a high density of silver grains. Binding o f J2sI bFGF on hippocampus
As the hippocampus has been a favorite structure in the brain for which a physiological role for FGF has been proposed, we focused our study on the developing hippocampus from the 16-17-day embryo to the adult stage. At ED 18, the pattern of labeling matchs precisely the hippocampic morphogenesis. It is localized on the pyramidai cell layer of the subiculum and hippocampus (data not shown). At 1 month post-natal (PN), the hippocampus reaches its final differential stage and is formed by several stratified layers with sharply defined neurons. At the level of the ammonic pyramid.al cell layer, there was no binding while a very dense binding is observed on the stratum radiatum, with a decreasing intensity to the stratum lacunosum and moleculare (fig 3). There was also a weaker labeling on the stratum oriens. At the level of the area dentata the binding is strongly localized in the hilus surrounded by the stratum granulosum which shows an apparent absence of labeling as already observed in the ammonic pyramidal layer (CA1 and CA3). The juxtaposition of the hematoxylin-eosin stained sections with the autoradiographic sections shows clearly that the poorly labeled areas correspond to the nuclear cell layers for both the hippocampus and the dentate gyrus. The high density of labeling is localized on cell processes but with different gradation. The other structures on which the labeling is intense correspond to the blood vessel walls and to the choroid plexus as observed at early stages. To quantify the binding to the different structures the grey levels were analysed by image analysis (table I). The data show a highly significant difference of binding between the pyramidal cell layers and the areas of dendritic ramifications (stratum oriens, radiatum and moleculare). Even within the zone of dendritic ramifications, there is a highly significant differential labeling between the radiatum layer and the molecular layer. Similar results are obtained in the dentate gyrus where the binding level is different in the granular cell layer and the hilus. The specificity of bFGF binding on the different regions of the
adult hippocampus is also confirmed by a chase experiment (fig 4). After heparitinase treatment, the labeling on the capillaries, blood vessels, meninges and choroid plexus completely disappears (fig 5). It corresponds to the binding site on one of the mean basement membrane components: the heparan sulfate proteoglycans (HSP). The specific cellular binding pattern of the hippocampus is conserved. During hippocampus maturation (from birth to 15 days PN) the pyramidal ammonic cells have aChieved their maturation and the labeling of the ammonic horn is clearly distributed on the cellular processes and not on the cell bodies (fig 6). Comparison of the labeling in the stratum radiatum shows that there is no significant difference between the 15 day PN versus 1-month-old (fig 6a, c). Conversely, in the dentate gyrus, at 15 days PN (where few granular cells have achieved their differentiation), there is a significant difference in the labeling of the dentate gyrus between the cell granular layer and the hilus (fig 6a, c). At 1 month the mossy fiber zone displays a mature pattern. The comparison of the labeling in the hilus at 15 days versus 1-month-old showed a highly significant difference with more intense labeling in the latter (table II) and especially in the mossy fiber layer at the level of the CA2, CA3, CA4 regions of the Ammon's horn. Binding o f 12SI-bFGF on choroid plexus
The choroid plexus occupies a large area in the lumen of the ventricles, especially during embryonic development. They are formed by an epithelial cell layer. There is a strong labeling on the basolateral face corresponding to the low affinity sites on the basement membrane, whereas the high affinity receptors likely correspond to the weaker labeling on the digitated apical face and the surface of the epithelial cells (fig 7a, c). The dense patches of silver grains around the ventricular surface correspond to the intensely labeled cells of the ependymal layer (fig 7d, e). Binding o f leSI-bFGF on cerebellum
The cerebellum was studied because it displays a stratified pattern reminiscent of characteristic patterns seen in other neural tissues, ie retina and hippocampus. The binding is observed on the cerebellar cortex in three distinct regions (fig 8). A high density of bFGF binding sites is localized on the molecular layer, a region poor in neural cell bodies but which contains the dendrites of Purkinje cells (PCs).
Table I. In situ binding of iodinated bFGF (20 ng/ml) to different regions and principal cell layers of hippocampus. Mean grey levels determinated by image analysis. Cell layers
Stratum oriens Stratum pyramidale Stratum radiatum Stratum lacunosum moleculare
Hippocampal formation Ammons's horn
Dentate gyrus
166
Stratum moleculare
119
+ 179 141 110 + 100 +19J 135 _+ 18
Stratum granulosum
+ 147 15]-1 + ll/ 68 J +4
Hilus fasciae dentatae
Data are presented as the mean of 15 measurements on each region of slides corresponding to figure 4, + standard deviation. Statistical significance was assessed, between the mean values by Student's t-test, t I difference statistically significant at P < 0.001.
bFGF binding in mouse brain
~
7
Fig 4. Brain sections through the hippocampus from 1-month-old mouse incubated with iodinated b F G F (a) in the presence of unlabeled b F G F (h) (bar = 300/~m). The organization of these horizontal sections is described in figure 6d. In competition with native growth factor, the initial binding is significantly decreased on the hippocampus.
Fig S. Brain sections through the hippocampus from 1-month-old mouse incubated with iodinated b F G F (a) after heparitinase treatment (b, c) (bar = 300 t~m). b, e correspond at higher magnification to the two areas surrounded in (a). The organization of these horizontal sections is described in figure 6d. Binding on the choroid plexus (ch pl) and meninges (m) completely disappears after specific heparitinase digestion of the heparan sulfate proteoglycans. The cellular labeling pattern of the hippocampus is still present.
8
|
NA Fayein e t a l
~!,.~-~~k,.~,.
.
~-,~ ..... ~--~:~:~....
.
. - L
~--:;:-]rc~'{
k~,f~.,, ~'~,~-~
,,,
-'
~',~
a-'
z~-
i
t~.~ : - ~
. ' - - ~-~-" i . " :,~,~,
• ;
.~
~-~',
,..
""~ L
WB~#
J
~;~:'-i-~',
~ ~. ~' . ~¢ , . ~ '~ ~. , f ~ . : ! ~" : : : ' ~ , , , - . , 5 :. 9 . .~.. . . ,. •~ ".~.. " ~ . . . . . ..2~ ""....~,~';~.~i~'" ~ ' ~. . . . . ,.":, . .>~
!i~"
i~j~rv~.~
I ' ~ , ~ . ' :-.
~;,~,~,,:,....
• .~._
~..~_,~'¢~3.~ . : . : ,
j
~'.'-',~:~-~.;22.'~'-~:~:'~',~.~ " ~ ""
-
.- ..: ' ~ h 4 - ' - ~ : .v-.~%,~,.~,L:,
.
"
'~,,-~,,.
~
"
"
:~..:.
.
,-'-~ .-,..., .., ,. ,..,.
:"~.~~.~.. , ~~, : • "'~'~" '
~k.
!
•
.
..
~'I"
X-, • ~.-~_~ ,,,-
. . . . "m,¢ k~"
"
~i~.~.-
- ~ ~ . " : . ~ - ~ ' - "
q . ,~., . - , : ,
~-','. . . . .
B
,~.
"
N
-. I . ' L - " ' i _ ".
cAa
I ! |
f t
¢ .... ¢_'...,," ihil.f.d.'',~Zff,,."-
.1
"
"
'
r
.
g
r
~
.,,-,
_o
R
~
\
\
X
•
•
d
Fig 6. Brain sections through the hippocampus at two different stages of PN development (15 days: a, b; 1 month: e) incubated with iodinated bFGF (a, c) (bar = 300/~m). d. Schematic representation of the hippocampus organization. At 15 days PN the layers of the cell bodies are clearly stained with hematoxylin-eosin (b). From 15 days to 1 month, the autoradiography shows a labeling progressively increased in the hilus of the dentate area and in the layer of mossy fibers (mr).
Table II. In situ binding of iodinated bFGF (20 ng/ml) to different cell layers of hippocampus during post natal development. Mean grey levels determinated by image analysis.
Hippocampal cell layers
15 days
One month I
Hilus fasciae dentatae
58.5-i + 2.8/
Stratum granulosum
-66.2/_~ +2.1
Stratum radiatum
59.7 _+ 2.8
I- . . . . . .
I
50.4 +__ 5.3
-1
58.5 + 2.8
Data are presented as the mean of 15 measurements on each region of slides corresponding to figure 7, +_ standard deviation. Statistical significance was assessed, between the mean values by Student's t-test., ~Difference statistically significant at P < 0.001. t - - - n Difference not statistically significant.
The granular layer, packed with different neural cell types, shows the lowest amount of b F G F binding sites. The third region, which corresponds to the fibers going to the cerebellar nucleus, shows an intermediate intensity of labeling. Thus, in cerebellum and hippocampus, bFGF binding
sites (ie high affinity receptors) are localized mainly on the areas rich in neuritic extensions. To confirm this observation and to define whether Purkinje cells were responsible for the high density binding observed in the molecular layer and the external granular layer, we incubated with ~25I-bFGF sections from two cerebellar mutant mice strains called P C D (Purkinje cell degeneration) [37] and TBL (tambaleante) [65], at a stage where PC degeneration is almost complete. In P C D (fig 8) as well as in TBL (results not shown) the pattern of labeling shows a major decrease in labeling intensity. This supports the view that the major part of the labeling was due to the dendritic extensions of PCs in the molecular layer. A residual labeling could be relevant to the other neural cell types. The structure of this molecular layer is conserved as shown by the histological staining.
Discussion
The results reported in this work demonstrate the presence of two different types of F G F receptors on mouse brain during embryonic and post-natal development. Competition experiments with unlabeled b F G F attest to the specificity of the binding. These specific receptors correspond to the two types previously described. The first one is localized on the basement membranes corresponding to the meninges, the choroid plexus and the blood vessels. This
bFGF binding in mouse brain
9
Fig 7. Sections through the choroid plexus from mouse brain incubated with iodinated bFGF (a, b, d, e) or stained with hematoxylineosin (c) (bar = 100 ~m). a, c, d. Bright field, b, e. Phase contrast. There is a strong labeling on the basolateral (bl) face corresponding to the basement membrane and a weaker labeling on the plasma membrane of the cells (apical face: ap). The intensely labeled cells around the ventricular surface correspond to the ependymal cells (ep).
binding site is associated with the heparan sulfate proteoglycans localized in these basement membranes [57]. We have already reported [6, 10, 22, 53, 57] that lens capsule, inner limiting membrane of the retina, Bruch's and Descemet's membranes, vessel basement membranes, and EHS tumor contain such binding sites. Its K d value is approximately 20 nM. The binding of FGFs to heparan sulfate chains of proteoglycans appears to protect the growth factors from degradation. Moreover, proteoglycan binding has been thought to be important in providing a matrix-bound or cell surface-bound reservoir of FGFs [49]. More recently, Yayon et al [67] showed that the binding of bFGF to its receptor requires prior binding either to the heparan sulfate side chains of a membrane heparan sulfate proteoglycan or to free heparan sulfate (heparin) chains. This new information on the cooperativity between high and low affinity receptors ([41], Mascarelli and Courtois, unpublished results) to participate in the formation of the active FGF receptor allows us to interprete the 'results of the specificity experiments performed on frozen sections in this and previous studies. Firstly, the low affinity sites are not saturated with 125I-bFGF. Secondly, bFGF (as well as aFGF) can oligomerize very easily in presence of heparan sulfate proteoglycans or heparin. Thus in a chase experiment with 125I-bFGF an increasing amount of unlabeled FGF will begin by occupying all free sites, then
will oligomerize with bound 125I-FGF and will partially displace it only at extremely high doses. This explains why a residual labeling is still present (fig 1). Conversely, heparitinase will digest easily all the potential low affinity binding sites and will leave mostly'intact the high affinity receptors. This is illustrated by the high sensitivity of the basement membrane labeling toward the heparitinase pretreatment. The second site is essentially localized on the neuritic extensions of the neurons in hippocampus and cerebellum. Their relative insensitivity to heparitinase treatment makes them likely to be similar to the high affinity receptors with a K d of 2 0 - 2 0 0 pM observed in cultured cells or in purified membrane preparations. It is similar to the one described on fetal hippocampal neurons [62] and also on PC12 cells [40], plasma membrane preparations from rod outer segments [31], adult bovine Muller cells [32] and baby hamster kidney (BHK) cells [39]. Our data, however, do not permit us to distinguish between the different types or subtypes of receptors (RI: flg in human, bFGF-R in mouse, cek 1 in chick, R2, R3, etc). We tried to assess the specific bFGF labeling at the different stages of neuronal development in the brain. At 10-12 days ED [10] the neuroblasts are actively dividing and the labeling is homogeneously distributed on the neural tube as described previously. At 13-14 days ED, the telencephalon is the highest labeled structure and the hetero-
10
NA Fayein et al .
~
~
.
.
.
.
,
.
,
z
~
~
~
~,..
,
. . , .
......
Fig 8. Brain sections through the cerebellum from 1-month-old normal (a, b) and PCD (c, d) mice incubated with iodinated bFGF (a, e) or stained with hematoxylin-eosin (b, d) (bar = 300 mm). The binding sites are preferentially localized in the molecular layer (ml) of the cerebellar cortex and seem reduced in the granuler cell layer (gr). In PCD mice, a much less intense labeling remains in the molecular layer. It is heterogeneously distributed, by: blood vessels; c: capillaries; m: meninges.
geneous labeling may correspond to both stages of proliferation and migration [63]. Cortical neurons are produced near the lateral ventricle early in gestation and migrate radially toward the cortical surface. The cortical layers are generated in a defined, 'inside out' sequence, with neurons destined for the deepest cortical layers formed first and neurons destined for more superficial layers formed in sequence later. At the middle of the period of cortical neurogenesis (16-17 days ED) the labeling corresponds to the ventricular and subventricular layers. It can be correlated with both the migration of neurons to their final location, and to the differentiation as well as the establishment of synaptic contacts. The migration concerns both neurons and glial cells. Finishing their last mitosis, the neuroblasts move to their final position guided by radial cells, whose extensions go from the germinative layer to the cortical surface [47]. Thus within a few days, in this structure there is a considerably redistribution of the binding sites. The mast cells, intensely labeled at the same stage under the ectoderm, are known to synthesize heparin and probably bind FGF through a heparin binding site. The development of the hippocampus proceeds in graded fashion with the formation of the pyramidal cell layer before birth, and of the dentate gyrus which is mainly postnatal. The maturation of the pyramidal cells occurs be-
tween the 18th embryonic day and day 15th post-natal. A large periventricular zone of labeling is observed before birth during the period of neurogenesis and of neuroblast migration, whereas at the 15th day post-natal the binding sites are clearly compartmented in the extension cells layers of the CA1, CA2 and CA3 sectors of the Ammon's horn. These findings support the role of FGF in development at an early embryonic stage. This may include proliferation and migration of the neuroblasts from the ventricular surface according to the inside-out sequence and later, when the neurons have reached their final destination, the growth and differentiation of the pyramidal cell processes. In the dentate gyrus, it appears that the distribution and compartmentalization of the labeling in the molecular layer, hilus and mossy fibers layer parallel the cytoarchitectonic maturation of the granular cells. The labeling within the hilus that is weaker at 15 days PN develops a mature pattern at 1 month PN. At 15 days PN, the late forming cells had not yet completed their differentiation. The density of silver grains seems to correlate with the differentiation of granule cells, which give rise to the mossy fibers, develop a dense plexus of collaterals and the main axons continue in to the inferior region of Ammon's horn CA3, CA4, where they form prominent mossy fiber layer [12]. We observed at 1 month, but not at 15 days
bFGF binding in mouse brain PN, high density binding sites in the narrow band which correspond to the mossy fibers layer. These observations suggest that bFGF is involved in different functions in the development of the dentate gyrus, including differentiation of granule cells as well as elongation of mossy fibers and could also play a role in synapse formation between mossy fibers and the proximal dendrites of pyramidal cells. The presence of an intense labeling in the cerebellar molecular layer suggests that the rich dendritic arborisation of the PCs possess a major part of the receptors. Wanaka et al [64] did not observe a FGF-R1 mRNA expression in the rat PCs while Heuer et al [17], with a sequence encoding the extracellular domain of the chicken FGF-R, found a mRNA expression in the cerebellar granular cell layer and PCs. Purkinje cells could express a specific type of receptors [43]. In mouse cerebellum, by in situ hybridization, we observed a mRNA expression in PCs with a specific FGF-R2 probe derived from TK14 [19] (unpublished data). The pattern of iodinated bFGF binding correlated with specific developmental events suggests that FGFs play a significant role in both central nervous system development and neuronal viability in the adult brain. There are many studies supporting that view. Expression and localization of both FGF and FGF receptors in neuronal cells and biological effects of FGF on the same cells suggest a paracrine or autocrine action. Both acidic and basic FGFs are abundant in the central nervous system (CNS). They have been purified from bovine brain [14, 20, 44, 54] and rat brain [46]. They are present in all or almost all brain neurons [21, 45]. A regional analysis of the expression of bFGF mRNA in brain reveals that it is widely distributed in the cortex (frontal, parietal, and occipital), the hippocampus, hypothalamus and pons [9]. aFGF mRNA has been localized in the developing and adult rat brain using in situ hybridization histochemistry [66]. Prenatally, hybridization to aFGF mRNA was observed throughout the brain, with the strongest signal associated with cells of the developing cortical plate. Postnatally, labeling was localized to specific neuronal populations. In the hippocampus, labeling of the pyramidal cell layer and dentate granule cells was observed and became progres-" sively more intense with maturation. Labeling was also observed in both the external and internal granule cell layers of the developing cerebellum. In addition to FGF localization and expression in brain, a recent report has demonstrated the distribution of an mRNA for aFGF receptor in rat brain using in situ hybridization and an flg clone [64]. The most intense hybridization signal is observed in the hippocampus and in the positive cholinergic neurons. Cerebellar granule cells and spinal cord neurons are also positive for FGF-R mRNA. The biological effects of a- and bFGF on neuronal cells are multiple. Over the last 3 years, several in vitro studies demonstrated that a- and bFGFs can promote the survival of embryonic neurons of various regions of the brain and induce neurite outgrowth in a variety of cells (reviewed in [3]). a- and bFGF support initial survival and subsequent fiber outgrowth of dissociated rodent fetal neurons in culture [36, 61]. bFGF stimulates cell proliferation'[13] neurite outgrowth [60], and cell survival in PC12 cells [51] and neurons from various brain regions of postnatal and fetal rats, including the hippocampus, neocortex, striatum, septum and thalamus [33, 59]. bFGF also displays a potent neurotrophic action on rat cerebral cortical neurons [35] and hippocampal neurons [58], implying that these cells possess FGF receptors [40]. The stimulation of neurite
11
growth is strongly correlated with the amount of 125IbFGF bound by different glycosaminoglycans (heparin, heparan sulfate or hyaluronic acid) [60]. In addition to its role on neurite outgrowth and synaptogenesis aFGF seems to maintain neuronal viability and • connectivity in various experimental brain injuries [1, 16, 30, 42, 52]. In conclusion, this study provides evidence that the presence of bFGF receptors can be correlated with brain developmental events such as proliferation, migration, differentiation and synaptogenesis. It would be interesting to determine the type, number, affinity, level of expression of such receptors in degenerative or survival processes. Work in this area is in progress.
Acknowledgments The authors thank Dr D Mc Devitt for his review of the manuscript, Dr C Sotelo for his scientific advice about cerebellum experiments, Dr JL Gu~net for providing mutant mice, L Jonet for excellent technical assistance, H Coet for the photographs and N Breugnot and Y Maville for typing.
References 1 Anderson KJ, Dam D, Lee S, Cotman CW (1988) Basic fibroblast growth factor prevents death of lesioned cholinergic neurons in vivo. Nature 332, 360-361 2 Baird A, Waiicke PA (1989) Fibroblast growth factor. Br Med Bull 45,438-452 3 Barde YA (1989) Trophic factors and neuronal survival. Neuron 2, 1525-1534 4 Besnard F, Perraud F, Sensenbrenner M, Labourdette G (1989) Effects of acidic and basic fibroblast growth factors on proliferation and maturation of cultured rat oligodendrocytes. Int J Dev Neurosci 7, 401-409 5 Bikfalvi A, Dupuy E, Inyang AL, Fayein N, Leseche G, Courtois Y, Tobelem G (1989) Binding, internalization and degradation of basic fibroblast growth factor in human microvascular endothelial cells. Exp Cell Res 181, 75-84 6 Cirillo A, Arruti C, Courtois Y, Jeanny JC (1990) Localization of basic fibroblast growth factor binding sites in the chick embryonic neural retina. Differentiation 45, 161-167 7 Durand J, Karpousas I, Salvodelli M, Keller N, Pouliquen Y (1990) Corneal stroma quantimetric analysis: Data acquisition optimization. Cybernetica 2, 107-114 8 Eccleston PA, Silberberg DH (1985) Fibroblast growth factor is a mitogen for oligodendrocytes in vitro. Brain Res 353, 315-318 9 Emoto N, Gonzalez AM, Walicke PA, Wada E, Simmons DM, Shimasaki S, Baird A (1989) Basic fibroblast growth factor (FGF) in the central nervous system: identification of specific loci of basic FGF expression in the rat brain. Growth Factors 2, 21-29 10 Fayein NA, Courtois Y, Jeanny JC (1990) Ontogeny of basic fibroblast growth factor binding sites in mouse ocular tissues. Exp Cell Res 188, 75-88 11 Feige J J, Baird A (1988) Glycosylation of the basic fibroblast growth factor receptor. The contribution of carbohydrate to receptor function. J B i o l Chem 263, 14023-14029 12 Gaarskjaer FB (1985) The development of the dendrate area and the hippocampal mossy fiber projection of the rat. J Comp Neurol 241, 154-170 13 Gensburger C, Labourdette G, Sensenbrenner M (1987) Brain basic fibroblast growth factor stimulates the proliferation of rat neuronal precursor cells in vitro. FEBS Lett 217, I-5 14 Gospodarowicz D, Cheng J, Lui GM, Baird A, Bollen P (1984) Isolation of brain fibroblast growth factor by heparine-sepharose affinity chromatography: identity with
12
15
16
17
18 19
20
21 22
23
24 25 26 27
28
29
30 31 32 33 34
NA Fayein et al pituitary fibroblast growth factor. Proc Natl A cad Sci USA 81, 6963-6967 Gospodarowicz D, Neufeld G, Schweigerer L (1986) Molecular and biological characterization of fibroblast growth factor, an angiogenic factor which also controls the proliferation and differentiation of mesoderm and neuroectoderm derived cells. Cell Differ 19, 1-17 Hefti F, Hartikka J, Kniisel B (1989) Function of neurotrophic factors in the adult and aging brain and their possible use in the treatment of neurodegenerative diseases. Neurobiol Aging 10, 515-533 Heuer JG, Von Bartheld CS, Kinoshita Y, Evers PC, Bothwell M (1990) Alternating phases of FGF receptor and NGF receptor expression in the developing chicken nervous system. Neuron 5, 283-296 Hicks D, Courtois Y (1988) Acidic fibroblast growth factor stimulates opsin levels in retinal photoreceptor cells in vitro. FEBS Lett 234, 475-479 Houssaint E, Blanquet PR, Champion-Arnaud P, Gesnei MC, Torriglia A, Courtois Y, Breathnach R (1990) Related fibroblast growth factor receptor genes exist in the human genome. Proc Natl Acad Sci USA 87, 8180-8184 Jacquot Dourges MA, Gulino D, Courty J, Badet J, Barritault D, Jozefonvicz J (1990) Affinity chromatography of fibroblast growth factors on substituted polystyrene. J Chromatogr 526, g5-45 Janet T, Miehe M, Pettrnann B, Labourdette G, Sensenbrenner M (1987) Ultrastructural localization of fibroblast growth factor in neurons of rat brain. Neurosci Lett 80, 153-157 Jeanny JC, Fayein N, Moenner M, Chevallier B, Barritault D, Courtois Y (1987) Specific f'Lxationof bovine brain and retinal acidic and basic fibroblast growth factors to mouse embryortic eye basement membranes. Exp Cell Res 171, 63-75 Kiefer MC, Stephans JC, Crawford K, Onico K, Barr PJ (1990) Ligand-affinity cloning and structure of a cell surface heparan sulfate proteoglycan that binds basic fibroblast growth factor. Proc Natl A cad Sci USA 87, 6985- 6989 Kniss DA, Burry RW (1988) Serum and fibroblast growth factor stimulate quiescent astrocytes to re-enter the cell cycle. Brain Res 439, 281-288 Kornbluth S, Paulson KE, Hanafusa H (1988) Novel tyrosine kinase identified by phosphotyrosine antibody screening of cDNA libraries. Mol Cell Biol 8, 5541-5544 Laemmli VH (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685 Ledoux D, Mereau A, Pieri I, Barritault D, Courty J (1991) High affinity receptors to acidic and basic fibroblast growth factor (FGF) are detected mainly in adult brain membrane preparations but not in liver, kidney, intestine, lung or stomach. Growth Factors 5, 221-231 Lee PL, Johnson DE, Cousens LS, Fried VA, Williams LT (1989) Purification and complementary DNA cloning of a receptor for basic fibroblast growth factor. Science, 245, 57-60 Lipton SA, Wagner JA, Madison RD, D'Amore PA (1988) Acidic fibroblast growth factor enhances regeneration of processes by postnatal .mammalian retinal ganglion cells in culture. Proc Natl Acad Sci USA 85, 2388-2392 Logan A (1988) Elevation of acidic fibroblast growth factor mRNA in lesioned rat brain. Mol Cell Endocrinol 58, 275-278 Mascarelli F, Rauials D, Courtois Y (1989) Fibrohlast growth factor phosphorylation and receptors in rod outer segments. EMBO J 8, 2265-2273 Mascarelli F, Tassin J, Courtois Y (1991) Effect of FGFs on adult bovine Muller cells: proliferation, binding and internalization. Growth Factors 4, 81-95 Matsuda S, Salto H, Nishiyama N (1990) Effect of basic fibroblast growth factor on neurons cultured from various regions of postnatal rat brain. Brain Res 520, 310-316 Mereau A, Pieri I, Gamby C, Courty J, Barritault D (1989) Purification of basic fibroblast growth factor receptors from bovine brain. Biochimie 71, 865-871
35 Morrison RS, Sharma A, de Vellis J, Bradshaw A 0986) Basic fibroblast growth factor supports the survival of cerebral cortical neurons in primary culture. Proc NatlAcad Sci USA 83, 7537-7541 36 Morrison RS, Keating RF, Moskal JR (1988) Basic fibroblast growth factor and epidermal growth factor exert differential trophic effects on CNS neurons. JNeurosci Res 21, 71-79 37 Mullen R J, Eicher EM, Sidman RL (1976) Purkinje cell degeneration, a new neurological mutation in the mouse. Proc Natl Acad Sci USA 73,208-212 38 Murphy M, Drago J, Bartlett PF (1990) Fibroblast growth factor stimulates the proliferation and differentiation of neural precursor cells in vitro. J Neurosci Res 25, 463-475 39 Neufeld G, Godpodarowicz D (1985) The identification and partial characterization of the fibroblast growth factor receptor of baby hamster kidney cells. J Biol Chem 260, 13860-13868 40 Neufeld G, Gospodarowicz D, Dodge L, Fujii DK (1987) Heparin modulation of the neurotrophic effects of acidic and basic fibroblast growth factors and nerve growth factor on PC12 cells. J Cell Physiol 131, 131-140 41 Ornitz DM, Yayon A, Flanagan JG, Svahn CM, Levi E, Leder P (1992) Heparin is required for cell-free binding of basic fibroblast growth factor to a soluble receptor and for mitogenesis in whole cells. Mol Cell Biol 12, 240-247 42 Otto D, Unsicker K (1990) Basic FGF reverses chemical and morphological deficits in the nigrostriatal system of MPTPtreated mice. J Neurosci 10, 1912-1921 43 Partanen J, Vainikka S, Korhnen J, Armstrong E, Alitalo K (1992) Diverse receptors for fibroblast growth factors. Prog Growth Factor Res 4, 69-83 44 Pettmann B, Weibel M, Sensenbrenner M, Labourdette G (1985) Purification of two astroglial growth factors from bovine brain. FEBS Lett 189, 102-108 45 Pettmann B, Labourdette G, Weibel M, Sensenbrenner M (1986) The brain fibroblast growth factor (FGF) is localized in neurons. Neurosci Lett 68, 175-180 46 Presta M, Rusnati M, Maler JAM, Ragnotti G (1988) Purification of basic fibroblast growth factor from rat brain: identification of a M r 22000 immunoreactive form. Biochem Biophys Res Commun 155, 1161-1172 47 Rakic P (1986) Specification of cerebral cortical areas. Science 241, 170-176 48 Reid HH, Wilks AF, Bernard O (1990) Two forms of the basic fibroblast growth factor receptorqike mRNA are expressed in the developing mouse brain. Proc Natl Acad Sci USA 87, 1596-1600 49 Ruoslahti E, Yamaguchi Y (1991) Proteogiycans as modulators of growth factor activities. Cell 64, 867-869 50 Saunders S, Jalkanen M, O'Farrell S, Bernfield M (1989) Syndecan, an integral membrane proteoglycan. J Cell Biol 108, 1547-1556 51 Schubert D, Ling N, Baird A (1987) Multiple influences of a heparin-binding growth factor on neuronal development. J Cell Biol 104, 635-643 52 Sievers J, Hausmann B, Unsicker K, Berry M (1987) Fibroblast growth factors promote the survival of adult rat retinal ganglion cells after transection of the optic nerve. Neurosci Lett 76, 157-162 53 Soubrane G, Jerdan J, Karpouzas I, Fayein NA, Glaser B, Coscas G, Courtois Y, Jeanny JC (1990) Binding of basic fibroblast growth factor to normal and neovascularized rabbit cornea. Invest Ophthalmol Vis Sci 31, 323-333 54 Thomas KA, Rios Candelore M, Fitzpatrick S (1984) Purification and characterization of acidic fibroblast growth factor from bovine brain. ProcNatlAcadSci USA 81,357-361 55 Uhlrich S, Lagente O, Lenfant M, Courtois Y (1986) Effect of heparin on the stimulation of non-vascular cells by human acidic and basic FGF. Biochem Biophys Res Commun 137, 1205-1213 56 Unsicker K, Reichert-Preibsch H, Schmidt R, Pettmann B, Labourdette G, Sensenbrenner M (1987) Astroglial and fibroblast growth factors have neurotrophic functions for
bFGF binding in mouse brain cultured peripheral and central nervous system neurons. Proc Natl Acad Sci USA 84, 5459-5463
57 Vigny M, OUier-Hartmann MP, Lavigne M, Fayein N, Jeanny JC, Laurent M, Courtois Y (1988) Specific binding of basic fibroblast growth factor to basic fibroblast growth factor to basement membrane-like structures and to purified heparan sulfate proteoglycan of the EHS tumor. J Cell Physiol 137, 321-328 58 Walicke P, Cowan WM, Ueno N, Baird A, Guillemin R (1986) Fibroblast growth factor promotes survival of dissociated hippocampal neurons and enhances neurite extension. Proc Natl Acad Sci USA 83, 3012-3016 59 Walicke PA (1988) Basic and acidic fihroblast growth factors have trophic effects on neurons from multiple CNS regions. J Neurosci 8, 2618-2627 60 Walicke PA (1988) Interactions between basic fibroblast growth factor (FGF) and glycosaminoglycans in promoting neurite outgrowth. Exp Neurol 102, 144-148 61 Walicke PA, Baird A (1988) Neurotrophic effects of basic and acidic fibroblast growth factors are not mediated through glial cells. Dev Brain Res 40, 71-79
13
62 Walicke PA, Feige J J, Baird A (1989) Characterization of the neuronal receptor for basic fibroblast growth factor and comparison to receptors on mesenchymal cells. JBiol Chem 264, 4120-4126 63 Walsh C, Cepko CL (1988) Clonally related cortical cells show several migration patterns. Science 241, 1342-1345 • 64 Wanaka A, Johnson EM, Milbrandt J (1990) Localization of FGF receptor mRNA in the adult central nervous system by in situ hybridization. Neuron 5, 267-281 65 Wassef M, Sotelo C, Cholley B, Brehier A, Thomasset M 0987) Cerebellar mutations affecting the postnatal survival of Purkinje cells in the mouse disclose a longitudinal pattern of differentially sensitive cells. Dev Biol 124, 379-389 66 Wilcox BJ, Unnerstall JR (1991) Expression of acidic fibroblast growth factor mRNA in the developing and adult rat brain. Neuron 6, 397-409 67 Yayon A, Klagsbrun M, Esko JD, Leder P, Ornitz DM (1991) Cell surface, heparin-like molecules are required for binding of basic fibroblast growth factor to its high affinity receptor. Cell 64, 841-848