- Email: [email protected]
The binding of basic fibroblast growth factor to Alzheimer's neurofibrillary tangles and senile plaques
Neuroscience Letters, 122 (1991) 33-36 Elsevier Scientific Publishers Ireland Ltd.
33
NSL 07455
The binding of basic fibroblast growth factor to Alzheimer's neurofibrillary tangles and senile plaques Takeo K a t o I , Hajime Sasaki 3, Tadashi Katagiri 1, Hideo Sasaki 1, Kazunori Koiwai a, Hitomi Youki 3, Shiro Totsuka 2 and Tsuyoshi Ishii# 1Third Department of Internal Medicine, ZDepartment of Psychiatry, Yamagata University School of Medicine, Yamagata (Japan), 3Bio.Science Laboratory, Inc., Yamagata (Japan) and 4psychiatric Research Institute of Tokyo, Tokyo (Japan) (Received 21 August 1990; Revised version received 25 August 1990; Accepted 21 September 1990)
Key words: Alzheimer's disease; Basic fibroblast growth factor; Acidic fibroblast growth factor; Heparan sulfate; Proteoglycan; fl-Protein; A4; Basic fibroblast growth factor-binding site Brain sections from Alzheimer's disease (AD) patients and controls were treated with basic fibroblast growth factor (bFGF) and then immunostained with anti-bFGF. Additional sections were treated with biotinylated bFGF without using the anti-bFGF. Labelling was visualized by the ABC method. Both protocols above intensely labelled neurofibrillary tangles, senile plaques and amyloidotic vessels in AD brains. Omission of the bFGF treatment abolished the staining of the AD lesions. The pretreatment of sections with heparitinase also reduced their staining. These results indicate that AD lesions contain bFGF-binding sites and that the chemical substrate for bFGF binding to AD lesions was heparan sulfate.
Alzheimer's disease (AD) is one of the most common and important disorders which cause dementia in the elderly and aged. The disease is neuropathologically characterized by severe loss of neurons associated with numerous Alzheimer's neurofibrillary tangles (NFTs) and senile plaques (SPs) in the cerebral cortex, hippocampus, nucleus basalis of Meynert and other areas of the brain. Since NFT and SP may be clues to the pathogenesis and etiology of AD, many recent studies on AD have focused on ultrastructural and biochemical aspects of these lesions. Although these pathological hallmarks of AD brains have generally been regarded as resulting from 'degeneration' of nerve cells and their neurites in the affected brain, it is also possible that accumulation of proteins such as phosphorylated tau and fl-protein, which seem to be major components of NFTs and SP cores, respectively, may occur under active 'regenerative' conditions. In fact, there has been evidence suggesting that a regenerative process occurs in AD brains: abormal sprouting of neurites from nerve cells was observed histologically in AD brains [4]; and a markedly high neurotrophic activity of extracts from AD brain homogenates was found in vitro [10]. These findings could be Correspondence: T. Kato, Third Department of Internal Medicine, Yamagata Univ. School of Medicine, 2-2-2 Iida-Nishi, Yamagata 99023, Japan. 0304- 3940/91/$ 03.50 © 1991 Elsevier Scientific Publishers Ireland Ltd.
explained by either an increased level of neurotrophic substance(s) or a decreased level of its (their) inhibitory factor(s) in AD brains [11], neither of which has yet been identified. In this paper, we report that the pathological hallmarks of AD brains (NFTs and SPs) contain abundant binding sites for basic fibroblast growth factor (bFGF), which is a polypeptide of 146 amino acid residues with a potent trophic effect on neurons and astroglia as well as mesoderm-derived cells such as fibroblasts, chondrocytes, myocytes, and endothelial cells, and that these bFGF-binding sites in AD lesions may be heparan sulfate. Anti-bFGF antiserum was raised in rabbits injected subcutaneously with bFGF (bovine recombinant, Amersham), and the antiserum obtained was purified by affinity chromatography on immobilized bFGF. Anti-acidic FGF(aFGF) was also raised and purified in a similar way. Both enzyme immunoassay and Western blot studies showed that anti-bFGF and anti-aFGF specifically recognized bFGF and aFGF, respectively, and did not cross-react with each other (Fig. 1). Formalin-fixed, paraffin-embedded sections of the hippocampus and temporal lobes from 3 cases of AD and from two age-matched controls were used in this study. As listed in Table I, we designed 10 experimental protocols (nos. 1-10) on tissue sections using bFGF,
34 silver stain
anti-aFGF A
anti-bFGF
*/' , ~ , /
biotin
C
B
92k
anti-bFGF
66k
bFGF-binding site
7'
," , , ~ i ,
45k bFGF
I
!
biotin
anti-bFGF
31k L--J
2 ~lk
!
14k
/
O
-
I) \
O
0
bFGF-binding site
bFGF
\ a
markers
aFGF
b
bFGF
a
b
a
b
aFGF
bFGF
aFGF
bFGF
Fig. !. Immunobiotting of anti-aFGF (B) or anti-bFGF (C) on aFGF (a) and bFGF (b), showing that anti-aFGF and anti-bFGF label aFGF (B-a) and bFGF (C-b), respectively,and do not cross-react with each other (B-b, C-a). I00 ng of aFGF (a), bFGF (b), or marker proteins (left) in SDS sample buffer were electrophoresed on 8-16 %gradient polyacrylamide gels. A: silver stain. B,C: after transfer to nitrocellulose, aFGF and bFGF were immunostained with anti-aFGF (B) or anti-bFGF (C).
a F G F , a n t i - b F G F , a n d a n t i - a F G F . A m o n g these, only protocol no. 3 (Table I, Fig. 2) intensely labelled N F T s , SPs, a n d a m y l o i d deposits o f vessel walls (amyloidotic vessels; AVs) (Fig. 3A,C). I n this protocol, tissue sec-
TABLE I LABELLING CHARACTERISTICS OF AD LESIONS Ten experimental protocols (1-10) were designed by using bFGF, aFGF, anti-bFGF and/or anti-aFGF. Among these, only protocol no. 3 labelled NFTs, SPs and AVs in AD tissue section: the section was incubated with 2.5 #g/ml bFGF (Amersham) in Tris-buffered saline (TBS, 50 mM Tris and 150 mM NaC1, pH 7.6), washed with TBS, and reacted with affinity-purified anti-bFGF. The bound antibody was visualizedby the ABC method (Vectastain, Vector). The other 9 protocols did not label any AD lesions. +, labelled; - , not labelled; bFGF, basic fibroblast growth factor; aFGF, acidic fibroblast growth factor; NRS, normal rabbit serilm; NFT, neurofibrillary tangle; SP, senile plaque; AV, amyloidotic vessel. Experimental protocols 1
2 3 4 5 6 7 8 9 l0
anti-bFGF anti-aFGF bFGF ---,anti-bFGF bFGF ---,anti-aFGF aFGF --, anti-bFGF aFGF ~ anti-aFGF NRS -, anti-bFGF NRS ~ anti-aFOF bFGF ~ NRS aFGF ~ NRS
NFT
SP
AV
+ -
+ -
+ -
•~
biotin
~" bFGF-binding site
Fig. 2. Schematic drawings of 3 experimental protocols. 1: protocol 1 in Table I (see Fig. 3B) 2: protocol 3 in Table I (see Fig. 3A,C) 3: protocol using biotinylated bFGF (see Fig. 3D).
tions were p r e - i n c u b a t e d with b F G F (2.5/~g/ml) overnight at r o o m temperature (RT), washed with 3 changes of Tris-buffered saline (TBS, p H 7.6) for 15-60 min, a n d then i m m u n o s t a i n e d with a n t i - b F G F overnight at R T (Fig. 2). The labelling was visualized by the a v i d i n - b i o t i n - p e r o x i d a s e complex (ABC, Vector) method. The other 9 protocols did not label any o f these pathological structures (Table I). The most plausible e x p l a n a t i o n for these findings is that b F G F itself binds to N F T s , SPs and AVs. T o test this interpretation, we employed several different c o n c e n t r a t i o n s of b F G F (2.5 pg/ml, 1.0 /tg/ml, 0.1 /tg/ml a n d 0.01 Hg/ml) in the p r e - i n c u b a t i o n solution. The resultant staining-intensity progressively decreased with the d i l u t i o n of b F G F , a n d the labelling was completely abolished by omission of b F G F (protocol no. 1 in Table I) (Fig. 3B). In order to o b t a i n direct evidence that b F G F binds to the pathological hallmarks of A D brains, b F G F was labelled with biotin ( N H S - L C biotin, Pierce) a n d it was applied to the tissue sections. The biotinylated b F G F b o u n d to the tissue sections was visualized by the A B C method. As expected, the N F T s , SPs a n d AVs were clearly stained by this procedure (Fig. 3D). F u r t h e r m o r e , the staining intensity was m a r k e d l y reduced by a d d i n g excess unlabelled b F G F into the biotinylated b F G F solution, indicating that c o m p e t i t i o n occurred between biotinylated a n d unlabelled b F G F for the b i n d i n g sites in A D b r a i n sections. As a first step in elucidating what is the chemical substrate for b F G F b i n d i n g to A D lesions, we incubated, overnight at 4°C, a solution c o n t a i n i n g 2.5 pg o f b F G F a n d 1 mg of one of the following synthesized fragments of fl-protein (A4) (residues 1-8, 6-12, a n d 10-26 from the N - t e r m i n u s offl-protein) in 1 ml of TBS with 1% nor-
35
~m
gg
F
Fig. 3. A: binding of basic fibroblast growth factor (bF(3F) to neurofibrillary tangles (NFTs) and senile plaques (SPs) in the hippocampus of an Alzheimer's disease (AD) brain. The arrow indicates the same vessel as in B. Protocol 3 (see Table I and Fig. 2-2) was performed on AD tissue section. B: an adjacent section to A, showing that omission of bFGF abolished the staining of NFTs and SPs. The section was incubated with the affinity-purified anti-bFGF and visualized by the ABC method (see protocol 1 in Table I and Fig. 2-1). The arrows in A and B point to the same vessel. C: cerebrovascular amyloid (arrow) as well as NFTs and SPs were labelled with bFGF (protocol 3). D: NFTs and SPs labelled with biotinylated bFGF. Biotinylation of bFGF was accomplished by using N-hydroxysuccinimidyl-6-(biotinamido)hexanoate (Pierce). The section was incubated with biotinylated bFGF, which was visualized by the ABC method. E: heparinase treatment of AD tissue section before protocol 3 did not affect the binding of bFGF to NFTs and SPs. The section was incubated for 24h at 37°C with 5 or 10 U/ml heparinase (Sigma) in 10 mM Tris and 10 -3 M CaC12. F: heparitinase treatment before protocol 3 of AD tissue section adjacent to E reduced bFGF binding to NFTs and SPs. The section was incubated for 24h at 37°C with 1 U/ml, heparitinase (Seikagaku Kogyo) in 10 mM Tris and 10 -3 M CaC12. (A-D: x 104, E,F: x 52.) (A,B: AD hippocampus; C-F: AD temporal cortex.)
mal goat serum (Vector), and applied each mixture to tissue sections, since the distribution of bFGF-binding sites was somewhat similar to that offl-protein [1, 3]. The staining intensity was not affected by these procedures, which suggests that the substrate for bFGF binding to AD lesions is unlikely to be fl-protein. Next, we incubated AD brain sections with 2 % periodic acid, a glycolytic agent [2], and then performed protocol no. 3, because saccharides have been reported to be associated with NFTs and SPs [5]. The resultant staining intensity was dramatically reduced or abolished, suggesting that the bFGF binding requires sugars incorporated into AD lesions. Since bFGF is known to have an affinity for heparin and heparan sulfate [7], AD brain sections were pretreated with heparinase or heparitinase, which degrades heparin or heparan sulfate, respectively, before protocol no. 3. While the pretreatment of the sections with heparinase did not affect the binding of b F G F to AD lesions (Fig. 3E), the heparitinase pretreatment
markedly reduced its binding (Fig. 3F). The results suggest that heparan sulfate is a component of NFT, SP and AV, and that it is the chemical substrate for bFGF binding to these AD lesions. In this context, of particular interest is that protease nexin I (PN I), a serine protease inhibitor, has been reported to bind to NFTs and SPs [6]. Since PN I has a binding affinity for heparin as bFGF does, it may be possible that PN I binds to saccharides incorporated into AD lesions. In 1987, Snow et al. suggested the presence of sulfated glycosaminoglycans in NFTs, SPs and AVs by using the sulfated Alcian blue and Alcian blue-MgCl 2 techniques [9]. In 1988, they reported an immunohistochemical study in which both monoclonal and polyclonal antibodies to the protein core of a basement membrane-derived heparan sulfate proteoglycan labelled SPs and AVs, but not NFTs [8]. Although they could not specify the type of sulfated glycosaminoglycans in AD lesions in
36
their 1987 study and the antibodies used in 1988 did not recognize heparan sulfate itself but the core protein of the proteoglycan from a basement membrane, their studies suggested the presence of heparan sulfate in AD lesions. In this study, the presence ofheparan sulfate in NFTs, SPs and AVs was demonstrated on the basis of the bFGF binding to these lesions and its enzymatic inhibition by heparitinase. The close association of heparan sulfate with the AD lesions may implicate its involvement in their pathogenesis and/or development. The authors thank Dr. David Allsop for his helpful suggestions during the preparation of this manuscript. 1 Allsop, D., Haga, S., Bruton, C., Ishii, T. and Roberts, G.W., Neurofibrillary tangles in some cases of dementia pugilistica share antigens with amyloid ,&protein of Alzheimer's disease, Am. J. Pathol., 136 (1990) 255-260. 2 Behrouz, N., Defossez, A., Delacourte, A., Hublau, P. and Mazzuca, M., Alzheimer's disease: glycolytic pretreatment dramatically enhances immunolabeling of senile plaques and cerebrovascular amyloid substance, Lab. Invest., 61 (1989) 576-583. 3 Hyman, B.T., Van Hoesen, G.W., Beyreuther, K. and Masters, C.L., A4 amyloid protein immunoreactivity is present in Alz-
4 5
6
7
8
9 10
11
heimer's disease neurofibrillary tangles, Neurosci. Lett., 101 (1989) 352-355. Ihara, Y., Massive somatodendritic sprouting of cortical neurons in Alzheimer's disease, Brain Res., 459 (1988) 138-144. Mann, D.M.A., Bonshek, R.E., Marcyniuk, B., Stoddart, R.W. and Torgerson, E., Saccharides of senile plaques and neurofibrillary tangles in Alzbeimer's disease, Neurosci. Lett., 85 (1988) 277 282. Rosenblatt, D.E., Guela, C. and Mesulam, M.-M., Protease nexin I immunostaining in Alzheimer's disease, Ann. Neurol., 26 (1989) 6284~34. Saksela, O., Moscatelli, D., Sommer, A. and Rifkin, D.B., Endothelial cell-derived heparan sulfate binds basic fibroblast growth factor and protects it from proteolytic degradation, J. Cell Biol., 107 (1988) 743-751. Snow, A.D., Mar, H., Nochlin, D., Kimata, K., Kato, M., Suzuki, S., Hassell, J. and Wight, T.N., The presence of beparan sulfate proteoglycans in the neuritic plaques and congophilic angiopathy in Alzheimer's disease, Am. J. Pathol., 133 (1988) 456-463. Snow, A.D., Willmer, J.P. and Kisilevsky, R., Sulfated glycosaminoglycans in Alzheimer's disease, Hum. Pathol., 18 (1987) 506-510. Uchida, Y., Ihara, Y. and Tomonaga, M., Alzheimer's disease brain extract stimulates the survival of cerebral cortical neurons from neonatal rats, Biochem. Biophys. Res. Commun., 150 (1988) 1263-1267. Uchida, Y. and Tomonaga, M., Neurotrophic action of Alzheimer's disease brain extract is due to the loss of inhibitory factors for survival and neurite formation of cerebral cortical neurons, Brain Res., 481 (1989) 190-193.
33
NSL 07455
The binding of basic fibroblast growth factor to Alzheimer's neurofibrillary tangles and senile plaques Takeo K a t o I , Hajime Sasaki 3, Tadashi Katagiri 1, Hideo Sasaki 1, Kazunori Koiwai a, Hitomi Youki 3, Shiro Totsuka 2 and Tsuyoshi Ishii# 1Third Department of Internal Medicine, ZDepartment of Psychiatry, Yamagata University School of Medicine, Yamagata (Japan), 3Bio.Science Laboratory, Inc., Yamagata (Japan) and 4psychiatric Research Institute of Tokyo, Tokyo (Japan) (Received 21 August 1990; Revised version received 25 August 1990; Accepted 21 September 1990)
Key words: Alzheimer's disease; Basic fibroblast growth factor; Acidic fibroblast growth factor; Heparan sulfate; Proteoglycan; fl-Protein; A4; Basic fibroblast growth factor-binding site Brain sections from Alzheimer's disease (AD) patients and controls were treated with basic fibroblast growth factor (bFGF) and then immunostained with anti-bFGF. Additional sections were treated with biotinylated bFGF without using the anti-bFGF. Labelling was visualized by the ABC method. Both protocols above intensely labelled neurofibrillary tangles, senile plaques and amyloidotic vessels in AD brains. Omission of the bFGF treatment abolished the staining of the AD lesions. The pretreatment of sections with heparitinase also reduced their staining. These results indicate that AD lesions contain bFGF-binding sites and that the chemical substrate for bFGF binding to AD lesions was heparan sulfate.
Alzheimer's disease (AD) is one of the most common and important disorders which cause dementia in the elderly and aged. The disease is neuropathologically characterized by severe loss of neurons associated with numerous Alzheimer's neurofibrillary tangles (NFTs) and senile plaques (SPs) in the cerebral cortex, hippocampus, nucleus basalis of Meynert and other areas of the brain. Since NFT and SP may be clues to the pathogenesis and etiology of AD, many recent studies on AD have focused on ultrastructural and biochemical aspects of these lesions. Although these pathological hallmarks of AD brains have generally been regarded as resulting from 'degeneration' of nerve cells and their neurites in the affected brain, it is also possible that accumulation of proteins such as phosphorylated tau and fl-protein, which seem to be major components of NFTs and SP cores, respectively, may occur under active 'regenerative' conditions. In fact, there has been evidence suggesting that a regenerative process occurs in AD brains: abormal sprouting of neurites from nerve cells was observed histologically in AD brains [4]; and a markedly high neurotrophic activity of extracts from AD brain homogenates was found in vitro [10]. These findings could be Correspondence: T. Kato, Third Department of Internal Medicine, Yamagata Univ. School of Medicine, 2-2-2 Iida-Nishi, Yamagata 99023, Japan. 0304- 3940/91/$ 03.50 © 1991 Elsevier Scientific Publishers Ireland Ltd.
explained by either an increased level of neurotrophic substance(s) or a decreased level of its (their) inhibitory factor(s) in AD brains [11], neither of which has yet been identified. In this paper, we report that the pathological hallmarks of AD brains (NFTs and SPs) contain abundant binding sites for basic fibroblast growth factor (bFGF), which is a polypeptide of 146 amino acid residues with a potent trophic effect on neurons and astroglia as well as mesoderm-derived cells such as fibroblasts, chondrocytes, myocytes, and endothelial cells, and that these bFGF-binding sites in AD lesions may be heparan sulfate. Anti-bFGF antiserum was raised in rabbits injected subcutaneously with bFGF (bovine recombinant, Amersham), and the antiserum obtained was purified by affinity chromatography on immobilized bFGF. Anti-acidic FGF(aFGF) was also raised and purified in a similar way. Both enzyme immunoassay and Western blot studies showed that anti-bFGF and anti-aFGF specifically recognized bFGF and aFGF, respectively, and did not cross-react with each other (Fig. 1). Formalin-fixed, paraffin-embedded sections of the hippocampus and temporal lobes from 3 cases of AD and from two age-matched controls were used in this study. As listed in Table I, we designed 10 experimental protocols (nos. 1-10) on tissue sections using bFGF,
34 silver stain
anti-aFGF A
anti-bFGF
*/' , ~ , /
biotin
C
B
92k
anti-bFGF
66k
bFGF-binding site
7'
," , , ~ i ,
45k bFGF
I
!
biotin
anti-bFGF
31k L--J
2 ~lk
!
14k
/
O
-
I) \
O
0
bFGF-binding site
bFGF
\ a
markers
aFGF
b
bFGF
a
b
a
b
aFGF
bFGF
aFGF
bFGF
Fig. !. Immunobiotting of anti-aFGF (B) or anti-bFGF (C) on aFGF (a) and bFGF (b), showing that anti-aFGF and anti-bFGF label aFGF (B-a) and bFGF (C-b), respectively,and do not cross-react with each other (B-b, C-a). I00 ng of aFGF (a), bFGF (b), or marker proteins (left) in SDS sample buffer were electrophoresed on 8-16 %gradient polyacrylamide gels. A: silver stain. B,C: after transfer to nitrocellulose, aFGF and bFGF were immunostained with anti-aFGF (B) or anti-bFGF (C).
a F G F , a n t i - b F G F , a n d a n t i - a F G F . A m o n g these, only protocol no. 3 (Table I, Fig. 2) intensely labelled N F T s , SPs, a n d a m y l o i d deposits o f vessel walls (amyloidotic vessels; AVs) (Fig. 3A,C). I n this protocol, tissue sec-
TABLE I LABELLING CHARACTERISTICS OF AD LESIONS Ten experimental protocols (1-10) were designed by using bFGF, aFGF, anti-bFGF and/or anti-aFGF. Among these, only protocol no. 3 labelled NFTs, SPs and AVs in AD tissue section: the section was incubated with 2.5 #g/ml bFGF (Amersham) in Tris-buffered saline (TBS, 50 mM Tris and 150 mM NaC1, pH 7.6), washed with TBS, and reacted with affinity-purified anti-bFGF. The bound antibody was visualizedby the ABC method (Vectastain, Vector). The other 9 protocols did not label any AD lesions. +, labelled; - , not labelled; bFGF, basic fibroblast growth factor; aFGF, acidic fibroblast growth factor; NRS, normal rabbit serilm; NFT, neurofibrillary tangle; SP, senile plaque; AV, amyloidotic vessel. Experimental protocols 1
2 3 4 5 6 7 8 9 l0
anti-bFGF anti-aFGF bFGF ---,anti-bFGF bFGF ---,anti-aFGF aFGF --, anti-bFGF aFGF ~ anti-aFGF NRS -, anti-bFGF NRS ~ anti-aFOF bFGF ~ NRS aFGF ~ NRS
NFT
SP
AV
+ -
+ -
+ -
•~
biotin
~" bFGF-binding site
Fig. 2. Schematic drawings of 3 experimental protocols. 1: protocol 1 in Table I (see Fig. 3B) 2: protocol 3 in Table I (see Fig. 3A,C) 3: protocol using biotinylated bFGF (see Fig. 3D).
tions were p r e - i n c u b a t e d with b F G F (2.5/~g/ml) overnight at r o o m temperature (RT), washed with 3 changes of Tris-buffered saline (TBS, p H 7.6) for 15-60 min, a n d then i m m u n o s t a i n e d with a n t i - b F G F overnight at R T (Fig. 2). The labelling was visualized by the a v i d i n - b i o t i n - p e r o x i d a s e complex (ABC, Vector) method. The other 9 protocols did not label any o f these pathological structures (Table I). The most plausible e x p l a n a t i o n for these findings is that b F G F itself binds to N F T s , SPs and AVs. T o test this interpretation, we employed several different c o n c e n t r a t i o n s of b F G F (2.5 pg/ml, 1.0 /tg/ml, 0.1 /tg/ml a n d 0.01 Hg/ml) in the p r e - i n c u b a t i o n solution. The resultant staining-intensity progressively decreased with the d i l u t i o n of b F G F , a n d the labelling was completely abolished by omission of b F G F (protocol no. 1 in Table I) (Fig. 3B). In order to o b t a i n direct evidence that b F G F binds to the pathological hallmarks of A D brains, b F G F was labelled with biotin ( N H S - L C biotin, Pierce) a n d it was applied to the tissue sections. The biotinylated b F G F b o u n d to the tissue sections was visualized by the A B C method. As expected, the N F T s , SPs a n d AVs were clearly stained by this procedure (Fig. 3D). F u r t h e r m o r e , the staining intensity was m a r k e d l y reduced by a d d i n g excess unlabelled b F G F into the biotinylated b F G F solution, indicating that c o m p e t i t i o n occurred between biotinylated a n d unlabelled b F G F for the b i n d i n g sites in A D b r a i n sections. As a first step in elucidating what is the chemical substrate for b F G F b i n d i n g to A D lesions, we incubated, overnight at 4°C, a solution c o n t a i n i n g 2.5 pg o f b F G F a n d 1 mg of one of the following synthesized fragments of fl-protein (A4) (residues 1-8, 6-12, a n d 10-26 from the N - t e r m i n u s offl-protein) in 1 ml of TBS with 1% nor-
35
~m
gg
F
Fig. 3. A: binding of basic fibroblast growth factor (bF(3F) to neurofibrillary tangles (NFTs) and senile plaques (SPs) in the hippocampus of an Alzheimer's disease (AD) brain. The arrow indicates the same vessel as in B. Protocol 3 (see Table I and Fig. 2-2) was performed on AD tissue section. B: an adjacent section to A, showing that omission of bFGF abolished the staining of NFTs and SPs. The section was incubated with the affinity-purified anti-bFGF and visualized by the ABC method (see protocol 1 in Table I and Fig. 2-1). The arrows in A and B point to the same vessel. C: cerebrovascular amyloid (arrow) as well as NFTs and SPs were labelled with bFGF (protocol 3). D: NFTs and SPs labelled with biotinylated bFGF. Biotinylation of bFGF was accomplished by using N-hydroxysuccinimidyl-6-(biotinamido)hexanoate (Pierce). The section was incubated with biotinylated bFGF, which was visualized by the ABC method. E: heparinase treatment of AD tissue section before protocol 3 did not affect the binding of bFGF to NFTs and SPs. The section was incubated for 24h at 37°C with 5 or 10 U/ml heparinase (Sigma) in 10 mM Tris and 10 -3 M CaC12. F: heparitinase treatment before protocol 3 of AD tissue section adjacent to E reduced bFGF binding to NFTs and SPs. The section was incubated for 24h at 37°C with 1 U/ml, heparitinase (Seikagaku Kogyo) in 10 mM Tris and 10 -3 M CaC12. (A-D: x 104, E,F: x 52.) (A,B: AD hippocampus; C-F: AD temporal cortex.)
mal goat serum (Vector), and applied each mixture to tissue sections, since the distribution of bFGF-binding sites was somewhat similar to that offl-protein [1, 3]. The staining intensity was not affected by these procedures, which suggests that the substrate for bFGF binding to AD lesions is unlikely to be fl-protein. Next, we incubated AD brain sections with 2 % periodic acid, a glycolytic agent [2], and then performed protocol no. 3, because saccharides have been reported to be associated with NFTs and SPs [5]. The resultant staining intensity was dramatically reduced or abolished, suggesting that the bFGF binding requires sugars incorporated into AD lesions. Since bFGF is known to have an affinity for heparin and heparan sulfate [7], AD brain sections were pretreated with heparinase or heparitinase, which degrades heparin or heparan sulfate, respectively, before protocol no. 3. While the pretreatment of the sections with heparinase did not affect the binding of b F G F to AD lesions (Fig. 3E), the heparitinase pretreatment
markedly reduced its binding (Fig. 3F). The results suggest that heparan sulfate is a component of NFT, SP and AV, and that it is the chemical substrate for bFGF binding to these AD lesions. In this context, of particular interest is that protease nexin I (PN I), a serine protease inhibitor, has been reported to bind to NFTs and SPs [6]. Since PN I has a binding affinity for heparin as bFGF does, it may be possible that PN I binds to saccharides incorporated into AD lesions. In 1987, Snow et al. suggested the presence of sulfated glycosaminoglycans in NFTs, SPs and AVs by using the sulfated Alcian blue and Alcian blue-MgCl 2 techniques [9]. In 1988, they reported an immunohistochemical study in which both monoclonal and polyclonal antibodies to the protein core of a basement membrane-derived heparan sulfate proteoglycan labelled SPs and AVs, but not NFTs [8]. Although they could not specify the type of sulfated glycosaminoglycans in AD lesions in
36
their 1987 study and the antibodies used in 1988 did not recognize heparan sulfate itself but the core protein of the proteoglycan from a basement membrane, their studies suggested the presence of heparan sulfate in AD lesions. In this study, the presence ofheparan sulfate in NFTs, SPs and AVs was demonstrated on the basis of the bFGF binding to these lesions and its enzymatic inhibition by heparitinase. The close association of heparan sulfate with the AD lesions may implicate its involvement in their pathogenesis and/or development. The authors thank Dr. David Allsop for his helpful suggestions during the preparation of this manuscript. 1 Allsop, D., Haga, S., Bruton, C., Ishii, T. and Roberts, G.W., Neurofibrillary tangles in some cases of dementia pugilistica share antigens with amyloid ,&protein of Alzheimer's disease, Am. J. Pathol., 136 (1990) 255-260. 2 Behrouz, N., Defossez, A., Delacourte, A., Hublau, P. and Mazzuca, M., Alzheimer's disease: glycolytic pretreatment dramatically enhances immunolabeling of senile plaques and cerebrovascular amyloid substance, Lab. Invest., 61 (1989) 576-583. 3 Hyman, B.T., Van Hoesen, G.W., Beyreuther, K. and Masters, C.L., A4 amyloid protein immunoreactivity is present in Alz-
4 5
6
7
8
9 10
11
heimer's disease neurofibrillary tangles, Neurosci. Lett., 101 (1989) 352-355. Ihara, Y., Massive somatodendritic sprouting of cortical neurons in Alzheimer's disease, Brain Res., 459 (1988) 138-144. Mann, D.M.A., Bonshek, R.E., Marcyniuk, B., Stoddart, R.W. and Torgerson, E., Saccharides of senile plaques and neurofibrillary tangles in Alzbeimer's disease, Neurosci. Lett., 85 (1988) 277 282. Rosenblatt, D.E., Guela, C. and Mesulam, M.-M., Protease nexin I immunostaining in Alzheimer's disease, Ann. Neurol., 26 (1989) 6284~34. Saksela, O., Moscatelli, D., Sommer, A. and Rifkin, D.B., Endothelial cell-derived heparan sulfate binds basic fibroblast growth factor and protects it from proteolytic degradation, J. Cell Biol., 107 (1988) 743-751. Snow, A.D., Mar, H., Nochlin, D., Kimata, K., Kato, M., Suzuki, S., Hassell, J. and Wight, T.N., The presence of beparan sulfate proteoglycans in the neuritic plaques and congophilic angiopathy in Alzheimer's disease, Am. J. Pathol., 133 (1988) 456-463. Snow, A.D., Willmer, J.P. and Kisilevsky, R., Sulfated glycosaminoglycans in Alzheimer's disease, Hum. Pathol., 18 (1987) 506-510. Uchida, Y., Ihara, Y. and Tomonaga, M., Alzheimer's disease brain extract stimulates the survival of cerebral cortical neurons from neonatal rats, Biochem. Biophys. Res. Commun., 150 (1988) 1263-1267. Uchida, Y. and Tomonaga, M., Neurotrophic action of Alzheimer's disease brain extract is due to the loss of inhibitory factors for survival and neurite formation of cerebral cortical neurons, Brain Res., 481 (1989) 190-193.