β-Amyloid induces the production of active, matrix-degrading proteases in cultured rat astrocytes

Brain Research 970 (2003) 205–213 www.elsevier.com / locate / brainres

Research report

b-Amyloid induces the production of active, matrix-degrading proteases in cultured rat astrocytes Suman Deb a , J. Wenjun Zhang a , Paul E. Gottschall a,b , * a

Department of Pharmacology and Therapeutics, University of South Florida College of Medicine, Tampa, FL 33612 -4799, USA b Alzheimer’ s Research Laboratory, University of South Florida College of Medicine, Tampa, FL 33612 -4799, USA Accepted 20 December 2002

Abstract The senile and neuritic plaque neuropathology of Alzheimer’s disease (AD) is accompanied by an inflammatory response that includes activated astrocytes and microglia. Activated mononuclear phagocytes and reactive astrocytes, in response to inflammatory cytokines, secrete a set of extracellular matrix (ECM)-degrading enzymes that include the matrix metalloproteinases (MMPs). The major peptide component of senile plaques of AD, b-amyloid (Ab), stimulates the production of several MMPs from cultured rat astrocytes and microglia. The purpose of this study was two-fold: (1) to compare the pattern of MMP induction in rat astrocytes on treatment with ‘soluble’ and ‘fibrillar’ Ab(1–40) and Ab(1–42), and (2) to examine whether treatment of astrocytes with Ab results in degraded fragments of ECM. Ab aggregation differentially affected the production of MMP-2 and MMP-9 in astrocyte cultures. Activation experiments with amino phenyl mercuric acetate suggested that the 52–54 kDa gelatin-degrading activity was an activated form of MMP-2. In addition, Ab peptide induced both MMP-3 and plasminogen activator-like activity from astrocytes. When medium from Ab-treated, astrocyte cultures was immunoblotted for fibronectin, several immunopositive, lower molecular weight bands were observed as compared to untreated conditioned medium, suggestive of the presence of an active fibronectin-degrading protease. Thus, Ab induces the secretion of several matrix-degrading proteases and stimulates matrix degradation in rat astrocytes. Since matrix-degrading proteases are elevated in AD brain, these proteases may influence the stability of ECM or other MMP substrates and thus may play a role in the neurotrophic / neurotoxic events associated with AD.  2003 Elsevier Science B.V. All rights reserved. Theme: Disorders of the nervous system Topic: Degenerative disease: Alzheimer’s-beta amyloid Keywords: Gelatinase; Matrix metalloproteinase; Plasminogen activator; Alzheimer’s disease; Zymogram; Cytokines; Fibronectin; Stromelysin-1

1. Introduction Fibril forming b-amyloid (Ab) peptides form the core of senile and neuritic plaques that are located predominantly in cortex and hippocampus of Alzheimer’s disease (AD) brain. Amyloidogenic processing of amyloid precursor protein (APP) leads to fibrillar Ab deposition, whereas the bulk of processed APP is non-amyloidogenic. Either overproduction of Ab or increased processing to a more *Corresponding author. Tel.: 11-813-974-2543; fax: 11-813-9742565. E-mail address: [email protected] (P.E. Gottschall).

fibril-forming peptide results in deposition of amyloid in the central nervous system [29]. Ab exists primarily as insoluble aggregates in AD brain. Aggregated peptide is associated with greater neurotoxicity in hippocampal cultures [5] whereas the non-aggregated form induces neuritic outgrowth and interacts with the extracellular matrix (ECM) [16], although more recent data suggests that soluble oligomers of Ab may also be neurotoxic [14]. ECM may act as a stratum for accumulation of Ab during the formation of diffuse plaques. This accumulation may lead to neurite retraction and neuron susceptibility to degeneration and thus alter the survival and plasticity of neurons in their extracellular environment [20,25,41].

0006-8993 / 03 / $ – see front matter  2003 Elsevier Science B.V. All rights reserved. doi:10.1016 / S0006-8993(03)02344-8

206

S. Deb et al. / Brain Research 970 (2003) 205–213

MMPs are a family of matrix degrading enzymes that are Zn 11 and Ca 11 dependent. Two major post-translational mechanisms regulate MMP activity, cleavage of pro-MMPs to an active form and inhibition by tissue inhibitor of metalloproteinases (TIMPs). Serine proteases, such as plasmin or trypsin can cleave and activate most of the MMPs in vitro, but little is known about what occurs in vivo. MMPs, along with the plasminogen activator (PA) / plasmin system of serine proteases, carry out proteolytic degradation of the ECM. The system consists of the inactive proenzyme, plasminogen, which is activated to the proteolytic enzyme plasmin by PAs, tissue-type-PA and urokinase-PA. u-PA binds to a selective cell surface receptor and brings about cell-mediated proteolysis by activation of latent matrix-degrading proteinases, e.g. MMPs. Localized activity of u-PA at the cell surface assists in cell invasion through extracellular matrices [35]. Microglial cultures from immature rat brain have been shown to express u-PA. t-PA activity was observed in neural tissue and correlated with morphological differentiation, i.e., neuronal regeneration, migration and neurite outgrowth [17]. t-PA may play a role in permitting excitotoxin-induced neuronal degeneration [32] at least in part by promoting the degradation of the ECM protein laminin [4]. In addition, elevated levels of MMPs have been observed in AD brain regions as compared to control [3,18,42] and are capable of degrading Ab [2,26]. Thus, these enzymes are a part of the proteolytic cascade involved in both physiological and pathological processes in the nervous system. Fibronectin is a cell adhesion and attachment factor that exists in ECM in an insoluble, fibrillar form and plays a major role in development, morphogenesis, cytoskeletal organization, tumorigenesis and wound healing. The functional domains of fibronectin act as binding sites for a variety of molecules at or near the cell surface and fragments produced by protease degradation possess biological functions absent in the native molecule. Increased fragmentation of fibronectin has been observed in osteoarthritic synovial fluids [39] and in chronic wounds [37]. The amino terminal 29 and 50 kDa, gelatin-binding fragments increase proteoglycan release [9,38] and inhibit proteoglycan synthesis [38] in chondrocyte cultures. Induction of MMP-3 and MMP-1 expression in synovial fibroblasts mediated via the fibronectin receptor has been reported [36]. The 29-kDa amino terminal fragment of fibronectin was shown to inhibit Schwann cell proliferation [22]. Several MMPs including MMP-2 release active fibronectin fragments which might be involved in inflammation or tissue remodeling [7]. The aim of this study was to determine whether soluble or fibrillar Ab peptides are more potent in inducing MMP production in astrocyte cultures, whether MMPs and PAs are active in vitro in enriched astrocyte cultures after treatment with Ab peptides, and to examine whether these proteases induced by Ab peptide degrade ECM proteins, in particular fibronectin.

2. Materials and methods

2.1. Astrocyte culture and peptide treatment Astrocytes were prepared from the brains of 1-day old Sprague–Dawley rat pups as described [8], according to a protocol approved by the Laboratory Animal Medical Ethics Committees at the University of South Florida College of Medicine. After thirteen days and one subculture, morphologically, the cells consist of a monolayer of confluent, protoplasmic astrocytes partly covered with phase dark, fibrous type cells. We have previously shown that these cultures are 99% immunopositive for glial fibrillary acidic protein. Ab peptides (Ab(1–40) Lot [ QM365, WM365, ZN571 and Ab(1–42) Lot [ ZN327, WN327; Bachem, Torrance, CA, USA) were added to these cultures as ‘unaged’ or ‘aged’ (a process to promote fibril formation) peptides according to the suppliers instructions. Ab(1–40) and Ab(1–42) were ‘aged’ by diluting the peptide in double distilled water to 6 mg / ml, further diluting to 1 mg / ml in PBS, and incubated at 378 C for 3–5 days. Peptide termed ‘unaged’ was diluted in serum-free, tissue culture medium (DMEM, high glucose containing 25 mM HEPES buffer, GIBCO BRL, Rockville, MD, USA) just prior (within 30 min) to addition to the cells. The aged peptides, and to a lesser extent, freshly prepared Ab(1–42), showed significant red–green birefringence to polarized light when stained with Congo Red (Sigma, St. Louis, MO, USA) as compared to freshly prepared Ab(1–40) indicating their fibrillar state and aggregated nature (data not shown). Cell culture supernatants were collected at 24 or 72 h, frozen and stored.

2.2. Gelatin-substrate zymography and plasminogen casein zymography Gelatin substrate zymography was carried out by subjecting conditioned media samples to 7.5% SDS–PAGE containing 1 mg / ml of gelatin as described [10]. Using this technique both latent and active species are visualized. A 15-ml volume of conditioned media sample was diluted with 15 ml of 23 non-reducing sample buffer (2% w / v SDS, 10% glycerol, 0.05 M Tris–HCl pH 6.8 and 0.005% bromophenol blue). Samples were electrophoresed at 150 V for about 45 min on a mini-gel apparatus. The gels were washed twice with 2.5% Triton X-100 for 15 min each and then incubated at 378 C for 16–18 h in substrate buffer (20 mM Tris–HCl, pH 7.6, 10 mM CaCl 2 and 0.04% NaN 3 ). The gels were then stained with 0.1% Coomassie brilliant blue R-250 diluted in 40% methanol and 10% acetic acidfor 1 h and destained until clear zones were obtained on a homogeneous blue background. These clear zones represent gelatinolytic activity. Gelatinase activity, relative to the amount of enzyme loaded is linear up to high concentrations where additional enzyme results in proportionally less activity [15,43]. Standard molecular weight markers (BioRad, Hercules, CA, USA) were used to

S. Deb et al. / Brain Research 970 (2003) 205–213

estimate the molecular weight of the gelatinases. In some experiments, conditioned media samples were incubated with aminophenyl mercuric acetate (APMA), 1 mM for 24 h prior to SDS–PAGE. In an effort to detect plasminogen activators, casein– plasminogen mini-gels were cast that contain 4.5 mg of b-casein (Sigma) and 0.15 U plasminogen (Sigma) [27]. Incubations were carried out in 100 mM Tris–HCl, pH 7.6 containing 0.02% sodium azide. Gels were stained and destained in a manner similar to gelatin zymograms.

2.3. Immunoblots Conditioned media from untreated and treated enriched astrocyte cultures were subjected to SDS–PAGE, the proteins electrophoretically transferred to nitrocellulose and probed using antibodies raised against rabbit anti rat fibronectin (1:500, Chemicon International, Temecula, CA, USA) and antisera raised against rat MMP-3 (rabbit antirat transin obtained from Dr. Lynn Matrisian, Vanderbilt University, 1: 1000). The blots were developed using either anti-rabbit IgG conjugated to alkaline phosphatase and developed with a colored substrate (BCIP/ NBT; Sigma) or to horseradish peroxidase and developed with a luminol substrate (ECL, Amersham, Piscataway, NJ, USA).

2.4. Data analysis Each of the above experiments were repeated at least three times using conditioned media samples from independent cultures. For some of the data, representative blots or gels are shown. For others mean optical densities from several experiments are shown. For the zymographs, the optical density (OD) of the area of each clear zone on the gel was determined by integration of the peaks on inverted OD scans generated from an imaging densitometer (Bio-Rad). In some experiments, when activity was high, each activity was observed as a single (overlapping) band, whereas in other experiments, the activities were observed as doublets. The OD of the doublets were summed for the purpose of presentation and analysis. These are likely different isoforms of the same enzyme. Results are expressed as X6S.E.M. Significant differences between data points were determined by ANOVA followed by the Tukey-Kramer pairwise comparison test. A P , 0.05 was considered significant. The arbitrary densitometric units were converted to a percent of the maximum response for each individual dose–response curve in Fig. 1. Samples from each dose–response were subjected to SDS–PAGE on a single gel.

3. Results Enriched rat astrocyte cultures were incubated for 72 h with freshly prepared ‘unaged’ or ‘aged’ Ab(1–40) and

207

Ab(1–42) (see Methods) to examine the effects of aggregation of the peptide (as assessed by red–green birefringence to polarized light) on metalloproteinase secretion. Analysis of medium from untreated cultures showed a high constitutive expression of a 58–66-kDa gelatinase previously identified as MMP-2 [6]. Fig. 1 (top panels) shows the effect of increasing concentrations of freshly prepared Ab(1–40) and Ab(1–42) on gelatinase activity in astrocytes. Treatment of astrocytes with freshly prepared Ab(1–40) and Ab(1–42) induced a dose-dependent increase in MMP-9 activity, whereas only Ab(1–40), and not (1–42), caused the appearance of the 52–54-kDa gelatinase activity at the highest concentration of 40 mM (Fig. 1, top, left panel, arrow). Zymograms from three separate experiments were quantified by densitometric scanning. Unaged Ab(1–40) and Ab(1–42), at concentrations of 20 mM, stimulated MMP-9 activity nearly 20- and 16-fold, respectively, as compared to untreated cultures. For freshly prepared Ab(1–40), a 40-mM concentration of peptide was required to significantly stimulate the 52–54-kDa gelatinase activity as compared to untreated cultures (Fig. 1, bottom panel). Surprisingly, when Ab was aggregated prior to adding it to the cultures, Ab(1–40) and Ab(1–42) failed to stimulate either MMP-9 or the 52–54-kDa activity (Fig. 2). Little or no activity other than constitutively expressed MMP-2 was observed in these cultures, although a decrease in MMP-2 activity was seen at the highest concentrations of Ab(1– 42) suggesting possible cytotoxicity of the aggregated peptide. Conditioned media samples containing MMP-2 from Ab(1–40)-treated primary cultures were subjected to activation using the organomercurial compound, APMA, which activates MMPs by inducing autocatalytic cleavage of the proenzyme. Untreated conditioned media containing MMP-2 (Fig. 3, lane C), when treated with 1 mM APMA, cleaved and activated MMP-2 (Fig. 3, lane C(APMA)) as observed on a gelatin substrate zymogram. Autoactivation of latent MMP-2 produces an active 52–54 kDa activity. Zymographic analysis revealed that the molecular weight of APMA-activated MMP-2, was similar or identical to the 52–54 kDa activity observed in Ab(1–40)-treated conditioned medium (Fig. 3, lane Ab). Also, when Ab(1–40)treated conditioned media samples were activated with APMA, an increase in the density of this 52–54 kDa activity was observed (Fig. 3, lane Ab(APMA)) as compared to the same sample that was not activated with APMA (Fig. 3, lane Ab). This observation supports the hypothesis that Ab-conditioned medium contains activated MMP-2. In addition, this 52–54 kDa, activated MMP-2 does not appear after treatment with IL-1b or LPS (Fig. 4), even though the proform of MMP-2 is clearly induced. MMP-3 (transin or stromelysin-1), which has the broadest substrate specificity of all the MMPs, was induced by 40 mM Ab(1–40) (Fig. 5), but was not observed at lower concentrations of Ab. The doublet seen in these blots is characteristic of cross-reactivity of this antiserum with

208

S. Deb et al. / Brain Research 970 (2003) 205–213

Fig. 1. Top panels: effect of ‘unaged’ Ab peptides (see Methods) on gelatinase production from enriched rat astrocyte cultures. Enriched rat astrocytes were treated with increasing concentrations of freshly prepared Ab(1–40) or Ab(1–42) and conditioned media that was collected 72 h later was measured for gelatinase activity using gelatin-substrate zymography. Note the presence of a 52–54-kDa activity (arrow) in samples treated with 40 mM Ab(1–40). Numbers to the left of the gel signify migration of molecular weight standards. Bottom panels: densitometric analysis of zymograms from above. Data are expressed as the fraction of the maximal response (in arbitrary densitometric units) obtained with each individual zymogram for the 52–54-kDa band (closed triangle), 58–66 kDa-band (open circle) and the 94-kDa band (closed circle). Points and bars represent the mean6S.E.M., respectively (n 5 3 samples from independent cultures).

MMP-10 (stromelysin-2 in human) [30], although the possibility of two different forms of MMP-3 cannot be ruled out. Ab peptides stimulated expression of plasminogen activator activity as well. Increased plasminogen activator-like activity, as measured by plasminogen–casein zymography, was observed in conditioned medium obtained from Ab(1–40)-treated astrocytes as compared to untreated cells. Plasminogen–casein zymograms showed bands at approximately 40 and 60 kDa. In untreated conditioned media, only the 40-kDa band was observed whereas in Ab-treated cultures, there was induction of the 40-kDa band along with the appearance of a 60-kDa band at 20 and 40 mM concentrations (Fig. 6). Based on their

molecular weight, it is likely that the lower molecular weight activity was u-PA and the higher molecular weight band was t-PA (Fig. 6). Neither MMP-3 or PA activity was detectable on zymograms containing casein alone (data not shown). Since MMP-2, MMP-9, MMP-3 and PA-like activities were induced in astrocytes in response to Ab treatment, then substrates for these proteases may be degraded by the active proteases. Astrocytes in culture are well known to produce significant amounts of soluble fibronectin [19] that is a substrate for MMP-2 as well as plasmin. Ab-treated conditioned media obtained from enriched rat astrocytes, when immunoblotted for fibronectin using polyclonal anti-

S. Deb et al. / Brain Research 970 (2003) 205–213

Fig. 2. Effect of aged Ab peptides on gelatinase activity in enriched astrocyte cultures. Enriched rat astrocytes were treated with increasing concentrations of ‘aged’ (see Methods) Ab (1–40) or Ab (1–42) and conditioned media that was collected 72 h later was measured for gelatinase activity using gelatin-substrate zymography. Note the absence of the 52–54-kDa activity (arrow) and MMP-9 in samples treated with 40 mM Ab (1–40) or (1–42). Numbers to the right of the gel signify migration of molecular weight standards. Data from three independent cultures were quantitated and analyzed as in Fig. 1. There were no significant differences within the concentration response curve.

rat fibronectin, showed several lower molecular weight products indicative of degradation of the protein (Fig. 7, top). These products were easily detectable in Ab(1–42)treated medium at 24 h compared to a near absence in untreated medium. Also IL-1b and LPS-treated enriched astrocytes, which induce the latent forms of MMP-2 and MMP-9 in astrocytes (Fig. 4) [8] failed to induce the degradation of fibronectin at 24 h compared to control cultures. Ab(1–40) treatment at 20 mM resulted in increased number and more abundant fibronectin fragments as compared to 24 h. The molecular weights of these fibronectin immunopositive products were 160, 100, 72, 56 and 47 kDa (Fig. 7, top). Thus, Ab appeared to stimulate the release of proteases capable of degrading fibronectin. These fragments were not observed in untreated conditioned medium obtained from astrocyte culture even when the ECL film was overexposed to a similar magnitude as that of the induced holoprotein (Fig. 7, bottom). Whether the appearance of these fibronectin immunoreactive fragments in Ab-treated conditioned medium were

209

Fig. 3. Effect of the metalloproteinase activator, aminophenyl mercuric acetate (APMA), on Ab-treated and untreated conditioned media samples. Samples from untreated cells or cells treated with Ab(1–40) were activated with 1 mM APMA. The samples were subjected to zymography. lane C: untreated conditioned media containing constitutively expressed MMP-2, lane C(APMA): untreated conditioned media activated with 1 mM APMA for 1 h, lane Ab: Ab-treated conditioned media, lane Ab(APMA): Ab-treated conditioned media activated with 1 mM APMA for 1 h. Numbers to the left of the gel signify migration of molecular weight standards. The arrow represents the molecular weight of the active form of MMP-2.

due to the activation of MMP-2 or due to induction of a plasminogen activator-like activity or some other activity is not known.

4. Discussion In this study, the pattern of MMP activity induced in rat astrocytes by aged (aggregated, fibrillar) and freshly prepared Ab peptides was dissimilar. Only freshly prepared Ab(1–40), and not the aged peptide, was capable of inducing the appearance of active MMP-2 and MMP-9 in rat astrocytes. Freshly prepared Ab(1–42) induced MMP-9 but failed to stimulate the appearance of active MMP-2, whereas the aged peptide failed to stimulate either MMP-2 or MMP-9. Freshly prepared Ab(1–16) or (25–35) also failed to induce MMP-9 or MMP-2 activity in astrocytes [6]. In addition, astrocytes secreted MMP-3 and PA-like activity in response to unaggregated Ab(1–40) as compared to untreated cells. Although it has been suggested that Ab results in neurodegeneration observed in AD, it is neurotoxic only under certain conditions in vitro and aggregation of the peptide may be required for its neurotoxic action [25]. Interestingly, Ab(1–40), when incubated

210

S. Deb et al. / Brain Research 970 (2003) 205–213

Fig. 6. Induction of plasminogen activator-like activity by Ab(1–40) in enriched astrocyte cultures. Media samples were collected 72 h after addition of Ab and subjected to plasminogen–casein zymography. The locations of molecular weight markers are indicated on the left.

Fig. 4. Effect of interleukin-1b (IL-1b) and lipopolysaccharide (LPS) on gelatinase activity secreted from cultured rat astrocytes. Enriched rat astrocytes were treated with LPS or IL-1b and conditioned medium was collected 72 h later. Note the absence of the 52–54-kDa gelatinase activity.

Fig. 5. Detection of MMP-3 in enriched astrocyte cultures treated with Ab(1–40) for 72 h. Samples were subjected to SDS–PAGE, blotted to nitrocellulose, and probed with rabbit anti-rat MMP-3 (rat transin, stromelysin). Note the immunopositive doublet obtained with this antiserum (arrow).

for 1–4 days in culture medium, undergoes aggregation [20]. Ab(1–42) is deposited in AD brain prior to Ab(1– 40) [11], forms aggregates more rapidly than Ab(1–40) and may act as a seed for plaque formation [12] that includes Ab(1–40). Several studies have shown that aggregated Ab peptides are capable of stimulating cytokines, chemokines, reactive oxygen species, and the activation of cultured rat astrocytes [10,13,28]. In one these experiments, the soluble peptide (or peptide oligomers) was demonstrated to activate glia as well [10]. It is possible that the process of Ab aggregation in the presence of astrocytes in culture was necessary for protease production, since ‘pre-aggregated’ peptides were ineffective in eliciting MMP-2 or MMP-9 secretion. This may be the reason such high concentrations of Ab are required to induce MMP production (or activate other processes) in vitro. However, others have shown that aggregated, and not non-aggregated, Ab was effective in eliciting tPA and uPA release in neuronal cultures [33]. In vivo, aggregating Ab(1–40) or more likely (1–42) may act as a stratum for accumulation of matrix molecules during the slow formation of diffuse plaques and locally, active protease may be produced. This accumulation could lead to neurite retraction and degeneration and / or dystrophic neurite outgrowth, thus altering the survival and plasticity of neurons in their extracellular environment and stimulate further Ab deposition. On the other hand, active forms of MMP-2 and

S. Deb et al. / Brain Research 970 (2003) 205–213

211

Fig. 7. Degradation of fibronectin by Ab-peptide-induced release of proteinases in enriched astrocyte cultures. Upper panel. Media samples untreated or treated with IL-1b, LPS, Ab(1–42) for 24 h and Ab (1–40) for 72 h were subjected to SDS–PAGE, blotted to nitrocellulose, and probed with rabbit anti-rat fibronectin. The molecular weight of the degraded fibronectin fragments are shown on the right. Note lower molecular weight fragments in Ab-treated samples signifying proteolytic degradation. Bottom panel. Media samples treated or untreated with freshly prepared Ab (1–42) 20 mM were immunoblotted for fibronectin. Even when the immunoblot using untreated conditioned media from astrocyte cultures was overexposed to the extent of that of Ab, the low molecular weight degraded fragments were not present. (bottom right panel).

MMP-9 degrade plaque-derived or synthetic Ab [2,26]. It is assumed that activated MMP-2 cleaves the soluble form of Ab(1–40), whereas Ab fibrils are likely more difficult for proteases to degrade. Thus, secreted proteases in their active form originating from reactive glia may play a role in degrading Ab and hence limit plaque development. However, once amyloidosis begins, proteases may instead act upon the surrounding ECM that interacts with Ab and influence the rate of plaque deposition. As MMPs are activated by SDS during a substrate gel electrophoresis assay, it is not known whether the proteases are active in the culture plate, i.e., are they capable of degrading substrates such as matrix proteins? Our results indicate the presence of degraded fibronectin fragments in Ab-treated conditioned medium suggesting that

Ab peptide is capable of inducing fibronectin-degrading activity. This is important since senile plaques in AD brain demonstrate the accumulation of matrix proteins such as fibronectin [34]. Proteoglycans, laminin, fibronectin as well as acute phase proteins such as a 1 -antichymotrypsin and a 1 -antitrypsin are all potential substrates for the MMPs [21]. Several of these proteins are integral components of the neuritic plaques of AD. MMP-9 levels are elevated in plaque-affected brains of AD patients as compared to healthy subjects [3,40] and MMP-9 was shown to be expressed by neurons [2]. Activated MMP-2 and MMP-9 degrade type IV collagen, a matrix molecule which forms the important component of capillary basal lamina in brain tissue. MMP-3, which has the broadest ECM substrate

212

S. Deb et al. / Brain Research 970 (2003) 205–213

specificity, was shown to be expressed in senile plaques in AD [42]. MMP-9 may be activated in vitro by MMP-3, plasmin and PA, however activation mechanisms in brain are not known. Plasmin, active MMP-2, -3 and -9 as well as membrane-type-MMP are all individually capable of degrading fibronectin. Also MMP-3 has been implicated in the cleavage of versican derived, glial hyaluronic acid binding protein that has been observed in senile plaques [24]. Inhibitors such as a 1 -anti-chymotrypsin [1], an inhibitor of serine proteases, a 2 -macroglobulin [31] and TIMP-1 [23], all inhibitors of MMPs have been shown to accumulate in amyloid plaques. Activation of latent metalloproteases would hence be controlled by such inhibitors and may play a role in the amyloidosis cascade. In summary, we have demonstrated that b-amyloid is effective in stimulating an ‘active’ gelatinase activity and plasminogen activator-like activity in cultures of rat astrocytes. Also proteases are effective in degrading an ECM substrate protein, fibronectin, after induction by Ab in astrocyte cultures. The local production of MMPs in areas of active plaque formation, for example in diffuse plaques, might suggest an important role for MMPs in plaque formation or in Ab degradation, two functions which are not mutually exclusive.

Acknowledgements The authors would like to acknowledge the gift of anti-rat MMP-3 serum from Dr. Lynn Matrisian, Vanderbilt University, Nashville, TN, USA. This work was supported in part by NIH grant AG12160 (P.E.G.).

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

References [1] C.R. Abraham, D.J. Selkoe, H. Potter, Immunohistochemical identification of the serine protease inhibitor alpha1-antichymotrypsin in the brain amyloid deposits of Alzheimer’s disease, Cell 52 (1988) 487–501. [2] J.R. Backstrom, G.P. Lim, M.J. Cullen, Z.A. Tokes, Matrix metalloproteinase-9 (MMP-9) is synthesized in neurons of the human hippocampus and is capable of degrading the amyloid-b peptide, J. Neurosci. 16 (1996) 7910–7919. [3] J.R. Backstrom, C.A. Miller, Z.A. Tokes, Characterization of neutral proteinases from Alzheimer-affected and control brain specimens: identification of calcium-dependent metalloproteinases from the hippocampus, J. Neurochem. 58 (1992) 983–992. [4] Z.L. Chen, S. Strickland, Neuronal death in the hippocampus is promoted by plasmin-catalyzed degradation of laminin, Cell 91 (1997) 917–925. [5] C.W. Cotman, C.J. Pike, A. Copani, beta-Amyloid neurotoxicity—a discussion of in vitro findings, Neurobiol. Aging 13 (1992) 587– 590. [6] S. Deb, P.E. Gottschall, Increased production of matrix metalloproteinases in enriched astrocyte and mixed hippocampal cultures treated with b-amyloid peptides, J. Neurochem. 66 (1996) 1641– 1647. [7] F. Fukai, M. Ohtaki, N. Fujii, H. Yajima, T. Ishii, Y. Nishijawa, K.

[20]

[21]

[22]

[23]

[24]

[25]

[26]

Miyazaki, T. Katayama, Release of biological activities from quiescent fibronectin by conformational change and limited proteolysis by matrix metalloproteinases, Biochemistry 34 (1995) 11453–11459. P.E. Gottschall, X. Yu, Cytokines regulate gelatinase A and B (matrix metalloproteinase 2 and 9) activity in cultured rat astrocytes, J. Neurochem. 64 (1995) 1513–1520. G.A. Homandberg, R. Meyers, D.L. Xie, Fibronectin fragments cause chondrolysis of bovine articular cartilage slices in culture, J. Biol. Chem. 267 (1992) 3597–3604. J. Hu, K.T. Akama, G.A. Krafft, B.A. Chromy, L.J. Van Eldik, Amyloid-beta peptide activates cultured astrocytes: morphological alterations, cytokine induction and nitric oxide release, Brain Res. 785 (1998) 195–206. T. Iwatsubo, A. Odaka, N. Suzuki, H. Mizusawa, N. Nukina, Y. Ihara, Visualization of Ab42(43) and Ab40 in senile plaques with end-specific Ab-monoclonals: evidence that an initially deposited species is Ab42(43), Neuron 13 (1994) 45–53. J.T. Jarrett, P.T. Lansbury, Seeding ‘one-dimensional crystalization’ of amyloid: a pathogenic mechanism in Alzheimer’s disease, Cell 73 (1993) 1055–1058. M. Johnstone, A.J. Gearing, K.M. Miller, A central role for astrocytes in the inflammatory response to beta-amyloid; chemokines, cytokines and reactive oxygen species are produced, J. Neuroimmunol. 93 (1999) 182–193. W.L. Klein, G.A. Krafft, C.E. Finch, Targeting small Abeta oligomers: the solution to an Alzheimer’s disease conundrum?, Trends Neurosci. 24 (2001) 219–224. D.E. Kleiner, W.G. Stetler-Stevenson, Quantitative zymography: detection of picogram quantities of gelatinases, Anal. Biochem. 218 (1994) 325–329. E.H. Koo, L. Park, D.J. Selkoe, Amyloid b-protein as a substrate interacts with extracellular matrix to promote neurite outgrowth, Proc. Natl. Acad. Sci. U.S.A. 90 (1993) 4748–4752. A. Krystosek, N.W. Seeds, Plasminogen activator secretion by granule neurons in cultures of developing cerebellum, Science 213 (1981) 7810–7814. A. Leake, C.M. Morris, J. Whateley, Brain matrix metalloproteinase 1 levels are elevated in Alzheimer’s disease, Neurosci. Lett. 291 (2000) 201–203. M. Mahler, Y. Ben-Ari, A. Represa, Differential expression of fibronectin, tenascin-C and NCAMs in cultured hippocampal astrocytes activated by kainate, bacterial lipopolysaccharide or basic fibroblast growth factor, Brain Res. 775 (1997) 63–73. M.P. Mattson, B. Cheng, D. Davis, K. Bryant, I. Leieberburg, R.E. Rydel, b-Amyloid peptides destabilize calcium homeostasis and render human cortical neurons vulnerable to excitotoxicity, J. Neurosci. 12 (1992) 376–389. L.J. McCawley, L.M. Matrisian, Matrix metalloproteinases: they’re not just for matrix anymore!, Curr. Opin. Cell Biol. 13 (2001) 534–540. D. Muir, M. Manthorpe, Stromelysin generates a fibronectin fragment that inhibits Schwann cell proliferation, J. Cell Biol. 116 (1992) 177–185. N. Peress, E. Perillo, S. Zucker, Localization of tissue inhibitor of matrix metalloproteinases in Alzheimer’s disease and normal brain, J. Neuropathol. Exp. Neurol. 54 (1995) 16–22. G. Perides, R.A. Asher, M.W. Lark, W.S. Lane, R.A. Robinson, A. Bignami, Glial hyaluronate-binding protein: a product of metalloproteinase digestion of versican?, Biochem. J. 312 (1995) 377–384. C. Pike, D. Burdick, A.J. Walencewicz, C.G. Glabe, C.W. Cotman, Neurodegeneration induce by b-amyloid peptides in vitro: the role of peptide assembly state, J. Neurosci. 13 (1993) 1676–1687. A.E. Roher, T.C. Kasunic, A.S. Woods, R.J. Cotter, M.J. Ball, R. Fridman, Proteolysis of Aß peptide from Alzheimer disease brain by gelatinase A, Biochem. Biophys. Res. Commun. 205 (1994) 1755– 1761.

S. Deb et al. / Brain Research 970 (2003) 205–213 [27] G.A. Rosenberg, J.E. Dencoff, P.G. McGuire, L.A. Liotta, W.G. Stetler-Stevenson, Injury-induced 92-kilodalton gelatinase and urokinase expression in rat brain, Lab. Invest. 71 (1994) 417–422. [28] F. Rossi, E. Bianchini, Synergistic induction of nitric oxide by beta-amyloid and cytokines in astrocytes, Biochem. Biophys. Res. Commun. 225 (1996) 474–478. [29] D.J. Selkoe, Alzheimer’s disease: a central role for amyloid, J. Neuropathol. Exp. Neurol. 53 (1994) 438–447. [30] T. Sreenath, L.M. Matrisian, W. Stetler-Stevenson, S. Gattoni-Celli, R.O. Pozzatti, Expression of matrix metalloproteinase genes in transformed rat cell lines of high and low metastatic potential, Cancer Res. 52 (1992) 4942–4947. [31] S. Strauss, J. Bauer, U. Ganter, U. Jonas, M. Berger, B. Volk, Detection of interleukin-6 and a2-macroglobulin immunoreactivity in cortex and hippocampus of Alzheimer’s disease patients, Lab. Invest. 66 (1992) 223–230. [32] S.E. Tsirka, A. Gualandris, D.G. Amara, S. Strickland, Excitotoxininduced degeneration and seizure are mediated by tissue plasminogen activator, Nature 377 (1995) 340–344. [33] H.M. Tucker, M. Kihiko, J.N. Caldwell, S. Wright, T. Kawarabayashi, D. Price, D. Walker, S. Scheff, J.P. McGillis, R.E. Rydel, S. Estus, The plasmin system is induced by and degrades amyloid-b-aggregates, J. Neurosci. 20 (2000) 3937–3946. [34] D. Van Gool, G. Carmeliet, E. Triau, J.J. Cassiman, R.D., Appearance of localized immunoreactivity for the alpha 4 integrin subunit and for fibronectin in brains from Alzheimer’s, Lewy body dementia patients and aged controls, Neurosci. Lett. 170 (1994) 71–73.

213

[35] J.D. Vassalli, M.S. Pepper, Membrane proteases in focus, Nature 370 (1994) 14–15. [36] Z. Werb, Proteinases and matrix degradation, in: W.N. Kelly, E.D.J. Harris, S. Ruddy, C.B. Sledge (Eds.), Textbook of Rheumatology, W.B. Saunders, Philadelphia, 1989, pp. 300–321. [37] A.B. Wysocki, F. Grinnell, Fibronectin profiles in normal and chronic wound fluid, Lab. Invest. 63 (1990) 825–831. [38] D.L. Xie, G.A. Homandberg, Fibronectin fragments bind to and penetrate cartilage tissue resulting in proteinase expression and cartilage damage, Biochim. Biophys. Acta 1182 (1993) 189–196. [39] D.L. Xie, R. Meyers, G.A. Homandberg, Fibronectin fragments in osteoarthritic synovial fluid, J. Rheumatol. 19 (1992) 1448–1452. [40] T. Yamada, Y. Yoshiama, H. Sato, M. Seiki, A. Shinagawa, M. Takahashi, White matter microglia produce membrane-type matrix metalloprotease, an activator of gelatinase A, in human brain tissues, Acta Neuropathol. 90 (1995) 421–424. [41] B.A. Yankner, Commentary and perspective on studies of betaamyloid neurotoxicity, Neurobiol. Aging 13 (1992) 615–616. [42] Y. Yoshiyama, M. Asahina, T. Hattori, Selective distribution of matrix metalloproteinase-3 (MMP-3) in Alzheimer’s disease brain, Acta Neuropathol. 99 (2000) 91–95. [43] J.W. Zhang, P.E. Gottschall, Zymographic measurement of gelatinase activity in brain tissue after detergent extraction and affinity-support purification, J. Neurosci. Methods 76 (1997) 15–20.