• Open Access

Wild-type amyloid beta 1-40 peptide induces vascular smooth muscle cell death independently from matrix metalloprotease activity

Authors


Isabelle Limon, University Paris 6, UR4, Vieillissement, Stress et Inflammation 7 quai Saint-Bernard, Bat A 5eme etage, 75252 Paris, France. Tel.: +33 144 273 716; fax: +33 144 274 140; e-mail:isabelle.limon@snv.jussieu.fr

Summary

Cerebral amyloid angiopathy (CAA) is an important cause of intracerebral hemorrhages in the elderly, characterized by amyloid-β (Aβ) peptide accumulating in central nervous system blood vessels. Within the vessel walls, Aβ-peptide deposits [composed mainly of wild-type (WT) Aβ1-40 peptide in sporadic forms] induce impaired adhesion of vascular smooth muscle cells (VSMCs) to the extracellular matrix (ECM) associated with their degeneration. This process often results in a loss of blood vessel wall integrity and ultimately translates into cerebral ischemia and microhemorrhages, both clinical features of CAA. In this study, we decipher the molecular mechanism of matrix metalloprotease (MMP)-2 activation in WT-Aβ1-40-treated VSMC and provide evidence that MMP activity, although playing a critical role in cell detachment disrupting ECM components, is not involved in the WT-Aβ1-40-induced degeneration of VSMCs. Indeed, whereas this peptide clearly induced VSMC apoptosis, neither preventing MMP-2 activity nor hampering the expression of membrane type1-MMP, or preventing tissue inhibitors of MMPs-2 (TIMP-2) recruitment (two proteins evidenced here as involved in MMP-2 activation), reduced the number of dead cells. Even the use of broad-range MMP inhibitors (GM6001 and Batimastat) did not affect WT-Aβ1-40-induced cell apoptosis. Our results, in contrast to those obtained using the Aβ1-40 Dutch variant suggesting a link between MMP-2 activity, VSMC mortality and degradation of specific matrix components, indicate that the ontogenesis of the Dutch familial and sporadic forms of CAAs is different. ECM degradation and VSMC degeneration would be tightly connected in the Dutch familial form while being two independent processes in sporadic forms of CAA.

Introduction

Cerebral amyloid angiopathy (CAA) refers to sporadic and hereditary cerebrovascular disorders frequently associated with cognitive impairment in the elderly, including Alzheimer’s disease. From a histopathological point of view, it is characterized by the amyloid accumulation in the media and adventitia of small and large arteries irrigating the central nervous system and/or amyloid deposition of Aβ peptides around the capillaries perfusing the cerebellum, cerebral cortex and leptomeninges (Smith & Greenberg, 2009). The Aβ-peptide accumulation within arteries induces vascular smooth muscle cell (VSMC) death (referred to, in the literature, as VSMC degeneration) and results in a loss of blood vessel wall integrity. This possibly translates into cerebral ischemia and microhemorrhages, both clinical features of CAA (Knudsen et al., 2001; Maia et al., 2007). The Aβ-induced VSMC degeneration is associated with impaired VSMC adhesion to the extracellular matrix (ECM) because of elevated pericellular proteolysis of the ECM components (Maruyama et al., 1990; Kawai et al., 1993; Mok et al., 2006).

The human matrix metalloprotease (MMP) family encoded by 24 genes is composed of Zn2+-dependent proteases known to degrade a large variety of ECM components and a number of bioactive molecules at the proximity of the cell surface. Among all the MMPs, six members are anchored to the cell membrane (membrane-type MMPs, MT-MMPs), whereas the other members are soluble and secreted into the extracellular space. Of note, there is increasing evidence that soluble MMPs (such as MMP-2) are recruited to the local cell environment by interacting with cell surface proteins (including MT-MMPs) and via the pericellular matrix (Murphy & Nagase, 2011). MMP activity is regulated at transcriptional and post-translational levels. All MMPs are synthesized as inactive zymogens. The cysteine switch containing N-terminal propeptide of the latent pro-enzyme interacts with the Zn2+ at the active site, blocking proteolytic activity (Van Wart & Birkedal-Hansen, 1990). The first post-translational modification cleaves the propeptide allowing the latent proenzyme to become activated, enabling proteolysis of its substrate molecules. The second one involves the broad-spectrum proteinase inhibitor β2-macroglobulin and four specific inhibitors named tissue inhibitors of MMPs (TIMPs-1 to 4, Visse & Nagase, 2003). A 1:1 stoichiometric interaction of MMPs with TIMPs inhibits the enzyme activity, whereas unbalancing this ratio in favor of the proteases results in increasing MMP activity.

Aβ-peptides are produced by proteolytic processing of the amyloid protein precursor (APP) by β- and γ-secretases. Amyloid deposits within the vessel walls are mainly composed of wild-type (WT) in sporadic CAAs or mutated forms in familial CAAs of the 40 amino acids Aβ-peptide species (Aβ1-40) (Alonzo et al., 1998). Previous studies have demonstrated that a mutated form, the Dutch mutant E22Q Aβ1-40 (Aβ1-40D) involved in a rare but severe hereditary CAA (Levy et al., 1990), triggers the expression and activation of MMP-2 in human smooth muscle cells (HSMC, Davis & Van Nostrand, 1996; Jung et al., 2003); they also suggest that it may contribute to the Aβ1-40D-induced HSMC death. However, the pathogenicity of WT Aβ1-40 (WT-Aβ1-40) and the possible role of MMPs in sporadic CAAs (which represent more than 90% of CAAs) have not been defined. In this study, we demonstrate that WT-Aβ1-40 induces both VSMC apoptosis and the MT1-MMP-dependent MMP-2 activation in vitro and decipher the molecular mechanism of this activation. We also prove that there is no cause-to-effect relationship between WT-Aβ1-40-induced cell death and MMP activity suggesting that MMP proteolytic activity is not the primary cause of VSMC apoptosis observed in sporadic CAA.

Results

WT-Aβ1-40 decreases VSM cells viability

Comparing cell morphology of WT-Aβ1-40 with that of inverted Aβ1-40-treated VSMC revealed modifications of their adhesion to the ECM evident by cell shape alteration (Fig. 1A). Hoechst nuclei labeling followed by immunofluorescence microscopy evidenced pycnotic nuclei, characteristic of cells undergoing apoptosis. Consistently, the percentage of hypodiploid nuclei [measured by propidium iodide flow cytometric assay (Mateo et al., 2007)] was more than 70% after a 72 h-treatment with WT-Aβ1-40 peptide; inverted Aβ40-1 peptide has no effect on spontaneous cell death (Fig. 1B). Apoptosis induced by WT-Aβ1-40 was also measured by evaluating phosphatidylserine (PS) exposure in the cytoplasmic outer leaflet membrane using APC-conjugated Annexin-V in combination with 7-amino-actinomycin D exclusion dye (data not shown).

Figure 1.

 Wild-type (WT)-Aβ1-40 peptide induces vascular smooth muscle cells apoptosis. Serum-starved cells were treated for 48 h with WT-Aβ1-40 (50 μm), inverted Aβ40-1 (50 μm) or vehicle (control). (A) Cell morphology was evaluated by phase contrast microscopy, and nuclei were stained with Hoechst (blue). Images are representative of three independent experiments. (B) Apoptosis was determined as described in Materials and methods section by propidium iodide staining. Data represent means ± SD of four independent experiments performed in triplicate; n.s, not significant and ***P < 0.001 compared with control (Ctl).

WT-Aβ1-40 induces MT1-MMP-dependent MMP-2 activation

MT1-MMP and MMP-2 activity are known to be essential for pericellular proteolysis (Itoh & Seiki, 2006), whereas MMP-9 is rather involved in inflammatory diseases, including peripheral arterial diseases (Busti et al., 2010). MMP-2 and MMP-9 are the predominant soluble MMPs in VSMCs (Newby, 2006). As both processes, pericellular proteolysis and inflammation, are the major pathogenic features in CAA, we next studied the expression and activity of these MMPs in WT-Aβ1-40-treated VSMCs. As shown in Fig. 2A, left panel, MMP-2 mRNA level (estimated by quantitative RT–PCR) was more than 6-fold increased (P < 0.001) in cells treated with WT-Aβ1-40, when compared to cells incubated with vehicle or the inverted Aβ40-1 control peptide. MMP-9 mRNA was undetectable either in control, inverted Aβ40-1 or WT-Aβ1-40-treated cells (Fig. 2A, right panel), while IL-1β induced the expression of MMP-9 transcripts in agreement with earlier reports (Delbosc et al., 2008). The active form of the MMP-2 protein (referred to as Act. MMP-2, Fig. 2B) could only be detected in WT-Aβ1-40-treated VSMCs. In vehicle or inverted Aβ40-1-treated cells, the observed bands corresponded to nonmature forms of the enzyme (Fig. 2B). Of note, the lower unidentified band (*) was always present independently of the MMP-2 antibody used. Accordingly, MMP-2 activity (revealed by a 68-kDa zymographic band referred to as Act. MMP-2 in Fig. 2C) was only detectable in media conditioned by cells treated with WT-Aβ1-40. The identity of the zymographic bands was controlled by immunoprecipitation using the anti-MMP-2 antibody and naïve Ig on conditioned media derived from rat lung incubated in serum-free medium. The active form of MMP-2 was identified as the lowest band by (aminomethyl) phosphinic acid (AMPA)-induced activation of a control sample prior to zymography (data not shown). According to the level of MMP-9 transcripts, WT-Aβ1-40 did not trigger MMP-9 activity, although IL-1β treatment significantly increased it.

Figure 2.

 Wild-type (WT)-Aβ1-40 peptide increases expression and activation of matrix metalloprotease (MMP)-2, but not MMP-9 in vascular smooth muscle cells. Serum-starved cells were treated for 48 h with IL-1β (10 ng mL−1), WT-Aβ1-40 (50 μm), inverted Aβ40-1 (50 μm), or vehicle (control). (A) Transcripts encoding MMP-2 (left panel) and MMP-9 (right panel) were assayed by RT–PCR. Data represent means ± SD of four independent experiments performed in triplicate; n.s, not significant, n.d, not detectable, and ***P < 0.001 compared with control (Ctl). (B) MMP-2 immunoblot on total proteins (20 μg). The Western blot is representative of two independent experiments. Int. MMP-2, intermediate form of MMP-2; Act. MMP-2, Active form of MMP-2; *Nonspecific band. (C) Culture medium was recovered after treatment; 45 μL were used for determination of secreted MMP-2 and MMP-9 activities on gelatin zymography as described in Materials and methods. The lysis bands corresponding to the pro- or active form (Act.) of MMP-2 and MMP-9 were visualized after Coomassie Blue staining and decoloration. The gelatin zymography image is representative of four independent experiments.

We also demonstrated an increase of MT1-MMP expression in WT-Aβ1-40-treated cells at both mRNA and protein levels by quantitative RT–PCR (Fig. 3A, left panel) and Western blot (Fig. 3A, right panel). Because MMP-2 activation is dependent on MT1-MMP activity (Strongin et al., 1995), we next examined whether WT-Aβ1-40-induced MT1-MMP expression is responsible for MMP-2 activation. To do so, VSMCs were transfected with a plasmid encoding the catalytically inert E240A mutant of MT1-MMP (Fig. 3B). MMP-2 status was analyzed by gelatin zymography. The overexpression of MT1-E240A protein was confirmed by Western blot (Fig. 3B, right panel). As expected, the forced expression of MT1-MMP E240A completely blocked the WT-Aβ1-40-induced MMP-2 activity (Fig. 3B, left panel). MMP-2 mRNA expression was not affected by the expression of MT1-E240A constructs whether with or without WT-Aβ1-40 treatment (data not shown). Altogether, these data highlighted a cause-to-effect relationship between the WT-Aβ1-40 induction of MT1-MMP expression and MMP-2 activity.

Figure 3.

 Wild-type (WT)-Aβ1-40 peptide increases MT1-matrix metalloprotease (MMP) expression, leading to MMP-2 activation in vascular smooth muscle cells. Serum-starved cells were treated for 48 h with WT-Aβ1-40 (50 μm), inverted Aβ40-1 (50 μm) or vehicle (control). (A) Left Panel. Transcripts encoding MT1-MMP were assayed by RT–PCR. Data represent the means ± SD of four independent experiments performed in triplicate, n.s, not significant, and ***P < 0.001 compare to control (Ctl). Right Panel. Total proteins (20 μg) were separated by electrophoresis. MT1-MMP and GAPDH were immunodetected with appropriate antibodies. (B) Culture media samples from mock or inert MT1-MMP E240A-transfected cells were recovered after indicated treatment and the MMP activity was evaluated by gelatin zymography (left panel). The protein bands corresponding to the pro- or active form of MMP-2 were visualized after Coomassie Blue staining and decoloration. The zymography image is a representative of three independent experiments. Expression level of MT1-MMP E240A was evaluated by Western blot (right panel).

WT-Aβ1-40-induced MMP-2 activation is regulated by TIMP-2 accumulation at the cell surface

It has been shown that the MT1-MMP–dependent MMP-2 activation requires strictly regulated amounts of TIMP-2 molecules (Strongin et al., 1995). In this activation mechanism, TIMP-2 bound to the catalytic domain of MT1-MMP acts as a ‘receptor’ for pro-MMP-2 at the cell surface. Within this protein complex, the hemopexin domain of pro-MMP-2 binds to the C-terminal domain of TIMP-2 allowing proteolytic activation of pro-MMP-2 by an adjacent molecule of MT1-MMP that is free of TIMP-2 (see Fig. S1). In this scenario, the TIMP-2 molecule, part of the trimolecular complex (composed of MT1-MMP/TIMP-2/pro-MMP-2), acts as an MMP-2 activator. Here, TIMP-2 implication in WT-Aβ1-40-induced MMP-2 activation was evidenced by the drastic reduction of MMP-2 active forms (Fig. 4A, left panel) visualized in cells transfected with TIMP-2 siRNA compared with control siRNA-transfected cells. TIMP-2 siRNA efficiency was attested by RT–PCR, Fig. 4A, left panel. To study the mechanism of MMP-2 activation, we analyzed TIMP-2 protein levels in control and WT-Aβ1-40-treated whole cell extracts (as opposed to cell media) assuming that it would reflect TIMP-2 recruitment to the cell surface. We also defined TIMP-2 subcellular localization. TIMP-2 protein was not present in controls (whether in vehicle- or inverted peptide-treated VSMCs), but highly accumulated in the WT-Aβ1-40 treated VSMCs (Fig. 4B left panel). Data obtained from immunocytochemistry experiments evidenced that TIMP-2 staining defines the plasma membrane in WT-Aβ1-40-treated cells (Fig. 4C). Altogether, these results strongly suggest the WT-Aβ1-40-dependent accumulation of TIMP-2 at the cell surface. Because WT-Aβ1-40 activates MMP-2 through MT1-MMP, we next examined whether TIMP-2 could be recruited at the VSMC surface by silencing MT1-MMP expression. As shown in Fig. 4D, the intensity of the TIMP-2 band was strongly reduced in MT1-MMP-silenced cells compared with control siRNA-treated cells. As neither the amount of TIMP-2 mRNA nor that of secreted protein (Fig. 4B, right panel) was modified by the WT-Aβ1-40 treatment, we concluded that the WT-Aβ1-40-dependent accumulation of TIMP-2 takes place at the cell surface, because of its interaction with the upregulated membrane type 1-MMP (MT1-MMP).

Figure 4.

 Matrix metalloprotease (MMP)-2 activation is dependent on the accumulation of TIMP-2 at the cell surface induced by wild-type (WT)-Aβ1-40 in vascular smooth muscle cells. Serum-starved cells were treated for 48 h with WT-Aβ1-40 (50 μm), inverted Aβ40-1 (50 μm) or vehicle (control). (A) Culture media obtained from cells previously transfected with control siRNA or TIMP-2 siRNA were collected after indicated treatment and MMP activities were evaluated by gelatin zymography (left panel). The gelatin zymography image shown is representative of three independent experiments. TIMP-2 siRNA efficiency was evaluated by RT–PCR (right panel); data represent the means ± SD of three independent experiments performed in triplicate, ***P < 0.001 compared with control siRNA related to each treatment. (B) Left Panel. Cell lysate samples (20 μg) were separated by electrophoresis; TIMP-2 and GAPDH were immunodetected with the appropriate antibodies; the western blot shown is representative of five independent experiments. Right Panel. Transcripts encoding TIMP-2 were assayed by RT–PCR (upper right panel). The data represent the means ± SD of four independent experiments performed in triplicate, n.s, not significant compared with control (Ctl). TIMP-2 immunoblot was performed on 40 μL of conditioned culture medium separated by electrophoresis (lower right panel). The western blot is representative of two independent experiments. (C) Immunostaining of TIMP-2 on PFA-fixed cells was performed after 24 h-treatment using monoclonal antibody against TIMP-2 and a secondary antibody coupled to FITC (Green Stain, 20×). Cell nuclei were stained with Hoechst (blue, 20× same microscopic field as FITC). (D) Total proteins (20 μg) obtained from treated cells previously transfected with control or MT1-MMP siRNA were separated by electrophoresis; TIMP-2 and GAPDH were immunodetected with appropriate antibodies. The western blot shown is representative of three independent experiments.

MMPs activity is not involved in WT-Aβ1-40 induced cell death

Anoïkis is a form of programmed cell death induced by cell detachment from the ECM. As MMP activity contributes to ECM degradation, we questioned whether WT-Aβ1-40-induced VSMC death is linked to MT1-MMP/MMP-2 activation. We analyzed the effect of MMP-2 or MT1-MMP silencing with siRNA on the WT-Aβ1-40-induced VSMC apoptosis. The efficiency of MMP-2 and MT1-MMP silencing was estimated by RT–PCR and gelatin zymography (Fig. 5A,C). To our surprise, MMP-2 silencing did not change the rate of apoptosis in cells treated with the WT-Aβ1-40 (Fig. 5B), neither did MT1-MMP siRNA (Fig. 5D), as the percentage of apoptotic cells was very similar whether MT1-MMP expression was silenced or not (siRNA MT1-MMP:79.2 ± 7.4; siRNA control: 69.3 ± 5.5). Similar results were obtained in the MCF-7 breast carcinoma cell subline deficient in MT1-MMP and MMP-2 expression (Rozanov et al., 2001) (Fig. S2). Altogether, these experiments demonstrated that MMP-2 and MT1-MMP are not involved in the WT-Aβ1-40-induced cell death.

Figure 5.

 Effect of matrix metalloprotease (MMP)-2 and MT1-MMP silencing on wild-type (WT)-Aβ1-40-induced cell death. Serum-starved cells were treated for 48 h with WT-Aβ1-40 (50 μm), or vehicle (control). (A,C) upper panel. siRNAs efficiency was evaluated by RT–PCR on transcripts obtained from cells transfected with control, MMP-2 siRNA (A) or MT1-MMP siRNA (C). Data represents the means ± SD of three independent experiments performed in triplicate, ***P < 0.001 compared with vehicle (Ctl)-treated cells transfected with control siRNA. ###P < 0.001 compared with WT-Aβ1-40-treated cells transfected with control siRNA. (A,C) lower panel. Culture medium samples issued from MMP-2 siRNA (A) or MT1-MMP siRNA (C) -transfected cells were collected after indicated treatment; MMP activities were evaluated by gelatin zymography. The zymography shown are representative of three independent experiments. (B,D) the apoptosis of MMP-2 siRNA (B) or MT1-MMP siRNA (D) -transfected cells was determined by Propidium Iodide staining. Data represents the means ± SD of three independent experiments performed in triplicate. n.s, not significant; ***P < 0.001; **P < 0.01 and *P < 0.05 compared with related control (Ctl).

As MMPs without gelatinase activity could be expressed in VSMCs, including MMP-1, MMP3, MMP-10, MMP-13 (Schonbeck et al., 1997; Mao et al., 1999; Wu et al., 2003), we evaluated the expression of these enzymes in WT-Aβ1-40-treated VSMCs and showed that the expression of MMP-3 and MMP-13 mRNA was induced by WT-Aβ1-40 (Fig. 6A). Nevertheless, although WT-Aβ1-40-induced MMP activity was inhibited by GM6001 or Batimastat (two broad-spectrum hydroxamate inhibitors of MMPs, Fig. 6B), these compounds had no effect on WT-Aβ1-40-induced cell death (Fig. 6C). These data clearly demonstrate that MMP proteolytic activity is not the primary cause of VSMC apoptosis induced by WT-Aβ1-40.

Figure 6.

 Effect of matrix metalloprotease (MMP) inhibitors on wild-type (WT)-Aβ1-40-induced cell death. Serum-starved cells were treated for 48 h with WT-Aβ1-40 (50 μm), or vehicle (control), with or without MMP inhibitors (GM6001 or Batimastat). (A) Transcripts encoding MMP-3 (left panel) and MMP-13 (right panel) were assayed by RT–PCR. Data represent the means ± SD of four independent experiments performed in triplicate, ***P < 0.001 and *P < 0.05 compared with control (Ctl). (B) Inhibition of WT-Aβ1-40-induced MMP activities by GM6001 (25 μm) or Batimastat (2.5 μm) treatment was evaluated by adding the MMP fluorescent substrate (40 μm) to the collected culture medium. The fluorescence of MMP-cleaved substrate (λex = 320 nm, λem = 390 nm) was measured with a fluorescence multiplate reader. Data represent the means ± SD of three independent experiments performed in triplicate, n.s, not significant; ***P < 0.001, ###P < 0.001 and ‡‡‡P < 0.001 compared with related control (Ctl). (C) Apoptosis of treated cells was determined by Eth-DIII staining. Data represent the means ± SD of three independent experiments performed in triplicate; n.s, not significant and ***P < 0.001 compared with related control (Ctl). The fluorescence of DNA-bound EthD-III (λex = 515 nm, λem = 620 nm) was measured with a fluorescence multiplate reader. The results are expressed in percentage of dead cells compared with cells treated by absolute ethanol (referred to as 100% dead cells). Data represent the means ± SD of six independent experiments performed in triplicate; n.s, not significant; ***P < 0.001, ###P < 0.001 and ‡‡‡P < 0.001 compared with related control (Ctl)-treated cells.

Discussion

Here, we show that the WT-Aβ1-40 peptide induces the expression of MT1-MMP and MMP-2 in VSMCs, similarly to the Dutch mutant of Aβ1-40 (Davis et al., 1999; Jung et al., 2003). In addition, we determine the molecular mechanism of MMP-2 activation by demonstrating a MT1-MMP–dependent MMP-2 activation (Fig. 3) through the recruitment of TIMP-2 to the cell surface (Fig. 4). This is consistent with the mechanism of pro-MMP-2 activation via MT1-MMP and TIMP-2 described by Strongin et al., 1995, in which TIMP-2 bound to the MT1-MMP catalytic domain acts as a cell surface ‘receptor’ for pro-MMP-2. The bound pro-MMP-2 is then activated by an adjacent MT1-MMP molecule free of TIMP-2.

Aβ-peptides upregulated MMP activity has been associated with having either protective or deleterious effects. The protective effect, which essentially translates into a reduction of cell degeneration in the blood vessel wall, is mostly attributed to the ability of MMP to degrade the Aβ peptides (Miners et al., 2008). The deleterious role includes ECM degradation, blood–brain-barrier disruption and, possibly, an induction of hemorrhagic phenotype (Rosenberg & Navratil, 1997; Rosell et al., 2006; Hernandez-Guillamon et al., 2011). Consistent with a recent study highlighting the participation of caspase-mediated mechanisms in the pro-apoptotic effect of Aβ-peptides (Fossati et al., 2010), we showed that, in addition to the induction of MMP-2, the WT-Aβ1-40 peptide led to VSM cell death, reinforcing the possible contribution of apoptotic mechanisms in sporadic CAA. We also clearly demonstrated that there is no cause-to-effect relationship between MMP activation and apoptosis. Interestingly in VSMCs, plasmin (generated from the cleavage of circulating inactive plasminogen by urokinase Plasminogen Activator and tissue Plaminogen Activator) has been also shown to induce both MMP-2 activation and apoptosis, while MMP inhibitors failed to rescue cell death (Meilhac et al., 2003); these suggest that plasmin, but not plasmin-activated MMP, is responsible for apoptosis. Besides, our data likely oppose Jung et al. (2003) study where they suggested, using the Dutch mutant of Aβ1-40, a link between MMP-2 activity, smooth muscle cell mortality and degradation of specific matrix components. However, our result, pointing out an MMP-independent VSM cell death, is consistent with the recent study demonstrating that MMP-2 activity is not directly involved in cell toxicity induced by Aβ peptides (Hernandez-Guillamon et al., 2010).

Therefore, instead of supporting the controversy surrounding MMP-2’s role in Aβ-peptide induction of cell toxicity, we believe that this apparent contradiction indicates that the ontogenesis of the Dutch familial and sporadic forms of CAAs is different. ECM degradation and VSMC degeneration are tightly connected in the Dutch familial form while being two independent processes in sporadic forms of CAA. Indeed, MMP broad-range inhibitor GM6001 can significantly reduce the apoptosis induced by a 48 h Aβ1-40Dutch-treatment (Fig S3B), but did not have this effect on WT-Aβ1-40-treated cells (Fig. 6). A likely divergence between WT, Dutch and Arctic Aβ-peptides has already been reported regarding anti-angiogenic effects in brain endothelium (Solito et al., 2009). However, although several data argue for distinct mechanisms between WT and Dutch Aβ1-40 peptides (including ours, demonstrating that the WT-Aβ1-40- but not the Aβ1-40Dutch-induced apoptosis is inhibited by fetal calf serum, see Fig S3C), one might be caution with comparing these two peptide effects as they are differently solubilized and display distinct secondary structures and oligomerization/fibrillization kinetics (Solito et al., 2009; Fossati et al., 2010).

It is not clear whether or not WT-Aβ peptide-induced MMP activity has a beneficial role (protecting the cells by degrading the WT-Aβ peptides) in our experimental conditions. If it was the case, inhibition of MMPs should have translated into an increased Aβ-induced apoptosis (Hernandez-Guillamon et al., 2010). A possible explanation for not being able to detect such an increase could be the dose of peptide used (50 μm). As a matter of fact, such a high concentration could prevent the effect of a minor Aβ-proteolytic degradation. Consistent with this, degradation of WT-Aβ peptides in a cellular context was only observed at a low dose (< 1 μm) using MMP overexpressing cells (Liao & Van Nostrand, 2010). In addition, because our cell culture condition medium does not contain plasminogen, it is also conceivable that the Aβ-dependent MMP activation (and therefore Aβ-peptide degradation) was limited. Indeed, plasmin activates MMP in VSMCs (Galis & Khatri, 2002). Interestingly, plasmin activators (uPA and tPA) are induced by Aβ-peptides (Davis et al., 2003 and Fig. S4), whereas the expression of the tPA/uPA inhibitor, PAI-1, is decreased (Fig. S4), suggesting that Aβ-treated VSMCs could convert plasminogen into plasmin. It is important to note that plasmin is also capable of degrading Aβ peptides (Tucker et al., 2000; Jacobsen et al., 2008) similar to other proteases including neprisylin, insulin-degarding enzyme, endothelin-converting enzyme, angiotensin-converting enzyme (Miners et al., 2008). Besides, plasmin, similar to MMP-2, can have both protective and deleterious effects, whereas plasmin activity can protect Aβ-peptide-treated VSMCs from apoptosis and can also lead to cell detachment (Davis et al., 2003), the process resulting in cell death by anoïkis (Michel, 2003).

Although the expression of MMP-2 was upregulated by WT-Aβ1-40 within a 72-h time frame, no effect could be detected on MMP-9. As expected, the pro-inflammatory cytokine IL-1β triggered its expression/activity (Fig. 2). Taking into account that MMP-9 is an inflammatory marker (del Zoppo, 2010), we suggest that it is unlikely that WT-Aβ1-40 initiates VSMC inflammation directly. Interestingly, Aβ1-40 peptides (Dutch or WT) induce the expression of adhesion molecules and pro-inflammatory cytokines in human arterial endothelial cells (Suo et al., 1998) arguing for an inflammatory context present in CAA (Vukic et al., 2009). Because MMP-9 (as well as MMP-2) has been extensively involved in blood–brain barrier breakdown and intracerebral hemorrhage (Rosenberg & Navratil, 1997; Rosell et al., 2006; Hernandez-Guillamon et al., 2011), two main features of CAA, it would be of interest to re-evaluate the WT-Aβ1-40 capacity of inducing MMP activity within the inflammatory context of CAA. This could first be approached in vitro on IL-1β-treated VSMCs.

In summary, our results tend to prove that the deposition of WT-Aβ1-40 (present in sporadic CAA) induces VSMC apoptosis independently from MMP activity. One might suggest that the WT-Aβ1-40-induced MMPs (because of Aβ accumulation) should actively participate in the destruction of the blood–brain barrier and the alteration of vessel wall integrity characteristics of the CAAs. Nevertheless, in vivo data should be obtained on relevant CAA models to further support our hypothesis.

Materials and methods

Reagents

Dulbecco’s modified Eagle’s medium, type I collagen from calf skin, glutamine, penicillin, streptomycin, fatty acid-free bovine serum albumin were purchased from Sigma-Aldrich (Saint Quentin Fallavier, France). Fetal calf serum and collagenase were from Gibco BRL (Cergy Pontoise, France). Elastase, protease inhibitors, LightCycler-DNA Master Plus SYBR Green and Fugene HD transfection reagent were obtained from Roche Diagnostics (Meylan, France). Oligonucleotides were sourced from MWG Biotech AG (Courtaboeuf, France). Kits for RNA extraction (RNeasy Mini kit) were obtained from Qiagen (Courtaboeuf, France). RT-MMLV, RNAsin, Lipofectamine RNAi Max transfection reagent were obtained from Invitrogen (Cergy Pontoise, France). siRNA were from Ambion (Invitrogen, Cergy Pontoise, France) or Qiagen (Courtaboeuf, France). Nitrocellulose membranes were from Schleicher and Schuell (Dassel, Germany). ECL reagent kit was from Amersham Pharmacia Biotech (les Ulis, France). Interleukin-1β (IL-1β) was from Santa Cruz Biotechnology (Heildberg, Germany). GM6001 and Batimastat broad-range MMP inhibitors were from Millipore (Molsheim, France) and Santa cruz Biotechnology, respectively. The FRET MMPs substrate Mca-Arg-Pro-Lys-Pro-Tyr-Ala-Nva-Trp-Met-Lys(Dnp)-NH2 was purchased from Bachem. Ethidium Homodimer III (ETHD-III) DNA dye was from Promocell (Heildberg, Germany).

Wild-type amyloid beta 1-40 (WT-Aβ1-40) (DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVV), Dutch mutant (E22Q) amyloid beta peptide (Aβ1-40 Dutch) (DAEFRHDSGYEVHHQKLVFFAQDVGSNKGAIIGLMVGGVV), or control inverted 40-1 (Aβ40-1) (VVGGVMLGIIAGKNSGVDEAFFVLKQHHVEYGSDHRFEAD) peptides were synthesized by GeneCust (Dudelange, Luxembourg) and analyzed by HPLC and mass spectrometry with a purity ranged between 94% and 98%. WT-Aβ1-40 and scrambled Aβ40-1 peptides were resuspended in ultrapure water at 2 mm and immediately frozen. WT-Aβ1-40 and scrambled Aβ40-1 peptides were directly added to culture medium at 50 μm. Aβ1-40 Dutch was resuspended at 1 mm in 1,1,1,3,3,3,-hexafluoro-2-propanol (Sigma-Aldrich) and incubated at room temperature until obtaining a clear solution. Aβ1-40 Dutch peptides were aliquoted and allowed to evaporate over night in a fume hood. Dried peptide films were stored at −20 °C. Prior to use, Aβ1-40 Dutch peptides films were resuspended at 5 mm into DMSO (Sigma-Aldrich) and diluted to 1 mm with PBS before added to cell culture at 50 μm.

Isolation and culture of smooth muscle cells

Rat vascular smooth muscle cells (VSMC) were isolated by enzymatic digestion of aortic media from male Wistar rats and cultured as described previously (Clement et al., 2006). All experiments were performed on cells at passages ranging from two to six. Twenty-four hours before any treatment, confluent cells were made quiescent by culturing them in a serum-free medium. All experiments were performed in serum-free medium. Of note, basal cell death 72 h after serum starvation is very similar to that of measured in presence of 10% fetal calf serum within this time frame (approximately 10%). Conditions of cell treatment are as indicated in the figure legends. All incubations were performed at 37 °C, 5% CO2. Cells were exposed to peptides diluted in culture medium at 50 μm for 24–72 h.

Real-time quantitative reverse transcription–polymerase chain reaction (RT–PCR) assays

Total RNA was extracted from VSMCs using the RNeasy kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Concentrations were determined spectrophotometrically. After annealing oligodT (1 μm) to template RNAs (0.5 μg) at 70 °C for 5 min, primer extension was initiated by adding the RT-MMLV enzyme plus 0.5 mm dNTP, 1U RNAsin, and 10 mm dithiothreitol (DTT) and carried out for 45 min at 37 °C. Quantitative PCR was performed using the LightCycler LC480 (Roche Diagnostics). The PCR mix included 5 μL of each reverse transcriptase (diluted 1:25) and 300 nm of each primer in 1 × LightCycler-DNA SYBR Green 1 Master Mix. Specific primers for complementary DNA (cDNA) were chosen with the LightCycler Probe Design2 program according to European Molecular Biology Laboratory accession numbers. The forward and reverse primers used to selectively amplify the cDNA encoding rat hypoxanthine phosphoribosyltransferase (HPRT), MMP-2, MMP-3, MMP-9, MMP-13, MT1-MMP and TIMP-2 are presented in Table 1.

Table 1.   Summary of PCR primers
Target geneForward primer (5′-3′)Reverse primer (5′-3′)Accession number
  1. HPRT, hypoxanthine phosphoribosyltransferase; MMP, matrix metalloprotease.

HPRTaggacctctcgaagtgtatccctgaagtgctcattataNM_012583.2
MMP-2atcgcccatcatcaagttccatgttctcgatggtgttcNM_031054.2
MMP-3atgtagatggtattcaatccctctataaagcagagctacacattNM_133523.2
MMP-9ttcgagggagcgtcctatttagtggtgcaggcagagtaNM_031055.1
MMP-13tggtcttctggcacacggagtggtccagaccgNM_133530.1
MT1-MMPtgggaaggaatccctgagtctctaccttcagcttctggttNM_031056.1
TIMP-2gacctgacaaggacatcgaatcccagggcacaataaagtNM_021989.2

The PCRs were performed using the following thermal settings: denaturation and enzyme activation at 95 °C for 5 min, with cycles at 95 °C for 10 s, 60 °C for 15 s, and 72 °C for 10 s. Amplification was followed up online, and the PCRs were stopped after the logarithmic phase. Melting curve analyses were also performed after PCR to check the reaction specificity. Controls and water blanks were included in each run; they were negative in all cases.

Real-time quantitative PCR data represent the amount of each target messenger RNA (mRNA) relative to the amount of a housekeeping reference gene mRNA, the HPRT, estimated in the logarithmic phase of the PCR. Serial dilutions were used to determine the fit coefficients of the relative standard curve. When the PCR target efficiencies were similar, individual cultures could be compared. If not, an internal calibrator was used.

Plasmid preparation and transient transfections

Expression plasmid for MT1-MMP mutant E240A was made as described in Rozanov et al., 2001. For plasmid expression, cells were transfected at 70–80% confluency by adding 10 μg DNA and 15 μL Fugene HD (ratio 2:3) per 10-cm dish according to the manufacturer’s instructions (Roche Diagnostics). Plasmids were inert MT1-MMP E240A mutant or control (pCDNA3, invitrogen). Twenty-four to 48 h post-transfection, cells were starved overnight and treated as indicated in the figure legends.

siRNA were transfected with Lipofectamine RNAiMAX reagent (Invitrogen) according to the manufacturer’s instructions. Briefly, 50% confluent cells were cultured for 4 h in complete medium without antibiotics and transfected by siRNA (5 nm final for 5 μL of reagent per well in 6-wells plate) for 24–48 h. After 12-h serum starvation, the cells were treated as indicated in the figure legends. The siRNA forward and reverse strand sequences are, respectively, 5′-CCAUCGAGACCAUGCGGAATT-3′ and 5′-UUCCGCAUGGUCUCGAUGGTG-3′ for MMP-2, 5′-CCAGUGUUUCAUUUGCCUATT3′ and 5′- UAGGCAAAUGAAACACUGGTT-3′ for TIMP-2; 5′-CUUCUACAAAGGGAACAAATT-3′ and 5′-UUUGUUCCCUUUGUAGAAGTA-3′ for TIMP-2.

Protein extraction and western blot analysis

Cells were washed with ice-cold phosphate-buffered saline (PBS) and resuspended in lysis buffer (20 mm Tris–HCl, pH 7.6, 150 mm NaCl, 1 mm EDTA, 1% Triton, 1% sodium deoxycholate) plus Complete Protease Inhibitor mixture (Roche Diagnostics), for 10 min on ice. The lysate was centrifuged at 13 000 g for 10 min at 4 °C, and the supernatant (cell lysate) was collected. Cell lysates (20–40 μg of protein) were analyzed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis on 12% resolving gels or 4–12% gradient gels (Invitrogen NuPAGE system) followed by transfer to nitrocellulose membranes. After blocking 1 h at room temperature with a Tris-buffered saline containing 0.1% Tween 20 (TBS-T) (20 mm Tris–HCl, pH 7.5, 100 mm NaCl, 0.1% Tween 20) and 5% fat-free milk or bovine serum albumin, membranes were incubated overnight with primary antibodies in TBS-T with 5% of milk or 5% bovine serum albumin, at 4 °C, washed in TBS-T, and incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies for 1 h at room temperature. Signals were detected with the ECL detection system and exposed to Fujifilm LAS-300 (Fujifilm Medical Systems, Stamford, CT, USA). Equal protein loading and transfer efficiency were determined by Ponceau Red staining after electrotransfer and GAPDH detection. The densitometry patterns were analyzed with ImageGauge software (Science Lab 2004; Fujifilm) and normalized to the density of GAPDH.

Antibodies

Anti-MMP-2 (mouse monoclonal; Santa Cruz), Anti-TIMP-2 (rabbit monoclonal; Cell signaling, Ozyme, St Quentin Yvelines, France), Anti MT1-MMP (mouse monoclonal; Millipore), anti-GAPDH (goat polyclonal; Sigma-Aldrich) were purchased. HRP-coupled secondary antibodies were from P.A.R.I.S Ltd, Compiègne, France.

Determination of MMP-2/-9 activity by gelatin zymography assay

Zymography of conditioned media samples was performed as previously described (Delbosc et al., 2008). Briefly, equal volume of sample issued from conditioned medium was loaded onto an 10% polyacrylamide gel containing 1% gelatin. After electrophoresis, gels were washed in a 2.5% Triton X-100 solution (2 × 30 min), incubated for 20 h in buffer containing 50 mm Tris and 2.5 mm CaCl2, stained with Coomassie Blue, and destained in methanol/acetic acid buffer until the bands could be detected.

Apoptosis detection

Cell death was determined as the percentage of hypodiploid nuclei (HN) assessed by propidium iodide staining (Mateo et al., 2007). Briefly, cells were resuspended in a hypotonic solution containing 0.1% sodium citrate, 0.1% Triton X-100 (Sigma-Aldrich), and 50 μg mL−1 propidium iodide (Sigma). Cell nuclei were analyzed on EPICS XL instrument using Expo32 software (Becton Coulter). The percentage of induced apoptosis was calculated as 100 × (percentage of experimental HN−percentage of spontaneous HN)/(100−% of spontaneous HN). Cell death was also determined by fluorescent ethidium homodimer (EthD-III) assay that produces a bright red fluorescence in dying cells. VSMCs were plated in black with clear bottom 96-well microplates at 10 000 cells/well 2 days prior treatment. After 48–72 h of treatment, the medium was removed and replaced by a volume of 100 μL per well of PBS containing 5 μm of ethidium homodimer III (EthD-III). After 30 min incubation, fluorescence was evaluated on a microplate reader (Infinite F500; Tecan France S.A.S, Lyon, France) at ex/em 520 nm/630 nm. One hundred percent dead cells were determined by measuring the fluorescence emitted by an absolute ethanol-treated well. Percentage of cell death was determined as 100 × (Fluorescence of experimental condition−background fluorescence)/(Fluorescence of ethanol-treated cells−background fluorescence). Broad-range MMP inhibitors GM6001 (25 μm) or Batimastat (2.5 μm) were added simultaneously to Aβ-peptide and renewed after 24 h.

Immunocytochemistry and microscopy

Vascular smooth muscle cells were seeded directly onto glass coverslips placed in 24-well culture plates. After treatment (always indicated in figure legends), cells were washed twice with PBS, fixed in 20 min at room temperature in 4% paraformaldehyde (neutralized with 50 mm of NH4Cl for 10 min), permeabilized in PBS+0.2% Triton X100, and incubated for 1 h in PBS+10% FCS with TIMP-2 monoclonal antibody. The incubation with DyLight 488-Conjugated Goat Anti-Rabbit secondary antibody was performed during 1 h and Hoechst staining for 5 min. Coverslips were mounted with Dako fluorescent mounting medium (Dako, Carpinteria, CA, USA). Cells were examined by phase contrast or fluorescent microscopy using a Nikon Diaphoto 300 microscope equipped with a mercury lamp (Nikon, Tokyo, Japan).

Statistical analysis

The data are reported as the mean ± SD. The numbers of independent experiments are reported in figure legends. Values were compared between groups with the Welch’s unpaired, corrected t-test.

Acknowledgments

We thank Dr. Xavier Houard for helpful discussion and advice and Dr. Alex Strongin for providing reagents. This project has been financially supported by ‘Pierre Fabre Innovation’.

Authors contribution

Isabelle Limon had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. All the authors approved the final manuscript version to be published. Régis Blaise, Véronique Mateo, Vladislav Golubkov, and Isabelle Limon contributed to the study conception and design. Régis Blaise, Véronique Mateo, Clotilde Rouxel, François Zaccarini, Martine Glorian, and Vladislav Golubkov contributed to the acquisition of data. Régis Blaise, Véronique Mateo, François Zaccarini, Gilbert Béréziat, Vladislav Golubkov, and Isabelle Limon carried out the analysis and interpretation of data. Régis Blaise, Véronique Mateo, Vladislav Golubkov, and Isabelle Limon contributed to the manuscript preparation. Régis Blaise and Isabelle Limon performed the statistical analysis.

Ancillary