Gelsolin restores A beta-induced alterations in choroid plexus epithelium.

Histologically, Alzheimer's disease (AD) is characterized by senile plaques and cerebrovascular amyloid deposits. In previous studies we demonstrated that in AD patients, amyloid-beta (A beta) peptide also accumulates in choroid plexus, and that this process is associated with mitochondrial dysfunction and epithelial cell death. However, the molecular mechanisms underlying A beta accumulation at the choroid plexus epithelium remain unclear. A beta clearance, from the brain to the blood, involves A beta carrier proteins that bind to megalin, including gelsolin, a protein produced specifically by the choroid plexus epithelial cells. In this study, we show that treatment with gelsolin reduces A beta-induced cytoskeletal disruption of blood-cerebrospinal fluid (CSF) barrier at the choroid plexus. Additionally, our results demonstrate that gelsolin plays an important role in decreasing A beta-induced cytotoxicity by inhibiting nitric oxide production and apoptotic mitochondrial changes. Taken together, these findings make gelsolin an appealing tool for the prophylactic treatment of AD.

1. Introduction

Alzheimer's disease (AD) is a neurodegenerative disorder characterized by a progressive loss of cognitive function and is associated with neuropathological hallmarks that include amyloid plaques, neurofibrillary tangles, synaptic loss, and neurodegeneration. Senile plaques mainly consist of an extracellular accumulation of the 40–42-aminoacid long peptide amyloid β (Aβ) [1, 2], although intracellular deposits of Aβ have also been reported [3, 4]. Besides accumulating in the brain parenchyma, Aβ also accumulates in choroid plexus epithelial cells [5] and in cerebrovascular walls, where it induces blood-brain barrier disruption [6-8]. Several studies have shown that Aβ1–40 and Aβ1–42 alter transmembrane and cytoplasmic tight junction proteins in brain microvessel endothelial cells, including ZO-1, which ultimately leads to disruption in the integrity of the blood-brain barrier [9, 10].

In support of these findings, recent results from our laboratory have suggested direct relationship between Aβ accumulation at the choroid plexus epithelium and the development of functional and structural dysfunctions [5, 11]. In addition, we demonstrated the existence of a link between Aβ-induced choroid plexus cell death, increased production of nitric oxide (NO), and mitochondrial dysfunction in the choroid plexus of patients with AD and amyloid precursor protein (APP)/PS1 mice [11].

The choroid plexus, which is made up of a single epithelial cell layer, is responsible for producing cerebrospinal fluid (CSF) and constitutes the blood-CSF barrier. Additionally, choroid plexus cells produce proteins involved in several processes important for normal brain function, such as prevention of Aβ fibrillization. One of such proteins is gelsolin [12], which is a potent actin-regulatory protein that controls cytoskeletal assembly and disassembly [13]. Gelsolin can be found both as an intrinsic cytoplasmic protein and as a secreted protein [14]. Besides controlling formation of cytoplasmic actin filaments, gelsolin plays an important role in apoptosis and amyloidosis. The secretory form of gelsolin is known to bind Aβ under normal physiological conditions [15], inhibit the fibrillation of Aβ, and defibrillize preformed fibrils of this peptide [16]. Some groups have suggested that gelsolin could be used in the prophylactic treatment of AD as Aβ sequestering agent [17, 18].

In this study, we hypothesize that the secreted form of gelsolin could be an effective therapeutic approach for the preservation of blood-CSF barrier integrity and function, and thus an attractive tool for the prophylactic treatment of AD. To test the hypothesis that secreted gelsolin can reduce Aβ cytotoxicity on choroid plexus epithelium, we analyzed cytoskeletal alterations, including the distribution and expression of ZO-1. In addition, we assessed Aβ-induced NO production, cell death, and mitochondrial changes in choroid plexus epithelial cells.

2. Methods

Cell Culture

Epithelial cell monolayers from P3-P5 Wistar rats were prepared as described previously [19]. Cells were grown to confluence for 5–7 days and serum starved for 2 hours. Human analog peptides corresponding to Aβ1–40, Aβ1–42, or scrambled Aβ1–42(5 μg/mL; AnaSpec, Inc.), and gelsolin, extracted from bovine plasma (5 μg/mL; Sigma), were added. Forty-eight hours after stimulation, cells were either fixed for immunocytochemical analysis or homogenized for immunoblot determination.

Antibodies

The following antibodies were used: mouse monoclonal anti-Aβ1–40 (Chemicon), rabbit polyclonal anti-Aβ1–42 (Chemicon), goat polyclonal anti-megalin (Santa Cruz Biotechnology), mouse monoclonal anti-gelsolin (Sigma), mouse monoclonal anti-pSer (Sigma), goat polyclonal anti-ZO-1 (Santa Cruz Biotechnology), mouse monoclonal anti-CoxVa (Molecular Probes), BIODIPY FL phallacidin (Invitrogen), Alexa-coupled (Molecular Probes), and HRP-conjugated (Bio-Rad).

Western-Blot Analysis

Western-blot (WB) analysis and immunoprecipitation were performed as described previously [19]. WB membranes were reblotted with the same antibody used for immunoprecipitation, and to normalize for protein load. Densitometric analysis was performed using ImageJ software (NIH). Nonimmune normal rabbit serum was used as a control for immunoprecipitation studies.

Blue Native Electrophoresis

Mitochondrial membranes were isolated according to the method described by Nijtmans et al. [20]. Enzyme activity of mitochondrial complexes from choroid plexus epithelial cell cultures was measured as described previously [11]. Gels were washed in distilled water, scanned, photographed immediately, and quantified with the aid of ImageJ software (NIH) [11].

Determination of NO Production

For NO detection, choroid plexus epithelial cell cultures were processed using the Nitric Oxide Colorimetric Assay Kit (BioVision, Inc.) as described previously [11].

Cell Death Quantification

After 48-hour incubations with Aβ1–42 and gelsolin, DNA fragmentation in choroid plexus epithelial cells undergoing apoptosis was detected with a Cell Death Detection ELISAPLUS kit (Roche) as described previously [11].

3. Results

Gelsolin Binds to Megalin and Forms a Complex with Aβ

We observed that megalin from rat choroid plexus cells binds to Aβ1–40 and to gelsolin (Figure 1(a)). Furthermore, our results indicate that both the cytoplasmic and the secreted forms of gelsolin bind to megalin (Figure 1(a), top panel). These findings were corroborated by imunoprecipitation and double immunostaining, which showed that gelsolin co-localizes with megalin and Aβ in choroid plexus epithelial cells (Figure 1(b)).

Figure 1

Gelsolin expression in choroid plexus epithelial cells. (a) Antimegalin immunoprecipitation of rat choroid plexus cell extracts, followed by blotting with respective antibodies, revealed an association between megalin, endogenous gelsolin, the exogenously added secreted gelsolin form, and the exogenously added Aβ. Immunoprecipitation with nonspecific serum showed no unspecific Aβ association. Binding of gelsolin with exogenously added Aβ was also observed. Representative blots are shown (n = 4). (b) Megalin colocalized with gelsolin and exogenously added Aβ in choroid plexus cultures. Confocal images also show gelsolin colocalization with Aβ. Scale  bars = 10 μm. IP: Immunoprecipitation; NRS: normal rabbit serum.

Gelsolin Inhibits Aβ-Induced Disruption on Choroid Plexus Epithelial Cell Cytoskeleton

Choroid plexus epithelial cells exposed to Aβ1–42 for 48 hours showed a disrupted plasma membrane pattern of ZO-1 with relocation of this protein to the cytoplasm (Figure 2(a)). Aβ1–42-induced disruption of epithelial barrier integrity was also confirmed by an increase on serine phosphorylation rate. Immunoprecipitation with anti-ZO-1, followed by immunoblotting with anti-pSer, showed an increase in serine ZO-1 phosphorylation and a reduction in ZO-1 expression in choroid plexus epithelial cells (Figure 2(b)). We found that coadministration of gelsolin restored these Aβ1–42-induced effects (Figures 2(a) and 2(b)). Moreover, this Aβ1–42-induced effect on the behavior of this tight junction protein was also accompanied by cytoskeletal disruption, as we observed the formation actin stress fibers (Figure 2(c), middle panel). The immunocytochemical analysis revealed a restoration of cytoskeletal assembly when gelsolin was added to the culture medium (Figure 2(c), right panel).

Figure 2

Secreted gelsolin inhibits Aβ-induced disruption on choroid plexus epithelial cell cytoskeleton. (a) Representative confocal images of choroid plexus confluent monolayer labeled with anti-ZO-1 antibody. Under control conditions, ZO-1 immunostaining is distributed along the plasma membrane. In contrast, after exposure to Aβ1–42 for 48 hours, a disruption of the plasma membrane pattern of ZO-1 was observed, resulting in increased cytoplasmic localization. Note the ability of gelsolin treatment to prevent this Aβ-induced alteration in ZO-1 pattern. (n = 3). (b) Aβ1–42 treatment resulted in increased serine phosphorylation of ZO-1 and decreased ZO-1 expression in choroid plexus epithelial cells. Gelsolin coadministration markedly attenuated Aβ1–42 alteration in ZO-1 (n = 3). (c) BIODIPY FL phallacidin staining of choroid plexus epithelial cells showed a disruption of the actin cytoskeleton after treatment with Aβ1–42 for 48 hours, and reversion when gelsolin was simultaneously added. Magnification: ×40. Scale bars = 10 μm.

Gelsolin Inhibits Aβ-Induced NO Production and Neuronal Death in Choroid Plexus Epithelial Cells

Gelsolin blocked Aβ1–42-induced NO production by cultured choroid plexus epithelial cells after 48 hours of treatment (Figure 3(a)). In vitro, Aβ1–42-induced cell death in choroid plexus epithelial cells was also reversed after 48 hours of gelsolin administration (Figure 3(b)).

Figure 3

Secreted gelsolin expression modulates NO production and cell death in choroid plexus epithelial cells. (a) Choroid plexus epithelial cells treated with Aβ1–42 for 48 hours exhibited a significantly enhanced NO production compared with untreated cells. Secreted gelsolin coadministration completely blocked this effect (n = 3); **P < .01. (b) Increased cell death was observed in choroid plexus cell cultures 48 hours after Aβ1–42 treatment, and gelsolin addition totally reversed this toxic effect (n = 3); *P < .05. (c and d) Aβ1–42 treatment reduced mitochondrial complex IV in-gel activity in choroid plexus epithelial cells, whereas secreted gelsolin administration increased complex IV activity and reversed this decrease in Aβ1–42-induced activity. Blue native analysis of these culture samples showed altered protein expression in the mitochondrial complex IV. Representative blue native blots and quantitative histograms are shown (n = 4 per group); *P < .05.

Gelsolin Increases Mitochondrial Respiratory Chain Activity in Choroid Plexus Epithelial Cells

When gelsolin was added to the choroid plexus culture, an increased activity of complex IV was observed as compared with control cells (Figure 3(c)). Moreover, gelsolin was able to reverse the inhibitory effect of Aβ in complex IV activity (Figure 3(c)). WB performed to assess protein level alterations revealed a parallel activation in complex IV, reflected by an increase in Cox Va subunit levels (Figure 3(c)).

4. Discussion

Gelsolin, an actin-regulatory protein, exists both as an intracellular and extracellular protein [14] and is present in all nervous system cell types, including neurons [21] and choroid plexus [12]. Our findings indicate that secreted gelsolin is involved in the pathology of AD through the regulation of brain Aβ and its neurotoxic effects. Plasma gelsolin has been found to bind and reduce brain Aβ [15, 17, 18]. In the present study we had confirmed the formation of a complex between cytoplasmic and secreted gelsolin with Aβ, accordingly with previously published studies [15, 22]. Our current results extend these observations, suggesting that megalin, an endocytic receptor involved in Aβ clearance [23-27], has a functional role in the formation of this complex. We show that megalin binds Aβ/gelsolin complex in the choroid plexus epithelium, suggesting a role in clearance of Aβ from CSF to the blood. The latter is not surprising in view of megalin's ability to transport a large variety of proteins [28].

Abnormalities in cytoskeletal organization are a common feature of many neurodegenerative disorders, including AD. Interestingly, Aβ1–42-induced cytoskeletal alterations known to be associated with the proteolytic degradation of the tight junction-associated protein ZO-1 [10, 11] were reversed after secreted gelsolin treatment, with relocation from the cytoplasm to the original position in the cell membrane of choroid plexus epithelial cells. In addition, our results suggest that the secretory form of gelsolin prevents the internalization of ZO-1 by blocking serine phosphorylation, as has been also demonstrated in other studies [29-31]. The presence of structures characterized by aggregates of polymerized actin (F-actin) has been described in AD [32]. Also, F-actin levels have been shown to increase in hippocampal neurons treated with Aβ [33]. Furthermore, cortical neurons expressing the APP intracellular domain suffer from pronounced changes in the organization of the actin cytoskeleton, including destabilization of actin fibers [34]. We found that fibrillar Aβ1–42 also caused alterations in cytoskeletal actin in choroid plexus epithelial cells, as reflected by an increase in the F-actin content, and that gelsolin was able to protect against this effect. In this way, gelsolin contributes to the maintenance of the choroid plexus monolayer and the blood-CSF barrier integrity. A possible explanation for these effects could be that Aβ increases metalloproteinase 9 [11, 35, 36], which is capable of cleaving cytoplasmic gelsolin [37], thereby resulting in the destabilization of actin filaments and the disruption of tight junctions. Secreted gelsolin, which binds and sequestrates Aβ, would then be able to prevent and diminish Aβ toxic effects.

On the other hand, our results also indicate that gelsolin prevents Aβ-induced cell death and NO production from choroid plexus cell cultures. Using several models of neuronal cell death, others studies have demonstrated that cytoplasmic gelsolin has antiapoptotic properties that correlate with its dynamic actions on the cytoskeleton [38]. Indeed, gelsolin-null neurons have enhanced cell death [39] and increased vulnerability to glutamate toxicity [40]. In a previous study we reported that Aβ-induced mitochondrial dysfunction could ultimately activate a programmed cell death pathway in the choroid plexus epithelial cells [11]. In this study, we show that secreted gelsolin prevents Aβ-induced cell death by increasing enzyme activity of the respiratory chain complex IV in the choroid plexus epithelial cells. These observations are in line with other studies showing an Aβ-induced reduction of mitochondrial membrane potential by cytoplasmic gelsolin [41, 42]. The intracellular form of gelsolin is associated to the mitochondrial membrane, where it can inhibit Aβ-induced loss of mitochondrial membrane potential, cytochrome c release, and regulate voltage-dependent channels [40, 41]. Since extracellular gelsolin has also been detected in CSF, where its concentration is significantly altered in certain neurological conditions [43], this extracellular isoform of gelsolin may well reduce choroid plexus Aβ-induced pathology in a similar manner to intracellular gelsolin [17, 18].

In conclusion, our results demonstrate that secreted gelsolin can modulate Aβ-induced alterations in the blood-CSF barrier. We suggest that secreted gelsolin have a neuroprotective role against Aβ neurotoxicity. In summary, enhancement of gelsolin levels may represent a novel way to protect against Aβ neurotoxicity and, in the future, could be considered a potential therapeutic strategy for the treatment of patients with AD.