Short- and Long-Term Treatment of Mouse Cortical Primary Astrocytes with β-Amyloid Differentially Regulates the mRNA Expression of L-Type Calcium Channels

Abstract

Background

It is well established that reactive astrocytes express L-type calcium channels (LTCC), but their functional role is completely unknown. We have recently shown that reactive astrocytes highly express the Ca_V1.2 α₁-subunit around β-amyloid (Aβ) plaques in an Alzheimer mouse model. The aim of the present study was to explore whether Aβ peptides may regulate the mRNA expression of all LTCC subunits in primary mouse astrocytes in culture.

Methods

Confluent primary astrocytes were incubated with 10 μg/ml of human or murine Aβ or the toxic fragment Aβ_25–35 for 3 days or for 3 weeks. The LTCC subunits were determined by quantitative RT-PCR.

Results

Our data show that murine Aβ₄₂ slightly but significantly increased Ca_V1.2 and Ca_V1.3 expression when incubated for 3 days. This acute treatment with murine Aβ enhanced β₂ and β₃ mRNA levels but decreased α₂δ-2 mRNA expression. When astrocytes were incubated for 3 weeks, the levels of Ca_V1.2 α₁ were significantly decreased by the murine Aβ and the toxic fragment. As a control, the protein kinase C-ε activator DCP-LA displayed a decrease in Ca_V2.1 expression.

Conclusion

In conclusion, our data show that Aβ can differentially regulate LTCC expression in primary mouse astrocytes depending on incubation time.

Introduction

Alzheimer’s disease (AD) is the most common progressive neurodegenerative disorder in elderly people. The extracellular deposition of β-amyloid (Aβ) plaques, intraneuronal tau pathology, synaptic loss, cholinergic neuronal cell death and inflammatory processes are the major neuropathological hallmarks of AD. A possible role for Ca²⁺ dysregulation in the aging brain and in AD has been postulated [1, 2]. L-type calcium channels (LTCC) are assumed to be involved in Ca²⁺ homeostasis [3]. LTCC consist of a pore-forming α₁-subunit (Ca_V1.1, 1.2, 1.3 or 1.4), which plays a role in voltage sensing and serves as a drug-binding site. Auxiliary β- and α₂δ-subunits are responsible for membrane trafficking and the modulation of channel properties [4, 5].

Besides their occurrence in neurons, LTCC are also expressed by astrocytes [6–8]. After brain injury, glial cells become reactive and highly express glial fibrillary acidic protein (GFAP) [9]. In contrast, neuron-glial antigen 2 glia also reacts to degenerative stimuli but does not express GFAP [10]. It is well established that brain injury leads to an upregulation of Ca_V1.2 α₁-subunits in reactive GFAP-positive astrocytes [11, 12]. The expression of Ca_V1.2 and Ca_V1.3 channels in primary astrocytes is well established [8, 13–15], but their function is not completely understood. We have recently shown that Ca_V1.2 α₁-subunits are highly expressed in reactive astrocytes surrounding Aβ plaques in a transgenic amyloid precursor protein-overexpressing mouse model [12, 16]. In addition, we have recently reported that reactive astrocytes around plaques do not coexpress the auxiliary β₄-subunit [16]. The role and function of these calcium channels in astrocytes are not known. In order to investigate the role of astroglial LTCC channels, it would be necessary to explore the cellular regulation of all LTCC subunits in the astroglia. Unfortunately, no specific and immunohistochemically selective antibodies are available for most of the LTCC subunits.

The principal question arises whether astroglial Ca_V1.2 α₁-subunit expression is caused by the AD pathology (including inflammation, oxidative stress or Ca²⁺ dysregulation) or whether it is caused directly by plaque generation induced by the Aβ cascade. In the latter case, the question arises whether the plaques per se (containing aggregated Aβ peptides) or the soluble Aβ peptides alone may induce astroglial Ca_V1.2 α₁-subunit expression. And third, it is possible that Ca_V1.2 α₁-subunit expression is a consequence of the activation and reaction of the astrocytes rather than independent of plaques. In order to study these questions, we cultured primary astrocytes from the murine cortex to explore whether exogenous Aβ peptides directly regulate expression of the LTCC channel. We show that Aβ peptides moderately enhance LTCC expression after short incubation but decrease the expression after longer incubation.

Material and Methods

Animals

Male and female wild-type B6129SF2/J mice were purchased from the Jackson Laboratory (Bar Harbor, Minn., USA). All mice were housed according to standard animal care protocols, fed ad libitum with regular animal diet and maintained in a pathogen-free environment. All experiments were approved by the Austrian Ministry of Science and Research and conformed to the Austrian guidelines on animal welfare and experimentation.

Primary Astrocyte Cell Culture

Primary astrocytes were cultured as previously reported, with slight modifications [17, 18]. Postnatal-day 4 mice were rapidly sacrificed. The cortices were dissected and pooled in 1.5 ml of pre-warmed sterile 0.25% trypsin solution. With a scalpel, the tissue was cut into small pieces. After an incubation time of 15 min at 37°C, the trypsin was discarded. To obtain single cells, the remaining tissue was passed through a 70-μm cell strainer (BD Falcon; Becton Dickinson, Franklin Lakes, N.J., USA) into a tube containing 1 ml of horse serum (16050-122; Gibco, USA) and 500 μl of DNase I (10 mg/ml; Roche Applied Sciences, Vienna, Austria). After 5 min of centrifugation (1,500 rpm), the pellet was resuspended in 10 ml of glial medium (5% horse serum, 0.5% FCS in 100 ml OptiMEM; Gibco). The cells were centrifuged again for 5 min at 1,500 rpm. The pellet was then resuspended in 2 ml of glial medium and transferred onto collagen-coated 6-well plates (150 μl of cells/well). The medium was changed 24 h after preparation and the cells were cultured for 2 weeks to become confluent. To remove residual microglia, the cells were shaken (180 rpm, 37°C) overnight and then again cultured in fresh medium. Astrocytes adhere to the surface, while other cell types peel off with the supernatant. We obtained a high astroglial purity with <0.1% oligodendrocytes or microglia.

Treatment with Exogenous Stimuli

Primary confluent astrocytes were incubated for either 3 days or 3 weeks by various treatments to induce cell activation: human Aβ₄₂ (10 μg/ml; Calbiochem, San Diego, Calif., USA); murine Aβ₄₂ (10 μg/ml; Calbiochem); the toxic Aβ_25–35 fragment (10 μg/ml; A4559; Sigma, St. Louis, Mo., USA); protein kinase C (PKC)-ε activator (3 μg/ml; DCP-LA, D5318; Sigma); and cAMP analog (12 μg/ml; Cpt-cAMP, C3912; Sigma). For some experiments, Aβ_25–35 treatments were performed at an acidic pH of 6.7 or with 45 μmol/l H₂O₂ in order to enhance astrogliosis. The 3-week treatments were conducted with 1% horse serum in glial medium.

Immunohistochemistry

Immunohistochemistry was performed as described [16]. Astrocytes were fixed in cold 4% paraformaldehyde/10 mmol/l PBS (4°C, 30 min). After fixation, the cells were washed with PBS and then incubated with 0.1% Triton X-100/PBS (T-PBS) for 30 min at room temperature, shaking. Subsequently, the astrocytes were pretreated for 20 min with 5% methanol/1% H₂O₂/PBS, and after thorough rinsing, the cells were blocked with 20% horse serum/0.2% BSA/T-PBS. Finally, they were incubated for 2–3 days at 4°C with a primary antibody diluted in 0.2% BSA/T-PBS. The primary antibodies used for this study were: rabbit anti-Ca_V1.2 α₁-subunit (1:2,000; Sigma C1603); chicken anti-GFAP (1:2,000; Millipore AB5541); and rabbit polyclonal to glutamine synthetase (1:2,000; GluS; Abcam). After incubation, the cells were washed in PBS and incubated with biotinylated secondary antibodies: α-rabbit (1:200; Vector V1011) or α-chicken (1:200; Vector K0722). After 1 h of incubation, the astrocytes were washed with PBS and incubated with an avidin-biotin complex solution (Vectastain Elite ABC reagent; Vector Laboratories) for another 1 h. After washing with 50 mmol/l Tris-buffered saline, the signal was detected by using 0.5 mg/ml 3,3′-diaminobenzidine including 0.003% H₂O₂ as a substrate in Tris-buffered saline for 3–8 min. The reaction was terminated by adding Tris-buffered saline. The stained cells were rinsed in 10 mmol/l PBS, at a pH of 7.4. Control experiments for all antibodies were conducted by omitting the primary antibody. Staining was visualized with an Olympus BX61 fluorescence microscope and pictures were captured with Openlab software connected to an Apple computer. For colocalization studies, fluorescence immunohistochemistry was performed similarly as for chromogenic staining but without methanol pretreatment. After incubation with the primary antibody, the cells were washed in PBS and incubated with fluorescent Alexa 488 secondary antibodies (Invitrogen Life Technologies, Vienna, Austria). Additionally, DAPI staining was performed. After 1 h of incubation, the astrocytes were washed and ready for analysis.

RNA Isolation and Quantitative TaqMan-PCR

Quantitative RT-PCR was performed as recently described [16]. Astrocytes were rinsed quickly with PBS to remove the medium. Subsequently, by the addition of 600 μl of Buffer RLT (RNeasy Mini Kit; Qiagen, Hilden, Germany), cells were harvested with a cell scraper. Three wells from a 6-well dish were pooled. The astrocytes were disrupted by using QIAshredder columns, and total RNA was extracted utilizing the RNeasy Mini Kit. RNA concentrations were determined photometrically using BioPhotometer 6131 (Eppendorf, Hamburg, Germany). Reverse transcription was performed on 1 μg of RNA using SuperScript II Reverse Transcriptase (Invitrogen, Carlsbad, Calif., USA) and random hexamer primers (Promega, Madison, Wisc., USA). The RT mix was incubated for 60 min at 37°C. The relative abundance of different Ca_V subunit transcripts was assessed by TaqMan quantitative PCR (qRT-PCR) using a standard curve method based on PCR products of known concentration in combination with normalization using the most stable control genes, as previously described [19]. TaqMan gene expression assays specific for all high-voltage-activated Ca²⁺ channel subunits (α₁, β and α₂δ), designed to span exon-exon boundaries, were purchased from Applied Biosystems (Foster City, Calif., USA). The following assays (Applied Biosystems) were used [name (gene symbol), assay ID]: Ca_V1.1 (Cacna1s), Mm00489257_m1; Ca_V1.2 (Cacan1c), Mm00437953_m1; Ca_V1.3 (Cacna1d), Mm01209919_m1; Ca_V1.4 (Cacna1f), Mm00490443_m1; Ca_V2.1 (Cacna1a), Mm00432190_m1; Ca_V2.2 (Cacna1b), Mm00432226_m1; Ca_V2.3 (Cacna1e), Mm00494444_m1; β₁ (Cacnb1), Mm00518940_m1; β₂ (Cacnb2), Mm00659092_m1; β₃ (Cacnb3), Mm00432233_m1; β₄ (Cacnb4) Mm00521623_m1; α₂δ-1 (Cacna2d1), Mm00486607_m1; α₂δ-2 (Cacna2d2), Mm00457825_m1; α₂δ-3 (Cacna2d3), Mm00486613_m1; and α₂δ-4 (Cacna2d4), Mm01190105_m1. The endogenous control genes (Applied Biosystems) included were [name (gene symbol), assay ID]: γ-cytoplasmic actin (Actb), Mm00607939_s1; β₂-microglobulin (B2m), Mm00437762_m1; glyceraldehyde-3-phosphate dehydrogenase (Gapdh), Mm99999915_g1; hypoxanthine phosphoribosyltransferase 1 (Hprt1), Mm00446968_m1; succinate dehydrogenase complex, subunit A (Sdha), Mm01352363_m1; tata box-binding protein (Tbp), Mm00446973_m1; and transferrin receptor (Tfrc), Mm00441941_m1. qRT-PCR (50 cycles) was performed in duplicates, using 20 ng total RNA equivalents of cDNA and the specific TaqMan gene expression assay for each 20-μl reaction in TaqMan Universal PCR Master Mix (Applied Biosystems). Analyses were performed using the 7500 Fast System (Applied Biosystems). The C_t values for each Ca_V gene expression assay were recorded for each individual preparation. To allow a direct comparison between expression levels in different tissues, we normalized all experiments to Tbp, Hprt1 and Actb, which were determined to be the most stably expressed reference genes across all preparations and time points [20]. Subsequently, normalized molecule numbers were calculated for each Ca_V subunit from their respective standard curve [19].

Data Analysis and Statistics

The data were organized and analyzed using MS Excel and KaleidaGraph statistical software. Statistical significance was determined by one-way ANOVA followed by the Fisher protected least significant difference (PLSD) post hoc test, and p < 0.05 represents significance.

Results

Astroglial Cultures

Immunohistochemistry for GFAP revealed a moderate staining of confluent astrocytes, while only occasionally, large reactive strong GFAP astrocytes were visible (fig. 1a, c). Immunostaining for glutamate synthetase showed that the whole confluent layer was strongly positive (fig. 1b). The reactive GFAP-positive astrocytes were also seen by fluorescent immunostaining over DAPI-positive nuclei (fig. 1e, f). No expression of Ca_V1.2 α₁-subunit-like immunoreactivity was seen in these astroglial cultures under all tested conditions (fig. 1d).

Fig. 1.

Immunohistochemical characterization of primary murine astrocytes. GFAP staining revealed a moderate confluent layer of astrocytes, and some few highly positive reactive astrocytes (a, c). The staining for glutamine synthetase (GluS) showed a strong staining of astroglia all over the whole confluent layer (b). Nuclear DAPI (e, f) showed a confluent layer of cells. No immunostaining was evident for the Ca_V1.2 α₁ channel (d). Scale bar = 20 μm (a, c–f), 10 μm (b).

Ca_V1.2 α₁-Subunit mRNA Expression after Three Days of Incubation

Neither the PKC-ε activator DCP-LA nor the cAMP agonist affected mRNA expression of α₁-subunits (fig. 2). Human Aβ₄₂ had no effect on mRNA expression, while murine Aβ₄₂ slightly but significantly enhanced Ca_V1.2 and Ca_V1.3 α₁-subunits (fig. 2). The toxic Aβ_25–35 fragment had no effect on expression of any tested α₁-subunits (fig. 2). However, Ca_V1.3 was significantly decreased when costimulated with the toxic Aβ_25–35 fragment and H₂O₂ (fig. 2).

Fig. 2.

qRT-PCR expression of Ca_V2.1, Ca_V1.2 and Ca_V1.3 in primary astrocytes after 3 days of incubation. Confluent astroglia was incubated without (Control) or with the PKC-ε activator DCP-LA or the cAMP agonist (cpt-cAMP) or with human or murine Aβ₄₂ or the toxic fragment Aβ_25–35 for 3 days. The toxic fragment was also coincubated under acidic conditions (pH 6.7) or with H₂O₂. After 3 days, cells of three 6-well dishes were scraped and pooled. After that, these cells were further processed for qRT-PCR. Values are given as means ± SEM normalized transcript level. Statistical analysis was performed by one-way ANOVA with a subsequent Fisher PLSD post hoc test. * p < 0.05.

Expression of Auxiliary Subunits after Three Days

In order to test expression of all auxiliary subunits in the primary astrocytes, we focused on murine Aβ₄₂, which had a significant effect on Ca_V1.2 subunit expression. Among all auxiliary subunits tested, only β₂ and β₃ were significantly enhanced, while the mRNA for α₂δ-2 was significantly downregulated by murine Aβ₄₂ (fig. 3).

Fig. 3.

qRT-PCR expression of Ca_V1.2 and all auxiliary subunits (β_1–4 and α₂δ-1 to -4) in primary astrocytes. Confluent control (Control) astroglia was incubated without any exogenous stimulus for 3 days or 3 weeks. Furthermore, cells were treated with murine Aβ₄₂ for 3 days. At the end of the experiment, cells of three 6-well dishes were scraped and pooled and further processed for RTPCR. Values are given as means ± SEM normalized transcript level. Statistical analysis was performed by one-way ANOVA with a subsequent Fisher PLSD post hoc test. * p < 0.05, ** p < 0.01, *** p < 0.001.

α₁-Subunit mRNA Expression after Three Weeks of Incubation

A 3-week incubation of astroglia showed a significant increase for the mRNA of Ca_V1.2 α₁ and β₃ and α₂ δ-2 compared with a 3-day control treatment (fig. 3). Surprisingly, a chronic treatment for 3 weeks with murine Aβ₄₂ and the toxic fragment Aβ_25–35 significantly decreased Ca_V1.2 α₁-subunit expression, but not Ca_V2.1 and Ca_V1.3. The PKC-ε activator DCP-LA significantly decreased Ca_V2.1, but not Ca_V1.2 and Ca_V1.3, subunit expression in primary murine astroglia (fig. 4).

Fig. 4.

qRT-PCR expression of Ca_V2.1, Ca_V1.2 and Ca_V1.3 in primary astrocytes after 3 weeks of incubation. Confluent astroglia was incubated without (Control) or with the PKC-ε activator DCP-LA or with human or murine Aβ or the toxic fragment Aβ_25–35. After 3 weeks of incubation with reduced serum (1%), the cells of three 6-well dishes were scraped and pooled and further processed for qRT-PCR. Values are given as means ± SEM normalized transcript level. Statistical analysis was performed by one-way ANOVA with a subsequent Fisher PLSD post hoc test. * p < 0.05, ** p < 0.01, *** p < 0.001.

Discussion

In the present study we found that Aβ peptides differentially regulate LTCC mRNA expression when incubated for a short (3 days) or long period (3 weeks).

Astroglial Cultures

It is assumed that astrocytes may contribute to several CNS pathologies, and reactive astrogliosis points to damaged tissue, which can easily be detected by GFAP staining, a reliable marker for activated astrocytes [9]. In addition, glutamine synthetase is another astroglia-specific staining. The primary astroglial culture model is well established in our research group [17, 18], displaying a confluent layer with some GFAP-positive reactive astroglia, and <1% of other cells. In order to culture astroglia for a longer time period, the cells are cultured with low serum to prevent overgrowth of the cells and subsequent cell death. In order to enhance astrogliosis, in some experiments the cells were incubated at a low pH (6.7) or together with hydrogen peroxide.

Effects of PKC-ε Activator and cAMP Agonist on LTCC Subunits

It has been reported that PKC-ε overexpression increases the mRNA levels of Ca_V2.1 and Ca_V1.2 α₁-subunits in primary astrocytes [13]. In addition, a novel PKC-ε activator, DCP-LA, reduced Aβ levels [21], providing an important link to our Aβ experiments. Thus, as a control, we incubated our primary astrocytes with DCP-LA [21]. In fact, our data only showed a slight upregulation of the Ca_V1.2 subunit treated for 3 days with DCP-LA, while after 3 weeks, Ca_V2.1 was significantly downregulated. As another control, the effect of cAMP was tested, because it is well known that LTCC activity is regulated by cAMP production and subsequent activation of protein kinase A, involving β-adrenergic receptors [22, 23]. Further, increased cAMP levels in cultured astrocytes can convert flat polygonal astrocytes into process-bearing stellate astrocytes [24]. In order to increase intracellular cAMP, we used the highly membrane-permeable agonist cpt-cAMP. However, our data show that enhanced cAMP did not affect LTCC expression in primary astrocytes at all time points.

Effects of Acute (Three-Day) Aβ Treatment on α₁-Subunits

We have recently shown that the Ca_V1.2 α₁-subunit is highly expressed in GFAP-positive astrocytes surrounding Aβ plaques in an AD mouse model [16]. However, in this in vivo model, we cannot distinguish whether the peptides alone or the plaque per se may activate the calcium channels in these reactive astrocytes. Therefore, we aimed to investigate the mRNA expression levels of all LTCC subunits in primary astrocytes of postnatal mice. In the present study, we used 3 different Aβ peptides: human Aβ₄₂, murine Aβ₄₂ and a toxic Aβ_25–35 amino acid fragment. We tested human and murine peptides because 3 amino acids are different between these two species [25] and it may be possible that human Aβ does not act on primary murine astrocytes. Aβ_25–35 was used because this small fragment is the main toxic component of the Aβ peptide. Previous work has shown that Aβ_25–35 induces cytoskeletal reorganization in cultured astrocytes [26] or revealed neurotrophic and neurotoxic effects on hippocampal neurons [27]. It was also suggested that this Aβ_25–35 peptide destabilizes calcium homeostasis and renders human cortical neurons more vulnerable to excitotoxicity [28]. All peptides were used at a very high concentration of 10 μg/ml because it has been reported that primary astrocytes are markedly affected under these conditions [29, 30]. Our present data show that the addition of murine Aβ₄₂ for 3 days slightly but significantly upregulated LTCC mRNA expression but not the protein of Ca_V1.2 and Ca_V1.3 in primary astrocytes. Thus, we conclude that an acute, brief exposure to Aβ does not markedly influence LTCC mRNA expression in cultured astrocytes.

Effects on Auxiliary Subunits and Potentiation of Astrogliosis

In order to test whether the Aβ peptide activates auxiliary subunits in vitro (independent of the α₁-subunit), we performed a qRT-PCR analysis of all subunits. We could identify that β₂ and β₃ were upregulated but α₂δ-2 was downregulated after incubation with murine Aβ₄₂ for 3 days. This differential expression profile is difficult to explain, and still the signaling process in primary astrocytes is unknown. More experiments using electrophysiology with the patch clamp technique are required to further investigate this issue. In order to enhance the reactivity of primary astrocytes, we incubated the cells at a low pH (acidosis, pH 6.7) or treated the cells with a high concentration of hydrogen peroxide together with the toxic Aβ_25–35 fragment. Indeed, chronic hypoxia dramatically increased the current density of membrane channel proteins in HEK 293 cells [31]. It has also been reported that Aβ induces cell death by generation of H₂O₂ [32]. However, costimulation with acidosis did not markedly affect mRNA expression of the LTCC subunits, while H₂O₂ led only to a slight decrease in expression, suggesting that, again, another more robust mechanism may account for the calcium channel activation.

Long-Time Exposure of Primary Astrocytes with Aβ

Thus, we hypothesized that in order to enhance LTCC expression, primary astrocytes must be exposed to Aβ peptides for a longer time period. In addition, the higher Aβ concentration and longer incubation over 3 weeks at 37°C would also likely induce aggregation of Aβ peptides and provide more plaque-promoting conditions. In order to decrease dramatic proliferation, the primary astrocytes were cultured under low-serum conditions together with the Aβ peptides for 3 weeks. Unexpectedly, we observed that the Ca_V1.2 mRNA were significantly downregulated and again no Ca_V1.2 α₁ protein expression was seen. This suggests that neither soluble nor aggregated Aβ peptides alone are sufficient to induce a dramatic upregulation of the calcium channel in vitro. Thus, we conclude that more potent exogenous stimuli are necessary to activate calcium channels in reactive astrocytes in vitro, such as inflammation, oxidative stress and/or large Aβ plaques, or even an exposure for months. In fact, we have also shown in vivo that calcium channels are only upregulated at a very late stage (>11 months) in mice.

Conclusions

In conclusion, our data show that Aβ peptides differentially regulate LTCC mRNA expression in cultured primary cortical astrocytes when incubated for short and long periods. It seems likely that Aβ peptides alone are not sufficient to activate astroglial calcium channels. More robust long-lasting effects are necessary to activate Ca²⁺ expression in GFAP-positive astrocytes.