Effect of calcium ions on the aggregation of highly phosphorylated tau

Abstract

Tau is typically an axonal protein, but in neurons of brains affected by Alzheimer's disease (AD), aggregation of hyperphosphorylated tau in the somatodendritic compartment causes neuronal death. We have previously demonstrated that tau mRNA is transported within dendrites and undergoes immediate translation and hyperphosphorylation of AD epitopes in response to NMDA receptor stimulation. Although this explains the emergence of hyperphosphorylated tau in dendrites, the relationship between tau hyperphosphorylation and aggregation is not well understood. In this study, we found that recombinant highly phosphorylated tau purified from NG108-15 rodent neuroblastoma/glioma cells transfected with both tau and GSK3β expression vectors bound calcium ions and formed sarkosyl-insoluble aggregates. In addition, thioflavin T analysis revealed that this highly phosphorylated tau tended to aggregate on its own, further facilitated by calcium ions. When NG108-15 cells expressing the highly phosphorylated tau were treated with calcium ionophore, sarkosyl-insoluble tau was generated. Interestingly, these cells exhibited resistance to both calcium ionophore-induced cytotoxicity and glutamate-induced excitotoxicity. We further found that sarkosyl-insoluble phosphorylated tau was increased in cultured hippocampal neurons due to glutamate-induced hyperactivity. Our data suggest that hyperphosphorylated tau synthesized in response to NMDA receptor stimulation contributes to regulation of neuronal activity by binding calcium ions, but that this calcium binding may cause tau to adopt an aggregated form.

Keywords: Aggregation; Calcium; Excitotoxicity; Highly phosphorylation; Tau.

Fig. 1

Highly phosphorylated tau binds calcium ions. (A) Synthesis of recombinant highly phosphorylated tau in NG108-15 cells. Cells co-transfected with 6 × His-tagged 2N4R-tau expression vector and GSK3β expression vector were immunostained with anti-tau (Tau-5), anti-tau pSer199, anti-tau pSer396 or anti-tau pThr231 (AT180) antibody. Scale bar: 10 μm. (B) Western blot analysis of highly phosphorylated tau synthesized in NG108-15 cells. Recombinant highly phosphorylated tau was purified using Ni-NTA agarose from CIAP-treated or -untreated cytosolic fractions and analyzed by Western blotting using anti-tau (Tau-5), anti-tau pSer199, pSer396 antibody or anti-tau pThr231 (AT180) antibody. (C) Analysis of the calcium binding property of highly phosphorylated tau. CIAP-treated and -untreated highly phosphorylated tau-trapped beads were incubated in CaCl₂ solution, and the relative amounts of calcium ions in each supernatant were compared. The data represent the mean and standard error obtained from three independent experiments. ∗∗P < 0.01 (Student's t-test).

Fig. 2

Formation of sarkosyl-insoluble aggregates of highly phosphorylated tau upon calcium binding. (A and B) CIAP-treated and -untreated highly phosphorylated tau (5 μM each) were incubated in 1 mM CaCl₂ solution. After centrifugation, the resulting precipitate was incubated in 1 % Sarkosyl solution, separated into Sarkosyl-soluble fraction (A) and -insoluble fraction (B) by centrifugation again, and Western blot analysis was performed using anti-tau antibody Tau-5. (C and D) Calcium-dependent sarkosyl-insoluble highly phosphorylated tau aggregation. Highly phosphorylated tau (5 μM) incubated with 0, 1 or 10 mM CaCl₂ was separated into a sarkosyl-soluble supernatant (C) and a sarkosyl-insoluble precipitate (D) as described above, and analyzed by Western blotting using anti-tau pSer199 antibody. (E) Thioflavin T-binding of the highly phosphorylated tau aggregates in the presence of calcium. A mixture of recombinant highly phosphorylated tau (2 μM) and CaCl₂ (0, 0.1 or 1 mM) was incubated in the presence of thioflavin T (20 μM), and the change in fluorescence intensity was measured at the indicated time point. Data represent the mean and standard error of three independent experiments with 10 replicates for each calcium concentration. ∗P < 0.05 versus control (one-way ANOVA, followed by Tukey-Kramer post hoc test).

Fig. 3

Calcium binding by highly phosphorylated tau in living cells attenuates excitotoxicity. (A) Left panel: NG108-15 cells co-transfected with the 2N4R-tau expression vector and GSK3β expression vector were treated with 5 μM calcium ionophore A23187 for 5 h in culture medium. As a control, cells transfected with 2N4R-tau vector alone were used. Cytosolic fractions were prepared and incubated in a final concentration of 1 % sarkosyl for 1 h at 37 °C, followed by centrifugation at 21,800×g for 15 min and the precipitated aggregates were analyzed by Western blotting using an anti-tau antibody Tau-5. Right panel: Cells transfected with both the ΔRepeat-tau and GSK3β vectors were treated with A23187, the cytosolic fraction was centrifuged at 21,800×g for 15 min, and the solubility of the ΔRepeat-tau contained in the precipitate in 1 % sarkosyl solution was examined. (B) NG108-15 cells transfected with both 2N4R-tau and GSK3β expression vectors or 2N4R-tau expression vector alone were treated with A23187 as described in (A). Highly phosphorylated tau and control tau were then immunoprecipitated from the post-nuclear supernatant, and the amount of bound calcium was quantified. The amount of tau bound to the beads was then examined (Supplementary data 1) and the molecular ratio of tau to calcium was determined. The data represent the mean and standard error for three independent experiments. ∗∗P < 0.01 (Student's t-test). (C and D) The effect of highly phosphorylated tau expression on calcium ionophore-induced cytotoxicity on undifferentiated NG108-15 cells (C) or glutamate-induced excitotoxicity on differentiated NG108-15 cells (D) was examined as described in Materials and Methods, and cell viability was measured by MTT assay. Values were normalized by the survival rate of untransfected cells after treatment with A23187 or glutamate. The effects of expression of 2N4R-tau alone or GSK3β alone were also examined (each lower panel). The data represent the mean and standard error for five independent experiments. ∗∗P < 0.01 (Student's t-test).

Fig. 4

Glutamatergic stimulation-responsive increase of sarkosyl-insoluble phosphorylated tau in hippocampal neurons. (A) Differentiated hippocampal neurons were treated with 0.5 mM glutamate for 2 h and analyzed by immunocytochemistry with anti-tau (Tau-5), anti-tau pSer199, anti-tau pSer396 or anti-AT8 antibody, and anti-MAP2 antibody as a dendritic marker. Scale bar: 10 μm. Quantitative analysis of total tau or each phosphorylated tau in dendrites is shown below each of the immunocytochemical data. Relative tau protein levels (fluorescence intensity) in dendrites of glutamate-treated and untreated hippocampal neurons are shown. Approximately 30 neurons were evaluated. The signal intensities were measured using the NIH ImageJ software package and normalized to the fluorescence intensity of each corresponding DAPI-stained neuron. Data are presented as the mean and standard error. ∗∗P <0.01 (Student's t-test). (B) Sarkosyl-insoluble fractions were prepared from glutamate-treated and -untreated neurons as described in Fig. 3A, and phosphorylated tau was detected by Western blotting using anti-tau pSer199, anti-tau pSer396 or anti AT8 antibody.