Tau assembly in inducible transfectants expressing wild-type or FTDP-17 tau

Abstract

Conditional expression systems for 4-repeat wild-type (WT) tau or the corresponding mutants V337M and R406W were established in human neuroglioma H4 cells to study the effect of tau mutations on the physicochemical properties of tau, and to develop a cellular model for the formation of filamentous tau characteristic of frontotemporal dementia with parkinsonism linked to chromosome 17 (FTDP-17) and Alzheimer's disease. Upon induction tau expression increased, reaching maximal levels at 5 to 7 days. WT tau was phosphorylated at amino acids T181, S202/T205, T231, and S396/S404. The R406W mutation decreased tau phosphorylation at each of these sites as did the V337M mutation except for S396/S404 sites that increased. Most tau in postnuclear cell lysates was recovered in the supernatant fraction after centrifugation at 200,000 x g. The amount of tau in the pellet fraction increased more in mutant transfectants compared to WT when the induction was extended beyond 5 days. This particulate tau could be partially extracted with salt, Triton X-100, or sarkosyl. Of the transfectants, R406W had the highest proportion of sarkosyl-insoluble tau by day 7. This insoluble fraction was thioflavin S-positive and contained 15- to 5-nm-wide filaments with tau immunoreactivities. The R406W filaments were more abundant than those detected in similar preparations from WT or V337M transfectants. At the light microscopy level, most tau was found with microtubules, or diffusely distributed in the cytoplasm, but none of this appeared thioflavin S-positive. The results suggest that conditional tau transfectants are in a pretangle stage making them an attractive model system for studying intracellular tangle accumulation and for testing potential therapeutic agents as inhibitors for tau aggregation.

Figure 1.

Schematic representation of human brain tau. A: Full-length 4-repeat tau with 441 amino acid residues [4R (+2, +3)]. B: Full-length 4-repeat tau with 383 amino acid residues lacking exons 2 and 3 [4R (−2, −3)]. The sites of the N-terminal inserts, MT-binding repeats, mutations, and epitopes recognized by antibodies to nonphosphorylated tau (WKS44, E1, and Tau46) and phosphorylated tau (AT270, CP-13, AT-100, AT-180, TG-3, and PHF-1) are marked.

Figure 2.

Double-immunofluorescence staining of tau transfectants. Cells were labeled with antibodies to tau (green) and tubulin (red). Transfectants expressed variable levels of tau after 3 days of induction, but by 7 days of induction most of the cells expressed tau. As the expression of tau increased the MT network becomes reorganized. Co-localization of tau and MT (orange to yellow color) was observed in cells expressing low levels of tau; but as levels of tau increased, tau was observed in MT bundles and in regions of cytosol lacking of MTs. By day 7, most of the cells have tau expression.

Figure 3.

Western blotting of postnuclear lysate from inducible transfectants. A: Time-dependent induction of tau expression: lysates obtained 0, 2, 3, 5, and 7 days after tetracycline-off induction were subjected to Western blotting using WKS44, an antibody to nonphosphorylated tau epitopes. Recombinant 4R (−2, −3) tau (rT) and molecular weight standards (not shown) were used as references. The lysates from WT transfectants contain tau species of 68, 64, 61, 58, 56 kd, and lower molecular weight. tau species larger than the recombinant standard (56 kd) in size are regarded as intact tau with posttranslational modifications. B: Phosphorylation of lysates from transfectants with a 5-day induced expression of WT (lanes 2 and 3), VM (lanes 4 and 5), and RW tau (lanes 6 and 7) were probed with WKS44, PHF1, CP13, AT180, and AT270. Four-repeat (2-, 3-) recombinant tau was included as a reference. The location of 68-kd and 56-kd regions are marked. All samples displayed phosphorylated tau immunoreactivities. However, the labeling pattern and the relative levels of different phospho-tau species were different between the WT and mutant tau (Table 1) ▶ .

Figure 4.

Immunofluorescence staining of phosphorylated tau in inducible transfectants. WT, V337M, and R406W transfectants after a 5-day induction were labeled with WKS44 (recognizes nonphosphorylated tau) and PHF1, CP13, or AT180 (recognizes phosphorylated tau). Many cells were intensely labeled with WKS44, but immunoreactivity with PHF1, CP13, or AT180 antibodies to phosphorylated epitopes varied greatly (Table 2) ▶ . Some of these images were adjusted to view the CP13 and AT180 better.

Figure 5.

Western blotting of fractionated lysates with WKS44. Lysates were extracted from transfectants with 7 days of induction under MT depolymerizing conditions. The extraction was performed with lysis buffer alone (Cont), supplemented with 500 mmol/L NaCl (A), 1% Triton X-100 (B), or 1% Triton X-100 plus 0.1% SDS and 0.5% deoxycholate (Mix-Det) (C). Supernatant (S) and pellet (P) fractions were separated by high-speed centrifugation. All pellets were dissolved in sample buffer in a volume equivalent to one fourth of that of the lysates. Analyses of multiple blots from six separate preparations demonstrated that mutant transfectants contain a higher proportion of tau in the pellet fraction than WT. Multiple blots from three separate preparations were analyzed. tau in R406W transfectants were the least susceptible to each extraction followed by WT and V337M.

Figure 6.

A: Western blotting of samples obtained from extraction of the pellets derived from Figure 5 ▶ (Cont, P) with lysis buffer supplemented with 500 mmol/L of NaCl, 10% sucrose, and 1% sarkosyl. The sarkosyl-soluble (S) and -insoluble pellet (P) fractions were separated by centrifugation at high speed. The pellets were resuspended in sample buffer at a volume equivalent to one half the volume of the extracts. Sarkosyl-insoluble tau was detected in transfectants after 7 days of induction. Such insoluble tau were more abundant in R406W transfectants and were also detected in cells replated after 8 days of induction and maintained for another 7 days in the absence of tetracycline. B: Thioflavin S binding to sarkosyl-insoluble preparations of H4 transfectants. To control for background fluorescence in noninduced (NI) transfectants, fluorescence intensities were normalized to NI signals. Experiments were repeated in triplicate. WT preparations had thioflavin S signals of 1.7 ± 0.1 × NI signals whereas V337M and R406W were 1.7 ± 0.3 × NI and 2.1 ± 0.2 × NI, respectively. The increase in R406W compared to WT was found to be significant (P < 0.02).

Figure 7.

Immunogold labeling of sarkosyl-insoluble preparations enriched with filaments. Sarkosyl-insoluble samples were separated on a sucrose step gradient. Samples collected from the 1.5 mol/L/2.0 mol/L interface were adsorbed onto electron microscopy grids and immunogold labeled with Tau46, a mouse monoclonal against the carboxy-terminus of tau (10-nm gold) and E1, a rabbit polyclonal antibody to the amino-terminus of tau (5-nm gold). More filaments were detected in R406W (C–D) than WT (A) or V337M (B) samples. These filaments are not uniform in diameter and do not display morphology characteristic of paired helical filaments. Scale bar, 195 nm. The inset in D displays the boxed tau filament with an additional ×2.2 of magnification.