Fig 1.
The APPNLI responder transgene construct.
The 695 amino acid-long amyloid precursor protein cDNA harboring the Swedish mutation (APPNL695) was inserted into the XhoI site of MoPrP.Xho fragment, which was further excised at two XbaI sites. The resulting fragment of prnp.APPNL was cloned into the unique XbaI site in the inducible expression vector pTRE. The London mutation (V717I) was further introduced into the pTRE.prnp.APPNL plasmid using site-directed mutagenesis.
Fig 2.
Expression and suppression of transgenic APP in rTg9191 mice.
(A) Bigenic activator-repsonder system. rTg9191 mice employ a bigenic system in which a calcium-calmodulin kinase IIα (CaMKIIα) protomer drives constitutive expression of the tetracycline-controlled transactivator (tTA) gene, and a responder transgene for human APP695 containing the Swedish and London mutations (APPNLI) is under control of the tetracycline response element (tetO). Regulatable expression of the APP transgene in the rTg9191 line is under the control of doxycycline (DOX). In the absence of DOX, tTA binds the tetO promoter and APPNLI is expressed; in the presence of DOX, the tTA-tetO interaction is blocked, and expression of APPNLI is suppressed. (B-C) Expression of APPNLI. (B) Representative immunoblot probed with monoclonal antibody 22C11, which recognizes both mouse and human APP; numbers above the blot show amounts of protein loaded in each lane. (C) Quantification. Thirty-five μg of protein from brains of of 2-month-old non-transgenic (neg) mice is required to produce the same APP signal as 7 μg of protein from age-matched rTg9191 littermates, indicating that transgenic mice have 5 times more APP (mouse + human) than non-transgenic mice. Therefore, rTg9191 mice express 4 times more APPNLI relative to mouse APP. DLU, densitometric light unit. (D-E) Suppression of APPNLI expression. (D) Representative immunoblot using monoclonal antibody 6E10, which recognizes human Aβ1–16; 10 μg of protein was loaded in each lane. Alpha-tubulin served as the loading control. 8Mon and 10Mon: 8- and 10-month-old rTg9191 mice without DOX treatment; 8-10Moff: 10-month-old rTg9191 mice, treated with DOX from 8 to 10 months of age. (E) Quantification. Administration of DOX (200 ppm in chow) to rTg9191 mice decreased levels of APPNLI by 87%. *** p < 0.0001, one-way ANOVA followed by Fisher’s post hoc analysis.
Fig 3.
Regional expression pattern of the APP transgene in rTg9191 mice.
The regional pattern of APPNLI expression in four distinct anatomical structures (cerebral cortex (Ctx.), hippocampus (Hpp.), olfactory bulb (Olf.), and cerebellum (Cbm.)) of brain was analyzed using mouse monoclonal antibody LN27, which specifically recognizes human APP, and 6E10. The APP transgene was expressed in cerebral cortex and hippocampus with a minor portion in olfactory bulb; however, no expression was observed in cerebellum. No immunoreactivity using these human-specific antibodies was seen in non-transgenic littermates (neg). Alpha-tubulin served as the loading control. Representative blots show the APPNLI expression pattern of female mice, and similar results were found in male mice.
Fig 4.
Age-related Aβ plaque progression in rTg9191 mice.
(A) A representative sagittal section of brain. Black rectangles indicate the regions of cerebral cortex and hippocampal formation in which Aβ plaques are shown in (B-E). (B-E) Representative photomicrographs showing age-dependent progression of Aβ plaques in female mice, visualized using 6E10 (B); 4G8, directed against a mid-region of Aβ (C); 139–5, an Aβx-40-specific antibody (D); and 1-11-3, an Aβx-42-specific antibody (E). Upper panels, cerebral cortex; lower panels, hippocampus. Scale bars: 100 μm (upper panels), 200 μm (lower panels). TTA, mice expressing only the tetracycline transactivator. (F) Quantification of 4G8-immunoreactive Aβ plaque load at various ages.
Fig 5.
Age-related production of Aβ proteins in water-soluble, detergent-soluble, and detergent-insoluble fractions of brain extracts from rTg9191 mice.
The levels of total Aβ38, Aβ40, and Aβ42 proteins in water-soluble (A), detergent-soluble (B) and detergent-insoluble (C) fractions of brain extracts of rTg9191 mice at 4, 12, 21, and 24 months of age were measured using ELISA.
Fig 6.
rTg9191 mice produce Aβ dimers in an age-dependent manner, but lack Aβ*56.
(A-B) Production of Aβ dimers. (A) A representative immunoblot (upper panel) shows levels of Aβ dimers in young, mid-age, and old mice; α-tubulin served as the loading control (lower panel). (B) Quantification. rTg9191 mice exhibit an age-dependent progression in levels of Aβ dimers. # p < 0.05, *** p < 0.0001 (compared to 26M), one-way ANOVA, followed by Fisher’s post hoc analysis. (C-D) rTg9191 mice lack Aβ*56. (C) Representative immunoblot shows levels of Aβ*56 in transgenic mouse lines rTg9191, TgArc6, and hAPP-J20. Although a faint band at ~56kDa was seen occasionally in extracts from rTg9191 mice, the intensity of this band was comparable to that seen in some samples from non-transgenic mice. As monoclonal antibody 6E10 recognizes human, but not mouse Aβ, this faint band represents non-specific background noise, and not a true Aβ signal. (D) Quantification. The intensity of the ~56 kDa band in rTg9191 mice (4, 12, 21, 24 and 26M of age) is comparable to that of non-transgenic littermates (neg, 26M) and TgArc6 (4M), but is significantly lower than that of hAPP-J20 (4M). *** p < 0.0001, each group compared to hAPP-J20, one-way ANOVA, followed by Fisher’s post hoc analysis. Neg, non-transgenic littermates of rTg9191 mice; pos, mice that express APP transgenes. Samples from both genders were loaded onto the representative blots, aligning in the order of left to right for (A): F, M, F, F, M, M, F, M, F, M, F, F (M, male; F, female), and for (C): F, F, M, M, F, M, F, M, F, M, F, F, M, F, M, F.
Fig 7.
rTg9191 mice age-dependently produce soluble Aβ oligomers immunoreactive to OC but not to A11 antibodies.
(A-B) Soluble OC-immunoreactive Aβ oligomers accumulate in brains of rTg9191 mice in an age-dependent manner. (A) Upper panels, representative dot blots showing levels of protein oligomers detected by polyclonal OC antibodies in water-soluble brain extracts from rTg9191 and Tg2576 mice and AD patients. Upper lane, brain extracts and synthetic aggregates; lower lane, brain extracts and synthetic Aβ aggregates immunodepleted of Aβ, using an array of antibodies (6E10, 4G8, 139–5, and 1-11-3). OC immunoreactivity disappears when samples are immunodepleted of Aβ, suggesting that the OC signals in the brain extracts arise from Aβ assemblies. Each dot contains either 0.5 μg of protein extracts or 2 ng of synthetic aggregates. Synthetic soluble Aβ aggregates with in-register parallel β-sheets were used as a positive control. Alpha-tubulin served as the loading control (lower panel) for both the untreated (upper lane) and Aβ-immunodepleted materials (lower lane). (B) Quantification. rTg9191 mice show an age-dependent increase in levels of OC immunoreactivity. At 21 months of age, rTg9191 mice have a comparable level of OC-immunoreactive signals to AD patients. ** p < 0.001, *** p < 0.0001, each group compared to AD, one-way ANOVA, followed by Fisher’s post hoc analysis. (C-D) rTg9191 mice lack A11-immunoreactive Aβ oligomers in brain. (C) Upper panels, representative dot blots showing levels of protein oligomers detected by polyclonal A11 antibodies in water-soluble brain extracts from rTg9191 and Tg2576 mice and AD patients. Upper lane, brain extracts and synthetic oligomers; lower lane, brain extracts and synthetic Aβ oligomers immunodepleted of Aβ, using an array of antibodies (6E10, 4G8, 139–5, and 1-11-3). Each dot contains either 1 μg of protein extracts or 1 μg of synthetic oligomers. Alpha-tubulin served as the loading control (lower panel) for both the untreated (upper lane) and Aβ-immunodepleted materials (lower lane). (D) Quantification. A11 immunoreactivities in brains of rTg9191 mice show no age-dependent change and are comparable to those of young and old non-transgenic littermates. Aged Tg2576 mice and AD patients, however, show significantly higher A11 immunoreactivity. *** p < 0.0001, each group compared to AD, one-way ANOVA, followed by Fisher’s post hoc analysis. nTg, non-transgenic littermates of rTg9191 or Tg2576 mice; APP/TTA, rTg9191 mice that harbor both the TTA activator and APP responder transgenes and therefore express APPNLI; APP, Tg2576 mice that express the APP transgene. For each detection antibody, all panels in an image came from a single representative blot; blots are shown segmented for clarity. Samples from both genders were loaded onto the representative blots, aligning in the order of left to right for both (A) and (C): F, M, F, M, F, M, F, M, F, F, M, F, F, M, F (M, male; F, female).
Fig 8.
TTA expression results in reduced forebrain weight and dentate-gyrus size.
(A) Weights of rTg9191 mice (APP/TTA), littermates harboring only the activator gene (TTA), only the responder gene (APP), and neither gene (non-Tg). There were no genotype-related differences in body weight in young and mid-aged mice. Aged rTg9191 mice, however, had lower body weights compared to their littermates. (B) rTg9191 mice and TTA littermates have lower forebrain weights than their APP and non-Tg littermates at all ages studied. The numbers of mice examined are shown for each genotype. # p < 0.05, * p < 0.01, ** p < 0.001, *** p < 0.0001, two-way ANOVA followed by Fisher’s post hoc analysis. The percentage of female mice in the genotype of APP/TTA, TTA, APP, and non-Tg is 50%, 38%, 67%, and 55%, respectively for the 2–6 month-old; 51%, 35%, 48%, and 50%, respectively for the 13–17 month-old; 46%, 61%, 47%, and 51%, respectively for the 24–27 month-old. Chi square/Fisher exact tests showed no significant difference in gender distribution between genotypes. (C-F) Representative photomicrographs showing hematoxylin and eosin staining of the hippocampal regions of 16.5-month-old rTg9191 mice and their age-matched littermates. Sections at ~1.20 mm lateral from the midline were used. The sizes of dentate gyri of rTg9191 (F) and TTA (E) mice are decreased compared to those of APP (D) and non-Tg (C) littermates. Scale bar: 200 μm, applies to C-F. Representative photomicrographs show hippocampus hematoxylin and eosin staining of female mice, and similar results were observed in male mice.
Fig 9.
Plaque-associated neuroinflammation and abnormal neuronal architecture in rTg9191 mice.
(A-I) rTg9191 mice show reactive gliosis in the vicinity of dense-core plaques. Brain sections from rTg9191 mice at 24 months of age (B, E, H), their age-matched non-transgenic littermates (A, D, G), and age-matched Tg2576 mice (C, F, I) were stained with antibodies directed against the astroglial marker S100β (A-C), a monoclonal antibody directed against the microglial marker ionized calcium-binding adaptor molecule 1 (Iba1) (D-F), and an antibody directed against the astrocytic marker glial fibrillary acidic protein (GFAP) (G-I). Astrocytes and activated microglial cells and reside near dense-core plaques visualized using Congo red (pink). Scale bar in I, 25 μm, applies to A-I. (J-K) rTg9191 mice exhibited abnormal neuronal architecture around plaques. Thioflavin S (green) was used to visualize plaques and monoclonal antibody SMI-312 was used to visualize axons (red). (J) No plaques were detected in age-matched non-transgenic littermates of rTg9191 mice, and neuronal morphologies were normal. (K) Plaques are surrounded by swollen, dystrophic axons (arrowheads) and curvy, distorted axonal processes (arrows) in brains of rTg9191 mice. Scale bar in K, 50 μm, applies to J and K. Representative photomicrographs show neuroinflammation and neuronal architecture of female mice, and similar results were found in male mice.
Fig 10.
Plaque-associated tau pathology in rTg9191 mice.
Brain sections from rTg9191 mice of 24 months of age (B, F, J, N, R, V, Z), their age-matched non-transgenic littermates (A, E, I, M, Q, U, Y), 23-month-old Tg2576 mice (C, G, K, O, S, W, AA) and 15-month-old rTg4510 mice (D, H, L, P, T, X, AB) were stained with a variety of antibodies directed against pathological conformation- and phosphorylation-dependent epitopes of tau: AT8 (A-D), CP13 (E-H), PG5 (I-L), PHF-1 (M-P), Alz50 (Q-T), MC1 (U-X) and TG-3 (Y-AB). Representative photomicrographs showed that hyperphosphorylated and/or misfolded tau proteins accumulated (brown puncta) around dense-core plaques visualized using Congo red (pink). Neuronal staining (brown) in rTg4510 mice served as positive control. Scale bars: 20 μm. Images in the same row have the same magnification. Representative photomicrographs show tau pathology of female mice, and similar results were found in male mice.