Human brain slides were provided by the brain bank of Karolinska Institutet (approval nr 2013/1301-31/2) (Supplementary Table 1). p62 immunoblotting was performed in brain homogenates from the Netherlands Brain Bank under ethical permit Dnr EPN 2011/962-31/1 and 2018/1993-32.

App^NL–F and App^NL–G–F knock-in mice have been described previously (Saito et al., 2014). App^NL–F mice contain the Swedish (KM670/671NL) and the Beyreuther/Iberian (I716F) mutations whereas App^NL–G–F mice have additionally the Arctic (E693G) mutation inserted into the mouse App gene (Saito et al., 2014). App knock-in mouse experiments were performed under ethical permit ID 407 approved by Linköping animal ethical board and 12570-2021 approved by Stockholm animal ethical board. Mice were kept on 12:12 light–dark cycle and with ad libitum access to food.

The mice were anesthetized by isoflurane. Thereafter, the mice were perfused with phosphate buffered saline (PBS) through cardiac perfusion and the mouse brains were collected. One hemisphere of the brain tissue was fixed in 10% formalin solution (Merck Millipore, MA, United States, Cat. HT501128) for immunohistochemistry, and the other hemisphere was dissected into hippocampus and cortex for biochemical analysis.

Six-well plates were coated by poly-D-lysine (Sigma-Aldrich, MO, United States, cat. P6407) for 1 h at room temperature and washed three times with Milli-Q water. Mouse brains from E17 embryos from WT and App^NL–G–F female mice were separated in Hanks’ balanced salt solution (HBSS) (Thermo Fisher Scientific, MA, United States, Cat. 14175095) and cortex/hippocampus were dissected under the dissection microscope. Cortex/hippocampus were transferred to a falcon tube and HBSS was removed. Mixture of 97% Neurobasal Medium (Thermo Fisher Scientific, MA, United States, Cat. 21103049), 1% Glutamax (Thermo Fisher Scientific, MA, United States, Cat. 35050038), and 2% B-27 (Thermo Fisher Scientific, MA, United States, Cat. 17504044) were added and tissues were separated by pipetting 20–30 times. Cells were counted by using a hematocytometer after trypan blue staining and 4.5 × 10⁵ cells were seeded. After 18 DIV, cells were treated with 100 nM bafilomycin A1 (Sigma-Aldrich, MO, United States, Cat. B1793) for 6 h and then the cells were collected and lysed in radioimmunoprecipitation assay (RIPA) buffer (Thermo Fisher Scientific, MA, United States, Cat. 89901) containing protease and phosphatase inhibitors for 10 min. Cell lysate was sonicated for 1 min and centrifuged at 15,000 rpm at 4°C for 20 min. Supernatant was transferred to the new tubes for western blot analysis.

For human brain samples, 3,3′-Diaminobenzidine (DAB) staining was performed with Dako EnVision Systems/HRP (Agilent Technologies, CA, United States, Cat. K401011-2). Slides were deparaffinized by xylene and ethanol and blocked by peroxidase (Dako kit) for 5 min and with 10% normal goat serum for 20 min at room temperature. The slides were incubated with anti-p62 antibody (Cell Signaling Technology, MA, United States, Cat. 5114) (1:500) in 3% normal goat serum overnight. After washing with 0.05% PBS-T buffer, slides were incubated with an anti-rabbit secondary antibody (Dako kit) for 30 min at room temperature, followed by incubating with chromogen solution, which was diluted in DAB substrate buffer (Dako kit), for 5 min at room temperature. Before mounting, slides were counterstained with Mayer’s Hematoxylin for 30 s and processed by dehydration. Images were acquired by the Nikon Eclipse E800 microscope with Nikon DS-Ri2 camera (acquisition parameters: 64 analog gain, 15-ms exposure time, and 4,908 × 3,264 pixel resolution) and were quantified by ImageJ software.

For mouse brain samples, paraffin-embedded brain tissues were sectioned into 4-μm thick sections. Antigen retrieval was performed, and after blocking, slides were incubated with anti-Aβ antibody (82E1) (Immuno-Biological Laboratories, Hokkaido, Japan, Cat.10323) (1:1,000), anti-LC3B (Novus Biologicals, CO, United States, Cat. NB100-2220) (1:2,000), and anti-p62 antibody (1:100). The following day, slides were incubated with secondary antibodies: alexa 546 goat anti-mouse (Invitrogen, MA, United States, Cat. A11030) (1:200), biotinylated goat anti-mouse IgG (Vector Laboratories, CA, United States, Cat. BA-9200) (1:200), and biotinylated goat anti-rabbit IgG (Vector Laboratories, CA, United States, Cat. BA-1000) (1:200). The biotinylated secondary antibodies were amplified with TSA Fluorescein System (PerkinElmer, MA, United States, Cat. NEL701001KT). Images were acquired by the Nikon Eclipse E800 microscope with Nikon DS-Qi2 camera (acquisition parameters: 64 analog gain, 500 ms or 1 s exposure time and 4,908 × 3,264 pixel resolution) and were quantified by ImageJ software.

Fresh frozen mouse brain samples were homogenized in RIPA buffer supplemented with phosphatases inhibitors (Sigma-Aldrich, MO, United States, Cat. P0044) and proteases inhibitors (G-Biosciences, MO, United States, Cat. 786-433) or separated into cytosolic and membrane fractions. For cytosolic fraction, tissues were homogenized in 10 mM Tris (pH: 8.0) and 0.25 M sucrose and centrifuged at 4,000 rpm for 15 min at 4^°C. Supernatant was centrifuged at 53,000 rpm for 1 h at 4^°C. The supernatant was kept as a cytosolic fraction. Tissue homogenates and cell lysates were boiled at 95^°C for 3 min except for analysis of Atg9A and Atg16L. A 20 μg portion of protein was loaded onto 4–20% SDS-PAGE for separation and transferred to polyvinylidene difluoride (PVDF) or nitrocellulose membranes. The membranes were blocked by 5% skim milk and were probed by primary antibodies, anti-p62 (1:500), anti-LC3 (Novus Biologicals, CO, United States, Cat. NB100-2331) (1:1,000), anti-p-Ulk1 S555 (Cell Signaling Technology, MA, United States, Cat. 5869) (1:500), anti-p-Ulk1 S757 (Cell Signaling Technology, MA, United States, Cat. 14202) (1:500), anti-Ulk1 (Cell Signaling Technology, MA, United States, Cat. 8054) (1:500), anti-Atg5 (Novus Biologicals, CO, United States, Cat. NB110-53818) (1:500), anti-Atg7 (Santa Cruz Biotechnology, CA, United States, Cat. sc-376212) (1:200), anti-Atg9A (Abcam, United Kingdom, Cat. ab108338) (1:500), anti-Atg16L (MBL Life Science, Tokyo, Japan Cat. PM040) (1:500), and anti-β-actin (Sigma-Aldrich, MO, United States, Cat. A2228) (1:10,000) overnight at 4^°C. The next day, the PVDF or nitrocellulose membranes were incubated with fluorescently labeled secondary antibodies (LI-COR Biosciences, NE, United States), IRDye 800CW donkey anti-rabbit (Cat. 926-32213) (1:10,000), IRDye 680RD goat anti-rabbit (Cat. 926-68071) (1:10,000), IRDye 800CW donkey anti- mouse (Cat. 926-32212) (1:10,000), or IRDye 680RD goat anti-mouse (Cat. 926-68070) (1:10,000) for 1 h at room temperature. Images were acquired by a fluorescence imaging system (LI-COR Biosciences, NE, United States, Odyssey CLx) and were analyzed by Image Studio Lite (LI-COR Biosciences, NE, United States) software.

Fresh mouse brain tissue was kept in RNAprotect Tissue Reagent (Qiagen, Venlo, Netherlands, Cat. 76104) and Ribonucleic acid (RNA) was extracted according to the manufacturer’s instruction of RNeasy Lipid Tissue Mini Kit (Qiagen, Venlo, Netherlands, Cat. 74804). After measuring RNA concentration, 200 ng of RNA was used for complementary deoxyribonucleic acid (cDNA) synthesis according to manufacturer’s instructions of the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, MA, United States, Cat. 4374966). The TaqMan Fast Advanced Master Mix (Thermo Fisher Scientific, MA, United States, Cat. 4444557) was used to perform the real-time polymerase chain reaction (RT-PCR) using TaqMan mouse gene expression assays (FAM) (Thermo Fisher Scientific, MA, United States, Cat. 4331182), Mm00448091_m1 for Sqstm1 (Gene aliases: p62). The gene expression level was normalized to TaqMan mouse gene expression assays (VIC) (Thermo Fisher Scientific, MA, United States, Cat. 4448489), Mm02619580_g1 for Atcb. Each sample was triplicated and run in the 7500 Fast Real-Time PCR System (Applied Biosystems, MA, United States).

Mice, 24-month-old, were anesthetized by isoflurane and perfused with 2.5% glutaraldehyde and 1% paraformaldehyde in 0.1 M phosphate buffer by cardiac perfusion. Thereafter, the mouse brains were collected and fixed in the same solution as used in perfusion process. Brain tissues were sectioned coronally into 1-mm thickness on the brain slicer matrix. The coronal brain slices were further dissected under the dissection microscope to separate the hippocampus. The hippocampal tissues were rinsed in 0.1 M phosphate buffer prior to the post fixation in 2% osmium tetroxide in 0.1 M phosphate buffer at 4^°C for 2 h. Tissues were dehydrated stepwise in ethanol and acetone and eventually embedded in LX-112. Prior to the ultrathin sections, the tissue slices were mounted on the glass slides and stained with a toluidine blue solution to visualize the hippocampal area. Ultrathin sections were prepared using an EM UC7 (Leica, Wetzlar, Germany) and the grids contrasted with uranyl acetate and lead citrate. Aβ42 was expressed and purified from Escherichia coli (Abelein et al., 2020). Aβ42 fibril was prepared by incubating 3 μM Aβ42 monomer at 37^°C for 24 h. After fibril formation, 10 μl of Aβ42 fibril was spotted on copper grid coated with formvar/carbon. Excess sample was wiped with Whatman filter paper and then grid was washed two times with 10 μl of Milli-Q water. It was then stained with 1% uranyl formate and then air dried. The sample was then imaged using TEM. Imaging was performed under the Hitachi HT7700 transmission electron microscope (Hitachi High-Technologies, Tokyo, Japan) operated at 80 kV equipped with a 2k × 2k Veleta CCD camera (Olympus Soft Imaging Solutions, Münster, Germany).

For the analysis of biochemical data, one-way ANOVA followed by Dunnett’s multiple comparisons test was performed for the three-group comparisons while two-tailed Student’s t-test was performed for the two-group comparisons in GraphPad Prism 8.

To investigate autophagy alteration in AD postmortem brain, we stained for p62. p62 is normally degraded along with the cargo during functional autophagy and therefore p62 accumulation and aggregation indicate impaired autophagy. We immunostained brains from healthy controls (Figures 1A–D) and AD subjects (Figures 1E–H) and quantified the p62-positive signals. In AD brains, p62 accumulation occurs in pyramidal neurons as previously shown (Kuusisto et al., 2002) in both cortex (Figure 1E) and hippocampus including dentate gyrus (DG) (Figure 1F) and CA1 (Figure 1G) of AD brains (semi-quantified in Figures 1I,J), whereas no p62 accumulation was found in brains of non-demented individuals (Figures 1A–C). p62 accumulated inside the neurons (Figures 1E,G arrow in depicted area and Figure 1F arrow in depicted area 1) and was also identified as small round puncta (Figure 1F arrows in depicted area 2). The puncta may represent axonal beadings, which are a series of swellings along the axons located in the molecular layer of DG. Interestingly, tunica intima of the vessels and corpora amylacea also contains substantial amount of p62 whereas no staining was observed in healthy controls (Figures 1D,H). Western blot analysis of brain homogenates from AD subjects with Braak stages 5–6 revealed significantly increased p62 levels as compared to healthy subjects (Figures 1K,L). In summary, p62 accumulation was frequently observed in AD brains, which indicates a disturbed autophagy-lysosomal system potentially caused by an inhibition of autophagy in both neurons and cells in the tunica intima of vessels.

To investigate the effect of the increasing Aβ amyloidosis in the App knock-in mice (Figure 2A) on autophagy, we analyzed markers of autophagic flux. The App knock-in mice exhibited alterations in autophagy as shown by western blot analysis for autophagy markers LC3 and p62 using 12-month-old WT, App^NL–F, and App^NL–G–F mouse brains. App^NL–F mice started to accumulate Aβ plaques at 12 months of age whereas App^NL–G–F mice exhibited a very pronounced Aβ pathology at this age (Figure 2A). We found that p62, LC3-II, and LC3-II/LC3-I ratio were significantly increased in the cortex of App^NL–G–F mice (Figures 2B,D and full blot of p62 in Supplementary Figure 1A, higher exposure of LC3 blots in Supplementary Figure 1B, and LC3-II/LC3-I in Supplementary Figure 1C) whereas p62 mRNA level was unaltered (Supplementary Figure 1E), indicating that autophagy was inhibited in the cortex of App^NL–G–F mice. The LC3-II levels were also similarly increased in the hippocampus of App^NL–G–F mice, although the ratio of LC3-II/LC3-I did not change significantly (Figures 2C,E and higher exposure time of LC3 blots in Supplementary Figure 1B, and LC3-II/LC3-I in Supplementary Figure 1D). In agreement with the p62 western blot data, immunostaining showed a significant accumulation of p62 in the cortex of App^NL–G–F mice as compared to WT mice (Figures 2F,G) whereas both p62 and LC3-II levels are unaltered in mouse primary neurons from App^NL−G−F mice as compared to WT (Supplementary Figures 1F–J). In contrast to App^NL–G–F mice, the LC3-II levels were significantly decreased in the hippocampus of App^NL–F mice whereas no changes of p62 levels in cortex and hippocampus were detected (Figures 2B–G).

Previous studies have identified alterations in several key autophagy proteins in AD brains (Heckmann et al., 2020). We therefore next analyzed the brains of the App knock-in mice for these proteins. The data interestingly showed a significant increase of p-Ulk1 S555 specifically in App^NL–G–F mouse cortex, which is involved in activating initiation of autophagy; whereas no change of p-Ulk1 S757, which is involved in inhibition of autophagy initiation, was observed (Figures 3A–E). In addition, a significant increase of both Atg7 and Atg9A was detected in the brains of App^NL–G–F mice whereas Atg5–Atg12 and Atg16L were unaltered (Figures 3F–J).

To further investigate the autophagy alterations upon aging of the App^NL–F mice, we next performed western blot analysis of LC3 and p62 using 18- and 24-month-old App^NL–F mice and no significant differences were observed, which was also confirmed with p62 immunostaining (Supplementary Figures 2A–J). However, EM analysis of 24-month-old App^NL–F mice exhibiting robust Aβ pathology (Figure 4A), including the region of striatum lacunosum-moleculare layer (Figures 4B,C), revealed a significant accumulation of autophagic vacuoles in the dystrophic neurites, around the Aβ plaques consisting of Aβ fibrils (Figures 4D–I) whereas no Aβ plaques or dystrophic neurites containing autophagic vacuoles were found in 24-month-old WT mouse hippocampus (Figures 4J,K). In agreement with the EM data, double immunostaining of Aβ and LC3 further confirmed the presence of LC3-positive puncta in association with the Aβ plaques in the 24-month-old App^NL–F mice (Figure 4L). Taken these data together indicate that the autophagic system of App^NL–G–F mice is characterized by an increase of both p62 and LC3-II, indicating that the autophagic system is inhibited, whereas the App^NL–F mice exhibited a region-specific alteration around the Aβ plaques. Hence, the degree of Aβ pathology correlates with the extent of autophagy alterations in the App knock-in mice.