Hair follicles were extracted by mechanical dissection after enzymatic digestion [15]. To isolate the hair follicles, hair-follicle pads of male Sprague-Dawley (SD) rats (21–26 g; purchased from the Animal Experiment Center of the Second Affiliated Hospital of Harbin Medical University) were dissected and digested with 0.1% collagenase (17018029, Gibco, Thermo Fisher Scientific, Waltham, MA, USA) in Dulbecco’s modified Eagle’s medium (DMEM) (11965092, Gibco, Thermo Fisher Scientific). The hair follicles were then carefully extracted from the pads and placed into 12-well tissue culture dishes (Beyotime, Shanghai, China) for cultivation. The hair follicles were cultivated in DMEM/F12 (11320033, Gibco, Thermo Fisher Scientific) medium with 1% penicillin and streptomycin (15140122, Gibco, Thermo Fisher Scientific) and 10% fetal bovine serum (0025, ScienCell Research Laboratories, Santiago, CA, USA) at 37 °C in a humidified conditions with 5% CO₂. Cell growth was monitored under a microscope, and half-medium alterations were conducted every 2–3 days. Primary cultures were first passaged after 14 days of cultivation, followed by subsequent passages every 3–5 days. Upon reaching 85–90% confluence, cells were dissociated via Accutase™ (A1110501, Gibco, Thermo Fisher Scientific), a neural stem-cell-specific cell dissociation reagent, during passage. The third-generation HFSCs were cultivated (3 $\times$ 10⁴ cells/cm²) in a six-well tissue-culture-plate medium. Each procedure was carried out in a sterile setting. In this study, third-generation HFSCs were used. All cell lines were authenticated via autosomal short tandem repeat (STR) profiling and confirmed negative for mycoplasma contamination.

The lentivirus vector was PLJM1-EGFP-puro (FH1717, Fenghui Biological Company, Changsha, Hunan, China). The third passage of cultured HFSCs was co-cultured with either NGF-overexpressing lentivirus (Gene: Lv-Rat-NGF-OE, NM_001277055) or blank lentivirus at a multiplicity of infection of 80 for 72 h. Specifically, 16 µL of virus stock with a titer of 1 $\times$ 10⁹ transducing units (TU)/mL was added per mL of culture medium. Transduced cells were selected with 2.0 µg/mL puromycin dihydrochloride (58582, Sigma-Aldrich, St. Louis, MO, USA) and utilized for following trials. The average transduction efficiency was assessed via a fluorescence microscope (Nikon Eclipse Ti2, Tokyo, Japan) and calculated as (number of EGFP-positive cells/total number of cells) $\times$ 100%, which was approximately 75%.

We performed surface-marker detection of the cell line using flow cytometry. Single-cell suspensions of third-passage HFSCs (4-week in vitro culture) were prepared, counted, and aliquoted ($\ge$1 $\times$ 10⁵ cells) into 1.5-mL EP tubes. After centrifugation (1000 rpm, 5 min) and supernatant elimination, resuspension of cells was conducted in 100 µL flow staining buffer (R32816, Yuanye Bio, Shanghai, China) and incubated with flow cytometry (FC) blocking reagent (101302, BioLegend, San Diego, CA, USA) (20 µL/tube) at room temperature for 30 min to reduce non-specific binding. Fluorochrome-conjugated primary antibodies were added at optimized concentrations (20 µL/1 $\times$ 10⁶ cells): anti-rat cluster of differentiation 45-allophycocyanin (CD45-APC) (1:100, F3104503, Lianke Bio, Hangzhou, Zhejiang, China), CD31-PE (1:150, sc-18916 PE, Santa Cruz Biotechnology, Dallas, TX, USA), CD90.1 (1:100, HIS51, 14-0900-81, Thermo Fisher Scientific), and CD29-fluorescein isothiocyanate (FITC) (1:100, eBioHMb1-1, 11-0291-80, Thermo Fisher Scientific). An incubation of tubes was conducted in the dark at room temperature for 1 h, then washing was conducted twice with 1 mL pre-cooled flow-staining buffer (1000 rpm, 5 min centrifugation) to remove unbound antibodies. Cells were resuspended in 500 µL flow-staining buffer, protected from light, and analyzed on a BD FACS Canto II cytometer (BD Biosciences, San Jose, CA, USA) within 2–6 h. Isotype controls (detailed in the Supplementary Material “Isotype controls.pdf”) were included to define background fluorescence. Gating excluded debris/aggregates via forward scatter (FSC) and side scatter (SSC) [FSC/SSC], with positive/negative populations distinguished using isotype control-derived fluorescence thresholds. Flow cytometry analysis showed that the cell surface markers CD29 (a well-recognized marker for HFSCs) and CD90 (cytokeratin) were highly expressed in our isolated HFSCs, and CD31 (a marker for vascular endothelial cells) and CD45 (a marker for hematopoietic cells) were expressed at low levels (Fig. 1D). These results are consistent with the identification criteria.

Fig. 1.

Results of isolation, culture, and identification of HFSCs from the bulge region of hair follicles and the ability of HFSCs to poly-differentiate. Optical microscopy of an intact hair follicle, microdissected and extracted from the hair follicle in the bulge region for primary cell culture. (A) HFSCs in primary, P1, and P3 were under an optical microscope. After 14 days of culture, osteogenic and adipogenic differentiation was confirmed by (B) Alizarin red and (C) Oil Red O staining. (D) Flow cytometry detection of HFSCs surfaces positive markers CD29, CD90, negative markers CD31, CD45. (A,C) scale bar = 25 µm, (B) scale bar = 50 µm. Abbreviations: HFSCs, hair follicle stem cells; FITC-H, fluorescein isothiocyanate-height; PE-H, phycoerythrin-height; APC-H, allophycocyanin-height; CD, cluster of differentiation; P, passage.

As directed by the manufacturer [32], the growth medium was exchanged for an osteogenic or adipogenic differentiation medium after a 24-h incubation. Three days were given for a shift in the osteogenic and adipogenic differentiation media while the cells were cultivated in an appropriate environment at 37 °C with 5% CO₂. Alizarin red (Y027377, Beyotime) and Oil Red O (Y264669, Beyotime) were used to confirm osteogenic and adipogenic differentiation after 14 days of culture.

Flow cytometry was applied in order to detect cell-surface markers. The 3rd-generation HFSCs cell suspension was made by digesting the cells with trypsin (0.25%, 25200056, Thermo Fisher Scientific), then rinsing two times with phosphate buffered saline (PBS), and finally adjusting to 1 $\times$ 10⁵ cells/mL. Afterward, the cells were incubated with an antibody against CD29 (0.5 mg/mL; 11-0291-82; Thermo Fisher Scientific) coupled with fluorescein isothiocyanate isoform, isophycocyanin conjugated antibody against CD90 (0.5 mg/mL; 14-0900-81; Thermo Fisher Scientific), phycoerythrin-conjugated antibody against CD31 (2 mg/mL; sc-18916; Santa Cruz Biotechnology), or phycoerythrin-conjugated antibody against CD45 (5 µL/test; F3104503; Multisciences, Hangzhou, Zhejiang, China) for 1 h at room temperature. Staining was stopped with 1% bovine serum albumin, after which loading the cells onto a flow cytometer was conducted (NovoCyte Advanteon, ACEA Biosciences, California, CA, USA). The same procedure was used for isotype controls.

The Animal Experiment Center of the Second Affiliated Hospital of Harbin Medical University provided clean-grade, male, SD rats (6–8 weeks old, 200–210 g) [License No. SCXK (Hei) 2023-001]. All rat experiments were conducted as per the principles outlined in the “Guide for the Care and Use of Laboratory Animals” published by the Ministry of Science and Technology of the People’s Republic of China. Sixty were randomly allocated into four groups (n = 15/gp): Sham+PBS, A$\beta$+PBS (AD+PBS), A$\beta$+HFSCs (AD+HFSCs), and A$\beta$+HFSCs/NGF (AD+HFSCs/NGF) groups (designated depending on the substance injected into the hippocampus). The experimental design for this hippocampus-targeted AD study was based on previous literature [33]. Experimental AD was established by stereotaxically injecting human A$\beta$1–42 (ab82795, Abcam, Cambridge, UK) diluted to 1 mg/mL, bilaterally, into the hippocampal cornu ammonis 1 (CA1) region (coordinates: A/P = +3.0 mm; M/L = +2.0 mm; D/V = +3.0 mm). Rats were anesthetized with 0.3% pentobarbital sodium (40 mg/kg, P3761, Shanghai Pharma, Shanghai, China) by i.p. injection and fixed in a stereotactic apparatus (RWD Life Sciences, Shenzhen, Guangdong, China). A microinjector (RWD Life Sciences) injected 10 µL A$\beta$1–42 solution into each side of the hippocampus (injection speed: 1 µL/min). In the Sham group, 10 µL of PBS was administered using the same method. After 14 days, rats were chosen randomly from Sham and A$\beta$ groups to run in the Morris water maze test, then anesthetized with 0.3% pentobarbital sodium (150 mg/kg) i.p. for euthanasia and then exposed to Congo red (MS4060-25G, Maokang Biotechnology, Shanghai, China) staining (n = 3). The remaining rats in the Sham and A$\beta$ groups were injected with 10 µL of PBS. A$\beta$+HFSCs and A$\beta$+HFSCs/NGF groups were subjected to 10 µL modified or unmodified HFSCs suspension (5 $\times$ 10⁵ third-generation) through the same needle track as the A$\beta$1–42 injection. On day 29, all rats underwent Morris water maze training and testing. Six days later, the rats were euthanized under intraperitoneal anesthesia with 0.3% pentobarbital sodium (150 mg/kg) and rat brains were harvested for analysis.

After being anesthetized, rats were subjected to transcardial perfusion with 0.9% physiological saline via thoracic cavity exposure. Brains were subsequently dissected, immersed in 4% paraformaldehyde (70-F0001; MultiSciences) for fixation, then PBS was utilized for rinsing. After that, the tissues were processed through gradient dehydration, paraffin-embedded, and sliced at a thickness of 40 µm. Then, slices were deparaffinized, followed by gradient rehydration and stained with Congo red dye for 10–20 min, then differentiated with alkaline ethanol differentiation solution for several seconds, and washed with water, then counterstained with hematoxylin (H3136; Sigma-Aldrich) for 2 min and rinsed with running water for 1–2 min. After that, the slices were dipped in ascending grades of ethanol. Finally, they were cleared with xylene (214736; Sigma-Aldrich) and sealed with neutral resin (G8590; Solarbio, Beijing, China). The slides were photographed under an optical microscope (ECLIPSE Ti2-U, Nikon).

The immersion of paraffin slices was conducted in 0.3% H₂O₂ for 30 min. Next, normal goat serum (c0265, Beyotime) was utilized to block the slices for 30 min at 25 °C, then an overnight incubation with primary mouse anti-$\beta$-Amyloid antibodies (6E10, 1:500 dilution; 803001, Biolegend) was conducted at 4 °C. The next day, they were rinsed 3 times with PBS, incubated at 37 °C for 1 h with the second antibody (1:500, ab205718, Abcam), and then washed with PBS. They were then visualized with a 3,3’-Diaminobenzidine (DAB) substrate kit (ab64238, Abcam). Sections were PBS-washed, counterstained with hematoxylin, and then hyalinized and dehydrated. An optical microscope was used to observe the staining effect and take photos.

The rats underwent a navigation test in a water maze (the water temperature was controlled at 25 $\pm$ 1 °C) for 5 consecutive days. The experimental environment was strictly standardized: Four fixed visual cues (30 $\times$ 30 cm; red circle, blue square, yellow triangle, green bar) were placed on the maze walls. Room conditions (light intensity: 200 lux; noise level: 40 dB) and the experimenter’s position were controlled to ensure reliable use of spatial cues by rats. The maze water was made opaque with 0.1% (w/v) food-grade titanium dioxide solution (Jianghu Titanium White Chemical Products Co., Shanghai, China) to hide the underwater platform. Prior to formal testing, all rats underwent habituation for 2 days before training: each was individually placed in the maze (without the platform) for 120 s of free swimming to become familiar with the pool, water temperature, and space. They were also adapted to the experimental room for 30 min daily to reduce environmental stress interference with behavior. For the 5-day navigation training, the rats were required to enter the water from four different quadrants with their heads facing the pool wall, and their movement trajectory and escape latency were recorded four times daily. Each training trial was limited to 60 s: if a rat reached the platform within 60 s, it was permitted to stay on the platform for 15 s to consolidate memory; if not, the experimenter guided it to the platform, where it stayed for 15 s. Escape latency is described as the time spent by the rat to first locate the underwater platform after entering the water. The spatial-probe test was conducted on the sixth day after the start of the water maze training. Rats were set into the water from the quadrant opposite to the original location of the platform, after the platform was eliminated, and the count of platform crossings was recorded.

Behavioral testing and image quantification were conducted blindly. All animals were ear-tagged for identification. Group allocation was done by a different person; both the operators performing behavioral tests and the personnel collating data were unaware of the groupings, thus reducing bias.

Rats from every group were euthanized and the hippocampus was dissected for measurement of the protein levels of A$\beta$1–40 and A$\beta$1–42. This was achieved by using A$\beta$40 ELISA kits (ER0754, Wuhan Fine Biotech Company, Wuhan, Hubei, China) and A$\beta$42 ELISA kits (ER0755, Wuhan Fine Biotech Company), as per the manufacturer’s guidelines. The quantified amounts of soluble A$\beta$, measured in picomolar units, were normalized to the total protein content of the cell lysate. The signal was identified via a Varioskan Flash multimode reader (Thermo Fisher Scientific).

The hippocampus was homogenized with a tissue extraction buffer that included protease inhibitors. This was accompanied by centrifugation at 12,000 rpm at a temperature of 4 °C for 10 min. The resulting supernatants were gathered and assessed for protein concentration with BCA protein assay kits (ab102536, Abcam). The protein samples, weighing 40 µg, were separated using 10% or 12.5% SDS-polyacrylamide gel electrophoresis (P0012A, Beyotime) at 100 V for 90–120 min. Then they were moved to a polyvinylidene fluoride (PVDF) membrane (88520, Thermo Fisher Scientific) using a current of 200 mA at 4 °C for 120 min. The PVDF membrane was preincubated with 3% nonfat milk in Tris-buffered saline with 0.1% Tween-20 (TBS-T) at 25 °C for 60 min to block nonspecific binding sites. The incubation of PVDF membranes was conducted overnight at 4 °C with primary antibodies targeting protein kinase B (Akt) (1:1000; 4685, Cell Signaling Technology, Danvers, MA, USA), phospho-Akt (Thr308) (1:1000; 13038, Cell Signaling Technology), glycogen synthase kinase (GSK)-3$\beta$ (1:1000; 9315, Cell Signaling Technology), phospho-GSK-3$\beta$ (ser9) (1:1000; 9336, Cell Signaling Technology), tau 5 (1:1000; ab80579, Abcam), p-tau (S396) (1:1000; CY5657, Abways, Shanghai, China), p-tau (S404) (1:1000; CY9003, Abways), p-tau (Thr231) (1:1000; CY6535, Abways), p-tau (Ser202/Thr205) (1:5000; 82568-1-RR, Proteintech, Rosemont, IL, USA), $\beta$-site amyloid precursor protein cleaving enzyme 1 (BACE1) (1:1000; CY6880, Abways), postsynaptic density protein 95 (PSD95) (1:1000; CY5407, Abways), synaptophysin (SYP) (1:1000; CY5273, Abways), PI3K (1:1000; 4292, Cell Signaling Technology), phosphorylated-PI3K (p-PI3K, 1:1000; 4228, Cell Signaling Technology) and loading control protein anti-$\beta$-Actin (1:3000; AF7018, Affinity, Changzhou, Jiangsu, China), anti-$\beta$-tubulin (1:1000; 2128, Cell Signaling Technology) and anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1:1000; 2118, Cell Signaling Technology). After being rinsed 3 times with TBS-T, the PVDF membranes were incubated with secondary antibodies conjugated with horseradish peroxidase (P2369; Beyotime) and exposed to the enhanced chemiluminescence Plus Western Blotting Detection Kit (35050, Thermo Fisher Scientific). Three replications were performed. The immunoblot pictures were obtained via the Omega-Lum G imaging system (ChemiScope 6300, Clinx Science Instruments, Shanghai, China). Image-Pro 11.1 (Media Cybernetics, Inc., Rockville, MD, USA) was used to calculate bands for semi-quantification. All values were standardized based on an equal loading control.

NGF levels in the untreated HFSCs, the HFSCs/Vector, and the HFSCs/NGF groups were quantified by qRT-PCR assay. In brief, total RNA isolation from the respective HFSCs was conducted via TriZol reagent (15596026CN, Invitrogen, Carlsbad, CA, USA). Reverse transcription was conducted via the PrimeScript RT reagent kit (RR037A, TaKaRa, Dalian, Shandong, China). Quantitative PCR was conducted via the SYBR PrimeScript RT-PCR Kit II (RR820A, Merck KGaA, Darmstadt, Germany) on an LC480 real-time PCR system (Influence Bio, Shanghai, China). The sequence of each primer is shown in Table 1.

Data are reported as the mean $\pm$ standard error of the mean (SEM). The density was measured via ImageJ software (Version 1.53t, Wayne Rasband, National Institutes of Health, Bethesda, MD, USA) . The statistical analysis and generation of statistical graphics were conducted using GraphPad Prism 9.5 software (GraphPad Software, LLC, San Diego, CA, USA). Data analysis was conducted with a 1-way analysis of variance and Tukey’s test for several comparisons. p $\le$ 0.05 were deemed to be significant.

HFSCs’ isolation from the upper lips of male SD rats was conducted; hair-follicle pads were cultivated until the 3rd generation (Fig. 1A). After inducing the osteogenic differentiation, HFSCs differentiated into osteoblasts, confirmed by positive Alizarin red staining (Fig. 1B). Moreover, HFSCs differentiated into adipocytes, as manifested by the presence of lipid droplets with orange coloration upon staining with Oil Red O dye (Fig. 1C). These steps illustrated the multidirectional differentiation capability of HFSCs. Flow-cytometry analysis revealed that the cell surface markers of HFSCs, such as CD29 and CD90, were overexpressed, while the endothelial cell-surface marker CD31 and the hematopoietic cell-surface marker CD45 were suppressed (Fig. 1D).

Fig. 2A illustrates the viral vector utilized in this investigation. The level of NGF mRNA expression was quantified by using PCR in the control, the HFSCs/Vector, and the HFSCs/NGF groups (Fig. 2B). The expression levels of NGF in these groups were assessed with WB analysis (Fig. 2C,D, the original WB images can be found in the Supplementary Material). These outcomes verified the HFSCs/NGF’s successful construction.

Fig. 2.

Successful construction of NGF over-expressed HFSCs. (A) The structure of the lentivirus vector with NGF over-expressing. (B) The NGF mRNA level was quantified with Real-time PCR. (C,D) The NGF expression levels were estimated through western blot analysis. Data analysis was conducted by one-way ANOVA with Tukey’s post-hoc test (B,D), *p 0.05, ****p 0.0001; ns, not significant, n = 3. Abbreviations: GAP, GTPase activating protein; RRE, rev response element; CMV, cytomegalovirus; EGFP, enhanced green fluorescent protein; CPPT, central polypurine tract; CTS, central termination sequence; hPGK, human phosphoglycerate kinase promoter; LTR, long terminal repeat; SV40, simian virus 40; AmpR, ampicillin resistance gene; RSV, rous sarcoma virus; HIV, human immunodeficiency virus; PLJMI-EGFP-Puro, PLJM1-enhanced green fluorescent protein-puromycin resistance gene.

A$\beta$1–42 or PBS was injected into the hippocampal CA1 region in multiple groups. Fourteen days later, three rats were randomly selected from the sham and A$\beta$ groups for the Morris water maze test and Congo red staining. The water maze test results indicated that rats in the A$\beta$ group wasted more time finding the platform than did those in the Sham group (Fig. 3A). In the spatial probe test, frequency of crossing the removed platform was lower (Fig. 3B), and the duration in the target quadrant was shorter for rats in the A$\beta$ group than for those in the sham group (Fig. 3C). To eliminate the possibility that Morris water maze performance differences were confounded by motor or visual impairments, additional assessments were conducted before the formal test. These confirm that the water maze differences reflect cognitive, not motor or visual impairment. Congo red staining of the hippocampus revealed that A$\beta$ deposition was found in the A$\beta$ group and almost no A$\beta$ deposition in the Sham group (Fig. 3D,E).

Fig. 3.

AD rat model results. (A) The latency for all groups to escape in the location-navigation exam. (B,C) In the spatial probe test, the number of platform crossings and the time in the target section in each group. (D) A$\beta$ deposition was detected in the rats’ hippocampus CA1 of the sham and A$\beta$ groups by Congo red, scale bars: 50 µm. (E) Quantitative assessment of Congo red stained plaque number in 2 groups. Data were analyzed by two-way ANOVA (A), and analyzed by one-way ANOVA (B,C,E), *p 0.05, **p 0.01, ***p 0.001; ns, not significant, n = 3. Abbreviations: AD, Alzheimer’s disease; CA1, cornu ammonis 1; A$\beta$, amyloid plaques.

To investigate the potential of transplanted HFSCs to restore behavioral deficits in AD rats, Morris water maze tests were performed 14 days post-cell transplantation. During the five-day training period, HFSC-transplanted rats exhibited a gradual increase in their ability to find a concealed platform (Fig. 4). Rats in the HFSCs/NGF transplanted group exhibited significantly better performance than did those in the HFSCs transplanted group (Fig. 4A). Throughout the spatial probe test, the frequency of crossing the eliminated platform was higher (Fig. 4B), and the duration in the target quadrant was longer in the HFSCs group and HFSCs/NGF group than in the AD group (Fig. 4C). Moreover, a significant difference in tracks of the spatial probe test on the 6th day was observed among the four groups (Fig. 4D); the rats transplanted with HFSCs and HFSCs/NGF improved their approaches for reaching the platform. In summary, these results indicated that transplantation of HFSCs and HFSCs/NGF reduced cognitive dysfunction in AD rats.

Fig. 4.

HFSCs and HFSCs/NGF transplantation improved the learning and memory impairment. (A) The latency for each group to escape in the location-navigation test, in the spatial-probe test, (B) the number of platform crossing and (C) time in the target section. (D) Illustrative tracks of the spatial probe test on day 6. When comparing with the AD rats, the rats transplanted with HFSCs and HFSCs/NGF showed significantly improved learning and memory. Data were assessed by two-way ANOVA (A), and analyzed by one-way ANOVA (B,C), AD+PBS group vs. AD+HFSCs group, *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001; AD+PBS group vs. AD+HFSCs/NGF group, ^#⁢#p 0.01, ^{#⁢#⁢#⁢#}p 0.0001; AD+HFSCs group vs. AD+HFSCs/NGF group, ^Δp 0.05, n = 6. Abbreviations: NW, northwest; SW, southwest; NE, northeast; SE, southeast.

To investigate the effects of HFSCs and HFSCs/NGF transplantation on the production of toxic amyloid plaques, immunohistochemistry was conducted using 6E10 antibodies in the brains of AD rats (Fig. 5A). Quantification of stained plaques in the hippocampus revealed significant variations between the HFSCs and the HFSCs/NGF groups (Fig. 5B). Protein levels were measured in four rat groups using A$\beta$1–40 and A$\beta$1–42 ELISA kits (Fig. 5C,D). Compared to PBS-treated AD rats, levels of A$\beta$1–40 and A$\beta$1–42 had significantly declined in both the HFSCs group and HFSCs/NGF group rats, as measured by ELISA. These findings led to the conclusion that HFSCs and HFSCs/NGF transplantation in AD rats could decrease the levels of A$\beta$1–40 and A$\beta$1–42. To elucidate the mechanisms underlying the reduction of A$\beta$ depositions in HFSCs and HFSCs/NGF group AD rats, the BACE1 levels were quantified by western blot analysis (Fig. 5E,F, the original WB images can be found in the Supplementary Material). The levels of BACE1 in the hippocampus of the HFSCs group and HFSCs/NGF group rats were significantly lower than those in the AD group rats. These data provide evidence that the level of BACE1 declined in the hippocampus during the HFSCs or HFSCs/NGF transplantation, thereby inhibiting the A$\beta$PP amyloid hydrolysis pathway.

Fig. 5.

HFSCs and HFSCs/NGF transplantation declined the level of A$\beta$40 and A$\beta$42. (A) Amyloid plaques were quantified by immunohistochemistry via 6E10 antibody. (B) Quantitative assessment of 6E10 stained plaque number in CA1 area. (C,D) Compared with PBS-treated AD rats, A$\beta$1–40 and A$\beta$1–42 levels were markedly decreased in HFSCs and HFSCs/NGF group rats as measured by ELISA. (E) The BACE1 protein level was assessed by western blot in the hippocampus of rats. (F) The relative quantification of BACE1 protein expression. Scale bars: 100 µm. One-way ANOVA with Tukey’s post-hoc test was utilized to analyze data (B–D,F), *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001; ns, not significant, n = 3. Abbreviations: BACE1, $\beta$-site amyloid precursor protein cleaving enzyme 1; ELISA, enzyme-linked immunosorbent assay.

Another important pathology of AD is tau hyperphosphorylation. To verify the consequences of HFSCs and HFSCs/NGF on tau hyperphosphorylation, the phosphorylation levels of tau 5 and other tau sites were quantified by using western blot analysis. The outcomes illustrated that HFSCs and HFSCs/NGF transplantation inhibited tau 5 and tau phosphorylation at Ser202/Thr205, Thr231, S396, and S404 sites in the AD rat hippocampus (Fig. 6, the original WB images can be found in the Supplementary Material).

Fig. 6.

HFSCs and HFSCs/NGF transplantation reduced tau hyperphosphorylation. (A) The tau phosphorylation site levels were quantified by western blot analysis, showing that HFSCs and HFSCs/NGF transplantation inhibited tau phosphorylation at Ser202/Thr205, Thr231, S396, and S404 sites in the hippocampus of AD rats. (B) The relative quantification of Ser202/Thr205, Thr231, S396, and S404 sites protein expression. (C) The tau 5 concentration was assessed by western blot analysis in the hippocampus. (D) The relative quantification of tau5 protein expression. One-way ANOVA with Tukey’s post-hoc test was utilized to analyze the data (B,D), *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001, n = 3. GAPDH, anti-glyceraldehyde- 3-phosphate dehydrogenase.

Proteins of the PI3K/Akt/GSK3$\beta$ pathway were examined with western blot analysis (Fig. 7, the original WB images can be found in the Supplementary Material), which revealed significantly higher levels of phosphorylated PI3K and phosphorylated Akt after transplantation of HFSCs and HFSCs/NGF than in both the Sham group and the AD group. Additionally, the levels of phosphorylated GSK3$\beta$ at the Ser9 site to suppress GSK3$\beta$ activity were significantly raised in the HFSCs and the HFSCs/NGF groups. Total PI3K, Akt, and GSK3$\beta$ were investigated as loading controls. These results revealed that HFSCs and HFSCs/NGF transplantation could reduce tau hyperphosphorylation by the PI3K/Akt pathway activation to suppress GSK3$\beta$ activity.

Fig. 7.

HFSCs and HFSCs/NGF transplantation inhibited the activity of GSK3$\beta$ through PI3K/Akt pathways activation. (A) The levels of GSK3$\beta$, p-GSK-3$\beta$, PI3K, p-PI3K, Akt, and p-Akt were quantified by immunoblots and (B) the relative quantification of protein expression in each group. One-way ANOVA with Tukey’s post-hoc test was utilized to analyze data (B), *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001; ns, not significant, n = 3. Abbreviations: GSK3$\beta$, glycogen synthase kinase 3$\beta$; PI3K, phosphoinositide-3 kinase; Akt, protein kinase B; p-PI3K, phosphorylated PI3K; p-Akt, phosphorylated Akt; p-GSK3$\beta$, phosphorylated GSK3$\beta$.

Of all the pathological changes that occur in the brains of individuals with AD, synaptic loss is most strictly related to cognitive decline [34]. Therefore, we analyzed the SYP and PSD-95 levels in the hippocampus of the four groups by immunohistochemical staining and western blotting (Fig. 8). The SYP and PSD-95 levels were significantly lower in AD+PBS rats than in Sham+PBS rats. However, after HFSCs and HFSCs/NGF transplantation, the level of SYP rose significantly (Fig. 8A,B, the original WB images can be found in the Supplementary Material). The PSD-95 level displayed a significant elevation in the HFSCs/NGF group, but no significant difference was shown between the AD+PBS and the AD+HFSCs groups (Fig. 8C,D).

Fig. 8.

HFSCs transplantation increased the level of synaptophysin and PSD-95 expression. (A) The SYP protein level was assessed by western blot analysis in the hippocampus of three groups. (B) The relative quantification of protein expression. (C) The PSD-95 protein level was assessed by western blot analysis in the hippocampus of three groups. (D) The relative quantification of protein expression. Data were analyzed by one-way ANOVA with Tukey’s post-hoc test (B,D), *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001; ns, not significant, n = 3. Abbreviations: PSD-95, postsynaptic density-95; SYP, synaptophysin.