Serum samples were collected from 18 AD patients and 18 healthy volunteers at The Second Affiliated Hospital of Nanchang University. The AD patients with diabetes, cancer, or other severe diseases were excluded. This study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the Second Affiliated Hospital of Nanchang University (No:202392). Written informed consent was collected from all participators.

SH-SY5Y cell line was purchased from American Type Culture Collection (ATCC, Gaithersburg, MD, USA) and cultured in dulbecco’s modified eagle medium (DMEM, 11965092, Thermo Fisher, Waltham, MA, USA) containing 10% fetal bovine serum (10099158, Thermo Fisher) at 37 °C with 5% CO₂. SH-SY5Y cells were authenticated by Short Tandem Repeat (STR) analysis and tested for mycoplasma contamination using the PCR assay (Result revealed that SH-SY5Y cell cell line is tested negative for mycoplasma). An in vitro AD model was established by stimulation with the oligomer of amyloid beta (ERMDA482IFCC, A$\beta$1–42, 10 µM, Sigma-Aldrich, St. Louis, MO, USA) for 24 h.

Overexpression plasmid for METTL14 (OE-METTL14) and its vector pcDNA3.1, short hairpin RNA for METTL14 (sh-METTL14, sequences: CCATGTACTTACAAGCCGATA), sh-TUG1 (sequences: GGAACGGGCGTGCGGTCGATC), sh-SMURF1 (sequences: GCCCAGAGATACGAAAGAGAT), sh-GDF15 (sequences: CCCTCAGAGTTGCACTCCGAA), and their negative control shRNA (sh-NC, sequences: GTTCTCCGAACGTGTCACGT) were purchased from RiboBio (Guangzhou, China). SH-SY5Y cells in logarithmic growth phase were seeded into 6-well plates. SH-SY5Y cells with 70% confluence under culture condition were transfected with these segments using Lipofectamine 2000 (11668500, Thermo Fisher).

Total RNA was extracted from serum samples, SH-SY5Y cells, or hippocampus tissues using the Total RNA Extraction Kit (R1200, Solarbio, Beijing, China). Subsequently, the iscript cDNA synthesis kit (1708890, Bio-Rad, Hercules, CA, USA) was used for cDNA synthesis. RT-qPCR was performed using the Trans Start Tip Green qPCR Super Mix (AQ141-01, TransGen, Beijing, China). Gene expression levels normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were calculated by the 2^{-Δ⁢Δ⁢Ct} method. The primer sequences are presented in Table 1.

Protein samples were isolated using the Radio Immunoprecipitation Assay (RIPA) reagent (P0013B, Beyotime, Haimen, China) and protein concentration was quantified using the Bicinchoninic Acid (BCA) Protein Assay Kit (PC0020, Solarbio). Then, the protein samples were loaded onto sodium dodecyl sulfate polyacrylamide gel electrophoresis. After blotting onto the polyvinylidene fluoride membranes and blocking with 5% non-fat milk, the membranes were incubated with primary antibodies against glutathione peroxidase 4 (GPX4, 1:500, bs-3884R, Bioss, Beijing, China), solute carri1er family 7 member 11 (SLC7A11, 1:1000, ab307601, Abcam, Cambridge, UK), SMURF1 (1:500, bs-9391R, Bioss), GDF15 (1:1000, sc-515675, Santa Cruz, TX, USA), NRF2 (1:1000, ab62352, Abcam), heme oxygenase-1 (HO-1, 1:2000, ab52947, Abcam), glyceraldehyde-3-phosphate dehydrogenase (GAPDH, 1:5000, bsm-52262R, Bioss) overnight at 4 °C, followed by incubation with the secondary antibody Goat Anti-Rabbit IgG Fc and F(ab)/F(ab’)2 (H&L) (1:2000, ab205718, Abcam) for 1 h. The bands were developed with efficient chemiluminescence (ECL) solution (WBULP-100ML, Millipore, Burlington, MA, USA). The band intensity of the Control group was set to 1 for normalized statistics.

To evaluate SH-SY5Y cell viability, 5 mg/mL 3-(4,5-dimethylthiazol-2-yl)-2,5 diphenyltetrazolium bromide (MTT) solution (G4100, Promega, Madison, WI, USA) was added into each well of cells that were seeded in 96-well plates at a density of 3000 cells per well. After incubation at 37 °C for 4 h, the formed formazan crystals were dissolved by dimethyl sulfoxide solution (DMSO, D8418, Sigma-Aldrich). The absorbance was measured at 490 nm on a Microplate Reader (VA000010C, Thermo Fisher).

The Iron Assay Kit (ab83366, Abcam) was used to determine Fe²⁺ and total iron levels. In brief, SH-SY5Y cells were lysed by Iron Assay Buffer. Then, cell lysates were incubated with the Assay Buffer for Fe²⁺ detection or the Iron Reducer for total iron detection at 37 °C for 30 min, followed by incubation with the Iron Probe for 60 min at 37 °C. The results were measured at 593 nm on a Microplate Reader (VA000010C, Thermo Fisher).

The ROS level in SH-SY5Y cells was measured using the ROS Detection Assay (ab287839, Abcam). The cells were seeded into 96-well plates and incubated with the ROS/Superoxide Detection Solution at 37 °C for 1 h away from light. The results were detected using a fluorescence microplate reader ((excitation) Ex = 488 nm, emission (Em) = 520 nm, VA000010C, Thermo Fisher).

The concentrations of interleukin-1beta (IL-1$\beta$), interleukin-6 (IL-6), and tumor necrosis factor-alpha (TNF-$\alpha$) were assessed by commercial Human IL-1$\beta$ ELISA kit (SEKH-0002), Human IL-6 ELISA kit (SEKH-0013), and Human TNF-$\alpha$ ELISA kit (SEKH-0047). All ELISA kits were purchased from Sorlarbio (Beijing, China). Briefly, the supernatant of SH-SY5Y cells was collected by centrifugation at 1000 g for 10 min at 4 °C. Then, the supernatant was added to each well, followed by incubation for 90 min. Subsequently, the samples were incubated with the Biotin-Conjugate antibody for 1 h, Streptavidin-HRP for 30 min, and Substrate solution for 15 min, respectively. After adding with stop solution, the results were detected at 450 nm on a microplate reader (VA000010C, Thermo Fisher). The concentrations of IL-1$\beta$, IL-6, and TNF-$\alpha$ were calculated according to standard curves.

Total RNA m6A level was detected using the m6A RNA methylation quantification kit (P-9008-48, EpiGentek, Farmingdale, NJ, USA). Briefly, total RNA was isolated from SH-SY5Y cells and incubated with the capture antibody and detection antibody. After incubation, color developing solution was added into cells. Finally, the optical density at 450 nm was measured and the relative total RNA m6A level was calculated.

To determined TUG1 stability, SH-SY5Y cells were treated with 5 µg/mL actinomycin D (SBR00013, Sigma-Aldrich) for 3, 6, 9 h. Then, TUG1 level was assessed by RT-qPCR as described above.

The m6A level of TUG1 was assessed by MeRIP using the riboMeRIPTM m6A Transcriptome Profiling Kit (C11051-1, RiboBio). Briefly, total RNA was fragmentated and incubated with anti-m6A antibody (ab208577, Abcam) or anti-IgG (ab18448, Abcam) conjugated in magnetic beads A/G at 4 °C for 2 h. After elution, the RNA was collected from the precipitates and detected by RT-qPCR.

RIP assay was performed using the Magna RIP kit (17-701, Millipore). SH-SY5Y cells were lysed in the RIP lysis buffer and then coimmunoprecipitated in Protein A/G MagBeads that were pre-coated with anti-SMURF1 antibody (H00057154-M01, Thermo Fisher) or isotype control IgG (ab18448, Abcam). The specific binding RNAs to the antibody-protein complex were isolated and the level of TUG1 was analyzed by RT-qPCR.

RNA pull-down assay was performed using the Magnetic RNA-Protein Pull-Down Kit (20164, Thermo Fisher). Briefly, the protein samples of SH-SY5Y cells were incubated with the Biotin-labeled TUG1 probe that was pre-coupled in streptavidin beads at 4 °C for 2 h. After washing for five times, the RNA-protein complexes were retrieved from the beads and subjected to Western blotting.

The cell samples were lysed in Radio Immunoprecipitation Assay (RIPA) Lysis Buffer (P0013B, Beyotime) containing various protease inhibitors. After centrifugation at 12,000 g for 10 min, cell lysates were immune-precipitated with protein G Plus-Agarose pre-bound with anti-GDF15 antibody (5 µg, sc-515675, Santa Cruz). After immunoprecipitation overnight, the immune-precipitates were eluted and subjected to Western blotting with GAPDH as loading control. For ubiquitination assay, SH-SY5Y cells were treated with 10 µM MG132 for 8 h, and the cell lysates were immunoprecipitated by GDF15 antibody (sc-515675, Santa Cruz), and GDF15 ubiquitination was analyzed by Western blotting with ubiquitin antibody (ab140601, Abcam).

C57BL/6J mice (aged 6–8 weeks, weighing 20–25 g) were purchased from Slac Jingda Laboratory Animal Co., Ltd. (Hunan, China). AD was induced by scopolamine according to a previous study [25]. A total of 32 mice were divided into four experimental groups (n = 8 per group): control, AD model, AD + vector, and AD + OE-Mettl14. The mice in AD groups were intraperitoneally injected with 2 mg/kg scopolamine (Sigma-Aldrich) for 7 consecutive days. The control mice were intraperitoneally injected with equal volume of saline vehicle. The mice in AD + vector and AD + OE-Mettl14 groups were intrahippocampally injected with Adeno-associated virus type 2 (AAV2) -vector or AAV2-OE-Mettl14 (10¹⁰ pfu/mouse, Genechem, Shanghai, China) before the injection with scopolamine. At 24 h after the final scopolamine injection, all mice were subjected to behavioral test and then euthanized by cervical dislocation. The hippocampal tissues were collected for further examination. All experimental protocols were approved by the Animal Care and Use Committee of the Second Affiliated Hospital of Nanchang University (No:2023095).

The learning and memory capacities of mice were analyzed by MWM assay using a steel pool (120 cm diameter $\times$ 40 cm deep). A platform (10 cm diameter) was submerged 1 cm underwater. The mice in the pool facing the wall were trained to find the platform in 5 consecutive days. The amount of time spent searching the platform, and swimming distance were noted and quantitatively analyzed. In addition, the mice were put in the quadrant farthest away from the platform. The percentage of time spent swimming in target quadrant and crossing the platform was analyzed.

The pathological changes in hippocampal tissues were determined by HE staining. The hippocampal tissues were fixed in 4% paraformaldehyde, embedded in paraffins, and sectioned into 5 µm slices. The slices were subjected to HE staining using the Hematoxylin and Eosin Staining Kit (C0105S, Beyotime). The stained slices were observed under a light microscope (BX51M, Olympus, Japan). The number of damaged neurons was quantified as previously described [26].

All data from three independent experiments are expressed as mean $\pm$ standard deviation (SD). Statistical analysis was performed using one-way analysis of variance (ANOVA) followed by Tukey’s post-hoc test or Student’s t test using GraphPad Prism 7 software (GraphPad Software, Inc., San Diego, CA, USA). p value less than 0.05 was designated as statistically significant.

First, RT-qPCR validated a lower expression of METTL14 in the serum samples of AD patients in comparison with healthy volunteers (Fig. 1A). Besides, SH-SY5Y cells were stimulated with A$\beta$1-42 to simulate AD pathology in vitro. Consistently, METTL14 was down-regulated in A$\beta$1-42-exposed SH-SY5Y cells (Fig. 1A). To investigate the biological function of METTL14 in AD development, METTL14 was overexpressed in SH-SY5Y cells by transfection with OE-METTL14 plasmid (Fig. 1B). Consistently, Western blotting indicated that METTL4 protein level was significantly increased after OE-METTL4 transfection (Fig. 1C). A$\beta$1-42-induced reduction in cell viability was effectively restored by METTL14 overexpression (Fig. 1D). Additionally, the ROS level was elevated by A$\beta$1-42 exposure, which could be reduced in METTL14-overexpressed cells (Fig. 1E). Moreover, A$\beta$1-42 stimulation enhanced Fe²⁺ and total iron levels, whereas METTL14 overexpression abolished these changes (Fig. 1F,G). Furthermore, overexpression of METTL14 reversed A$\beta$1-42-mediated down-regulation of ferroptosis-related proteins GPX4 and SLC7A11 (Fig. 1H). ELISA results indicated that the enhanced concentrations of IL-1$\beta$, IL-6, and TNF-$\alpha$ in A$\beta$1-42-stimlated cells were evidently lowered by METTL14 overexpression (Fig. 1I). The raw data of Fig. 1 as Supplementary Materials 1. These observations proved that METTL14 overexpression protected against A$\beta$1–42-induced neurotoxicity and ferroptosis in SH-SY5Y cells.

Fig. 1.

METTL14 overexpression inhibited A$\beta$1-42-induced ferroptosis and inflammation. (A) RT-qPCR analysis of METTL14 level in serum samples (n = 18) or A$\beta$1-42-challenged SH-SY5Y cells (n = 3). (B) METTL14 expression after transfection with OE-METTL14 was assessed by RT-qPCR (n = 3). (C) Western blotting analysis of METTL14 protein level after oe-METTL4 transfection. (D) MTT assay determination of SH-SY5Y cell viability (n = 3). Levels of ROS (E), Fe²⁺ (F), and total iron (G) in cells from different groups were detected (n = 3). (H) Western blotting detected the expression of GPX4 and SLC7A11 (n = 3). (I) ELISA detection of IL-1$\beta$, IL-6, and TNF-$\alpha$ production (n = 3). * p 0.05; ** p 0.01; *** p 0.001. Student’s t test (for A, B) or one-way ANOVA followed by Tukey’s post-hoc test (for C–H) was performed to analyze data. AD, Alzheimer’s disease; METTL14, methyltransferase like 14; RT-qPCR, real-time Quantitative PCR; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5 diphenyltetrazolium bromide; ROS, reactive oxygen species; GPX4, glutathione peroxidase 4; SLC7A11, solute carrier family 7 member 11; ELISA, Enzyme-Linked Immunosorbent Assay; IL-1$\beta$, interleukin-1beta; IL-6, interleukin-6; TNF-$\alpha$, tumor necrosis factor-alpha.

To further explore the downstream regulatory mechanism of METTL14 in AD progression, TUG1 was focused on. The silencing efficiency of sh-METTL14 in SH-SY5Y cells was validated by RT-qPCR and Western blotting (Fig. 2A,B). Notably, TUG1 expression was increased in METTL14-silenced cells, but reduced in METTL14-overexpressed cells (Fig. 2C). We also found that total RNA m6A level was declined after METTL14 knockdown, while elevated after METTL14 overexpression (Fig. 2D). Of note, SRAMP database predicted several m6A modification sites on TUG1. Moreover, METTL14 overexpression significantly enhanced the m6A level of TUG1, while METTL14 knockdown resulted in the opposite result (Fig. 2E). In addition, the stability of TUG1 RNA in actinomycin D-treated SH-SY5Y cells was decreased in METTL14-overexpressed cells, but increased in METTL14-depleted cells (Fig. 2F). There was no significant difference between control and sh-NC/vector group. These results revealed that METTL14 facilitated the decay of TUG1 RNA via m6A modification.

Fig. 2.

METTL14 reduced the stability of TUG1 via m6A modification. Expression of METTL14 was detected by RT-qPCR (A) and Western blotting (B). (C) TUG1 expression after transfection was evaluated by RT-qPCR (n = 3). (D) Total RNA m6A level was assessed by the m6A RNA methylation quantification kit (n = 3). (E) MeRIP was adopted to determine m6A level of TUG1 (n = 3). (F) After exposure to actinomycin D, the remaining TUG1 expression was determined by RT-qPCR (n = 3). * p 0.05; ** p 0.01; *** p 0.001. One-way ANOVA followed by Tukey’s post-hoc test (for A–F) was performed to analyze data. TUG1, taurine upregulated gene 1; NC, negative control; m6A, N6-methyladenosine; MeRIP, methylated RNA immunoprecipitation sequencing-quantitative real-time polymerase chain reaction; ANOVA, analysis of variance.

Given that TUG1 was a downstream target of METTL14, we further investigated the regulatory role of TUG1 in AD progression. As compared with normal volunteers, TUG1 expression was higher in the serum samples of AD patients, as well as in A$\beta$1-42-stimulated SH-SY5Y cells (Fig. 3A). The knockdown efficiency of sh-TUG1 was confirmed by RT-qPCR (Fig. 3B). MTT data indicated that TUG1 silencing restored A$\beta$1-42-induced declined viability (Fig. 3C). The elevated ROS, Fe²⁺ and total iron levels in A$\beta$1-42-treated cells were effectively reduced by TUG1 silencing (Fig. 3D–F). Additionally, down-regulation of ferroptosis-related proteins GPX4 and SLC7A11 by A$\beta$1-42 exposure was counteracted by TUG1 silencing (Fig. 3G). Moreover, knockdown of TUG1 inhibited A$\beta$1–42-mediated release of IL-1$\beta$, IL-6 and TNF-$\alpha$ (Fig. 3H). The raw data of Fig. 3 as Supplementary Materials 3. Taken together, A$\beta$1-42-mediated ferroptosis and neuronal damage were restrained after TUG1 knockdown.

Fig. 3.

A$\beta$1-42-triggerred ferroptosis was restrained by TUG1 silencing. (A) RT-qPCR analysis of TUG1 expression in serum samples (n = 18) or A$\beta$1-42-exposed cells (n = 3). (B) TUG1 abundance in sh-TUG1 transfected cells was detected by RT-qPCR (n = 3). (C) Cell viability was evaluated by MTT assay (n = 3). Levels of ROS (D), Fe²⁺ (E), and total iron (F) after various treatments were measured (n = 3). (G) Protein abundance of GPX4 and SLC7A11 was determined by Western blotting (n = 3). (H) IL-1$\beta$, IL-6, and TNF-$\alpha$ production was evaluated by ELISA (n = 3). * p 0.05; ** p 0.01; *** p 0.001. Student’s t test (for A, B) or one-way ANOVA followed by Tukey’s post-hoc test (for C–H) was performed to analyze data.

Next, the underlying molecular mechanism of TUG1 in AD was explored. As predicted by RPISeq database, TUG1 might interact with SMURF1 protein (interaction score, 0.7; SVM 0.99). RNA pull-down assay suggested that SMURF1 protein could be immunoprecipitated by TUG1 probe (Fig. 4A). Furthermore, RIP assay demonstrated that the enrichment of TUG1 was elevated after immunoprecipitation with anti-SMURF1 antibody (Fig. 4B). These results proved that there was a direct interaction between TUG1 and SMURF1. In addition, ubibrowser predicted that SMURF1 was a potential E3 ubiquitin ligase of GDF15. Co-IP assay further verified that SMURF1 could directly interplay with GDF15 protein (Fig. 4C). Transfection with sh-SMURF1 evidently reduced SMURF1 mRNA and protein levels in SH-SY5Y cells (Fig. 4D,E). Furthermore, SMURF1 silencing lowered the ubiquitin level of GDF15 and consequently enhanced GDF15 protein level in SH-SY5Y cells (Fig. 4F). Besides, silencing of TUG1 or SMURF1 reduced SMURF1 protein level, but enhanced GDF15, NRF2, and HO-1 protein levels in SH-SY5Y cells, which were reinforced by co-transfection with sh-TUG1 and sh-SMURF1 (Fig. 4G). Collectively, TUG1 depletion activated GDF15/NRF2 pathway via repressing SMURF1-mediated ubiquitination of GDF15.

Fig. 4.

TUG1 knockdown activated GDF15/NRF2 pathway through interaction with SMURF1. The direct interaction between TUG1 and SMURF1 was validated by RNA pull down (A) and RIP assays (B) (n = 3). (C) The interaction between SMURF1 and GDF15 proteins was validated by Co-IP (n = 3). (D,E) RT-qPCR and Western blotting analysis of SMURF1 mRNA level (n = 3). (F) The ubiquitin level of GDF15 was assessed by Co-IP (n = 3). (G) Western blotting determination of SMURF1, GDF15, NRF2, and HO-1 protein levels (n = 3). * p 0.05; ** p 0.01; *** p 0.001. Student’s t test (for B, D, E) or one-way ANOVA followed by Tukey’s post-hoc test (for G) was performed to analyze data. GDF15, growth differentiation factor 15; SMURF1, Smad ubiquitin regulatory factor 1; NRF2, inactivated nuclear factor erythroid 2-related factor 2; HO-1, heme oxygenase-1; RIP, RNA immunoprecipitation; Co-IP, co-immunoprecipitation.

Given that GDF15 was modulated by TUG1 in SH-SY5Y cells, we further investigated the involvement of GDF15 in the influence of TUG1 on AD progression. For this purpose, A$\beta$1–42-exposed SH-SY5Y cells were transfected with sh-TUG1, sh-GDF15, or a combination of them. As illustrated in Fig. 5A,B, GDF15 mRNA and protein levels were declined by transfection with sh-GDF15. As shown in Fig. 5C,D, GDF15 was down-regulated by A$\beta$1–42 treatment, which was intensified by sh-GDF15 transfection, but reversed by sh-TUG1 transfection. TUG1-mediated up-regulation of GDF15 was abolished by sh-GDF15 upon A$\beta$1–42 stimulation. Moreover, the viability of A$\beta$1-42-exposed SH-SY5Y cells was reduced by GDF15 knockdown, and TUG1 knockdown-mediated promotion in cell viability was reversed by GDF15 depletion (Fig. 5E). In addition, GDF15 knockdown enhanced ROS level of A$\beta$1-42-exposed SH-SY5Y cells, and reversed the decreased ROS level in TUG1-silenced cells (Fig. 5F). Accordingly, GDF15 down-regulation raised Fe²⁺ and total iron levels and abolished TUG1 silencing-induced reduction in Fe²⁺ and total iron levels (Fig. 5G,H). sh-GDF15 down-regulated GPX4 and SLC7A11 in A$\beta$1-42-exposed SH-SY5Y cells, and the promotive effect of sh-TUG1 on GPX4 and SLC7A11 expression was counteracted by sh-GDF15 co-transfection (Fig. 5I). Furthermore, the production of IL-1$\beta$, IL-6, and TNF-$\alpha$ was repressed by TUG1 interference, but promoted by GDF15 silencing. The inhibitory effect of TUG1 knockdown on IL-1$\beta$, IL-6, and TNF-$\alpha$ production was abrogated by GDF15 depletion (Fig. 5J). These results suggested that GDF15 was involved in TUG1 knockdown-mediated protection against ferroptosis in the in vitro model of AD.

Fig. 5.

TUG1 depletion repressed ferroptosis via activating GDF15. GDF15 expression level was detected by RT-qPCR (A) and Western blotting (B) (n = 3). (C,D) RT-qPCR and Western blotting analysis of GDF15 expression in cell from various groups. (E) Cell viability was determined by MTT assay (n = 3). Levels of ROS (F), Fe²⁺ (G), and total iron (H) in cells received multiple treatments were assessed (n = 3). (I) Protein levels of GPX4 and SLC7A11 was measured by Western blotting (n = 3). (J) IL-1$\beta$, IL-6, and TNF-$\alpha$ release was assessed by ELISA (n = 3). * p 0.05; ** p 0.01; *** p 0.001. Student’s t test (for A, B) or one-way ANOVA followed by Tukey’s post-hoc test (for C–J) was performed to analyze data.

Finally, the protective effect of Mettl14 was validated in AD mice in vivo. As evaluated by MWM, AD mice exhibited obvious memory impairment, however; Mettl14 overexpression effectively improved behavioral deficits of AD mice (Fig. 6A–D). HE staining indicated intact structure of hippocampal neurons, and no obvious necrosis was observed in WT mice. However, abnormal shape, unordered arrangement, and apoptosis were found in hippocampal neurons of AD mice (Fig. 6E). These pathological changes in hippocampus were prominently relieved by Mettl14 overexpression (Fig. 6E). Moreover, the damaged neurons were indicated as red arrows and quantified. We found that the number of damaged neurons was strikingly increased, which was reduced after Mettl14 overexpression (Fig. 6E). In addition, METTL14 and Gdf15 mRNA levels were down-regulated, while Tug1 level was up-regulated in the hippocampal tissues of AD mice, whereas Mettl14 overexpression reversed these changes (Fig. 6F). Furthermore, the protein levels of Gdf15, Nrf2, HO-1, Gpx4 and Slc7A11 were reduced in the hippocampal tissues of AD group, which could be counteracted after Mettl14 overexpression (Fig. 6G). RT-qPCR analysis showed that the increased Il-1$\beta$, Il-6, and Tnf-$\alpha$ concentrations in the hippocampal tissues of AD mice were partly reversed by Mettl14 overexpression (Fig. 6H). To sum up, Mettl14 overexpression ameliorated cognitive impairment and ferroptosis of AD mice through the regulation of Tug1/Gdf15/Nrf2 pathway.

Fig. 6.

Mettl14 overexpression improved cognitive disorder and suppressed ferroptosis in AD mice via TUG1/Gdf15/Nrf2 pathway. (A–D) Morris water maze was performed to evaluate the learning and memory capacities of mice (n = 8). (E) The pathological changes in hippocampus were observed by HE staining (n = 8). Scale bars: 100 µm or 25 µm. (F) Expression of Mettl14, Tug1, and Gdf15 was detected by RT-qPCR (n = 8). (G) Protein abundance of Gdf15, Nrf2, HO-1, Gpx4, and Slc7a11 was determined by Western blotting (n = 8). (H) The mRNA levels of Il-1$\beta$, Il-6, and Tnf-$\alpha$ were measured by RT-qPCR (n = 8). * p 0.05; ** p 0.01; *** p 0.001. One-way ANOVA followed by Tukey’s post-hoc test (for A–H) was performed to analyze data. WT, wild type.