Eight-week-old wild-type C57BL/6 (WT) male mice and male NLRP3 knockout (NLRP3^−/−) mice on the C57BL/6J genetic background mice were purchased from the Jiangsu ALF Biotechnology Co., Ltd (Nanjing, China). Our laboratory extracted DNLA from Dendrobium nobile stem parts. According to LC–MS/MS, the alkaloids accounted for 79.8%. The chemical structures and the chromatograms of DNLA were shown in our latest articles (Nie et al., 2016). DNLA was dissolved in 1% Tween 80 and was orally administered to each group daily, and the equal volume of vehicle was used as control. WT and NLRP3 KO mice received intragastric administration with DNLA (20 or 40 mg/kg) once daily for 14 days. On the seventh day, the mice were injected with LPS (2 μg in 1 μL normal saline, Sigma-Aldrich, United States) in the hippocampus on both sides of the brain with the following Bregma coordinates: AP 0.5 mm, ML 1.0 mm, and DV −2.0 mm; animals in the control group accepted equal volume of saline. All experiments were in strict conformity with the Chinese Guidelines of Animal Care and Welfare, and this study was approved by the Animal Care and Use Committee of Zunyi Medical University (Zunyi, China).

Behavioral changes in the spontaneous alternation were assessed via the Y-maze test (Liu B. et al., 2020). The mice were placed in the start area and permitted to freely explore the left or right arm of the maze for 5 min, while the sequence of the arm entrances and total numbers of arm choices were monitored and recorded using a camcorder (TopScan, United States). The total number of arm entries and alternation behavior was counted, and the percentage of spontaneous alternation was calculated by the formula [number of alternation/(number of total arm entries −2)] × 100%.

Before the novel object recognition test, the mice were acclimated to the open field testing arena for 5 min on day 1. On the second day, the mice were allowed freely to explore two identical objects in the open field for 5 min (training period). On the third day, the mice were allowed to freely explore both objects (one of the objects is novel) for 5 min (test period). The time spent in exploring the objects was measured. The location preference (to assess the impact of position on preference) is described as the recognition index, which is calculated by the time taken to explore novel objects/total time taken to explore both objects *100%.

All mice were sacrificed after behavior tests. Mice brains were fixed with 4% paraformaldehyde, and brain slices (4 μm thick) were prepared and used for staining. The slices were washed in PBS for 5 min three times and then blotted in 5% goat serum for 1 h at room temperature. The slices were incubated with the primary antibodies overnight at 4°C. The primary antibodies used were: rabbit anti-IBA-1 (1:200, Abcam, United Kingdom), rabbit anti-NLRP3 (1:200, Novus Biologicals, United States), rabbit anti-caspase-1, rabbit anti-GSDMD-N, and rabbit anti-NeuN (1:100, Proteintech, China). After being washed in PBS three times, these slices were incubated with secondary antibodies at 37°C for 1 h, and the secondary antibodies used were Alexa Fluor 488-conjugated goat anti-rabbit IgG (1:400, Abcam, United Kingdom) and Alexa Fluor 594-conjugated goat anti-rabbit IgG (1:400, Abcam, United States). Then, these slices were washed three times again and stained with DAPI (Solarbio, China) for 5 min. The results were imaged using a fluorescence microscope (Nikon, Japan).

Protein from hippocampus tissue was extracted using lysing buffer according to the protocol from the previous study (Liu B. et al., 2020). The protein was quantified using a BCA protein assay kit (Generay, China) after centrifugation (12,000 rpm, 15 min at 4°C). Protein (30 µg) was loaded onto 10% Tris–glycine polyacrylamide gels and then transferred onto the PVDF membrane (Millipore Trading Co., Ltd.). The membranes were soaked in 5% fat-free milk for 2 h at room temperature and then incubated with primary antibodies overnight at 4°C: anti-rabbit -NLRP3 (1:1000, Novus Biologicals, United States), anti-rabbit ASC (1:1000, Proteintech, China), anti-rabbit-Caspase-1 (1:1000, Proteintech, China), anti-rabbit GSDMD-N (1:1000, HuaAn, China), and anti-rabbit GAPDH (1:5000, Proteintech, China). After washing three times, the membranes were incubated with rabbit HRP-conjugated secondary antibody (1:5000, Proteintech, China) for 1 h at room temperature. The intensity of the blots was detected using a chemiluminescence kit and quantified using Image Lab (Bio-Rad, United States).

The total RNA of the hippocampus tissues was prepared using RNeasy kit, and the detailed steps of RT-PCR were described previously (Liu J. J. et al., 2020). NLRP3, IL-18, IL-1β, Caspase-1, and GAPDH genes were tested. Accordingly, the gene expression was normalized with GAPDH. The primers are listed as follows Table1:

The hippocampal tissues were dried in an increasing sequence of ethanol (30–100%), placed on aluminum stabs with their treated surfaces facing up, and sputtered with gold in the JEE-4X equipment (Jeol, Japan). The Hitachi-3000N scanning electron microscope (Hitachi High-Tech, Japan) was used for observation.

Data were presented as mean ± SEM. The significant statistical difference was analyzed using one-way ANOVA by Dunnett’s post hoc t-test and two-way ANOVA with the Bonferroni post hoc test for comparison between WT and NLRP3-KO mice groups. The differences were considered significant at p < 0.05. Analyses were performed using GraphPad Prism (GraphPad Software, Inc., United States).

Neuroprotective effects of DNLA on the LPS-induced learning and memory function disorder were investigated in WT mice. We analyzed mice learning and memory function changes via Y-maze and novel object recognition tests. As shown in the Y-maze result (Figure 1A), LPS reduced the number of alternations compared with the control group (p < 0.05), and DNLA improved spatial memory (p < 0.05). Similarly, the novel object recognition test indicated that the percentage of time to explore new objects was reduced in the LPS group compared with the control group (p < 0.05). However, DNLA treatment increased the percentage of time to explore new objects (p < 0.05, Figure 1B). These findings suggested that DNLA improved LPS-induced spatial learning and memory deficits.

Neuron-specific nuclear protein (NeuN) is a mature neuronal marker, and the number of NeuN-positive cells reflects the growth and development of neurons. As shown in Figures 2A,B, the hippocampus in control and DNLA alone groups was normal with high NeuN-positive immunoreactivity. However, the reduced neuron numbers occurred in the hippocampus CA1 area of the LPS group (p < 0.05). DNLA pretreatment protected neurons against LPS-induced neurotoxicity by enhancing the NeuN-positive neuronal number (p < 0.05).

Pyroptosis plays an important role in the occurrence of progression of AD. To examine whether pyroptosis affected the neuroprotective effect of DNLA in the WT mice, we assessed the cell membrane structure of hippocampal neurons using the scanning electron microscope. As shown in Figure 2C, the cell membrane of hippocampus neuronal disruption by pores was discerned in the LPS group, and DNLA pretreatment attenuated the LPS-induced neuron membrane damage. These results show that DNLA could generate neuroprotective effect against LPS-induced neurotoxicity.

Neuroinflammation is critical in the pathophysiological process of AD. Activation of the NLRP3 inflammasome and microglia plays a key role in the regulation of neuroinflammatory response. To investigate the effects of DNLA on microglia-induced neuroinflammation, the co-localization of microglia and NLRP3 was determined by double immunofluorescence analysis (Figure 3A). In parallel with neuronal damage, LPS induced microglia and NLRP3 activation (p < 0.05), and DNLA suppressed these activations (p < 0.05, Figure 3B). Furthermore, we revealed that DNLA downregulated the mRNA expressions of NLRP3, caspase-1, IL-1β, and IL-18 in the hippocampus (p < 0.05, Figure 3C). Similar results were shown in protein expression of NLRP3, ASC, and caspase-1 in Figure 4A. Together, DNLA treatment attenuated LPS-induced neuroinflammation by inhibiting microglia, NLRP3 inflammasome activation, and release of pro-inflammatory factors.

To confirm whether DNLA inhibited the NLRP3/GSDMD signaling pathway in mouse hippocampus, immunofluorescence analysis was performed to determine the expression of caspase-1 and GSDMD-N. As shown in Figures 4A,B, the fluorescence intensity of caspase-1 and GSDMD-N was more prominent after LPS treatment (p < 0.05). Compared with the LPS group, DNLA reduced the protein expressions of caspase-1 and GSDMD-N (p < 0.05). Moreover, the protein expression of the NLRP3/GSDMD-N signaling pathway was measured by Western blotting; the higher level of GSDMD-N, caspase-1, ASC, and NLRP3 protein expressions was observed in the LPS group (p < 0.05), which were inhibited by DNLA treatment (p < 0.05, Figures 4C,D).

Additionally, the results further exhibited the strong binding affinity between DNLA and NLRP3 with a binding energy of −6.15 kcal/mol. Thereafter, the presumptive binding modes and the pocket of the amino acid were tested by molecular docking, including Val72, Trp73, Tyr84, Ala87, Glu91, and Lys88, which further confirmed that DNLA bound to the hydrophobic pocket of NLRP3, which is consistent with the results in vivo (Figure 5). These findings indicated that DNLA might directly bind to NLRP3 to exert its pharmacological activities.

Since DNLA ameliorated NLRP3 inflammasome activation, the neuroprotection effect of DNLA was further evaluated by using NLRP3 KO mice. First, the DNLA-ameliorated LPS-induced learning and memory function disorder was not discerned in NLRP3 KO mice (Figures 6A,B). In addition, DNLA-mediated neuroprotection was absent in NLRP3 KO mice via NeuN immunostaining in the hippocampus. In detail, as shown in Figure 6C, LPS injection caused neuronal death in WT mice, while DNLA-attenuated neuronal loss in NLRP3 KO mice appeared to be less evident than that in WT mice. Similar results were shown in the cell membrane structure of hippocampal neurons using the scanning electron microscope (Figure 7A). Furthermore, DNLA reduced LPS-induced protein expressions of caspase-1 and GSDMD-N in WT mice, whereas DNLA-mediated reduction expressions of caspase-1 and GSDMD-N were not discerned in NLRP3 KO mice (Figures 7B–D). Collectively, these results suggested that inhibition of NLRP3 expression participated in DNLA-mediated neuroprotection in mice hippocampus.

The results further demonstrated that knockdown of NLRP3 attenuates GSDMD-N expression. Furthermore, the docking of NLRP3 and GSDMD-N was carried out by the ZDOCK method same as the previous study. The results showed that there were strong interactions between NLRP3 and GSDMD-N as evidenced by the output value score of 1330.175 (Figure 8). Of note, the results further demonstrated that hydrogen bond and charge interactions were generated through the amino acid residues including Gln45, Ile59, Arg100, Val101, Ser102, Asn103 and Thr105 in NLRP3 and Asp20, Thr22, Arg41, Tyr53, Arg212, Asp227, and Gln248 in GSDMD-N. These findings suggested that there existed a potent affinity between NLRP3 and GSDMD-N.