Porphyromonas gingivalis lipopolysaccharide induces cognitive dysfunction, mediated by neuronal inflammation via activation of the TLR4 signaling pathway in C57BL/6 mice

The OFT was used to assess spontaneous activity, anxiety-like behavior, and emotional change in the animals [16]. When placed in an open field for the first time, mice tend to remain along the periphery of the apparatus because of fear. Therefore, anxiety-like behavior could be assessed, based on the extent of exploratory behavior observed. The open field in the present study consisted of a rectangular arena (278 × 236 mm), enclosed by a white wall, 30 cm in height. The test was initiated by gently placing a single mouse in the middle of the arena, allowing the animal to move freely for 5 min, while being recorded.

The MWM test is used to test spatial learning and memory [17]. The MWM equipment (Datum Mobile, Minhang, Shanghai, China) consisted of a circular pool with a diameter of 120 cm and a depth of 50 cm. The pool was filled with water until a platform (9 cm in diameter) in the pool was submerged 1 cm below the surface. The water temperature was controlled to remain with a range equivalent to that of room temperature (22 ± 1 °C). A video analysis system was used to monitor and record the swimming activity of the mice.

The mice should have learned to use the visual tips around the pool to find the hidden platforms within 90 s. If a mouse could not find the platform within 90 s, it was guided to the platform and allowed to stay there for 30 s. Each mouse was trained four times a day with 30 s of rest per training interval. The probe trial was carried out on the sixth day. The platform was removed from the pool, and the mice were then placed into the water at the two quadrants furthest from the platform used on days 1–5. Each mouse was allowed to navigate for 60 s.

The PAT was used to evaluate passive avoidance learning in mice [18]. Mice tend to prefer darkness; thus, the device was divided by a retractable door into two compartments: a brightly lit compartment and a dark compartment. Upon completely entering the dark compartment, the animal received an electric shock (0.6 mA, 2 s).

The first day of the study was designated training day. Each mouse was placed in the light compartment, and 10 s later, the door was opened. After 5 min, the mouse was removed. These actions were repeated for each mouse. When the mouse entered the dark compartment, it immediately received the electric shock. The time taken (latency) for the mouse to enter the dark compartment was recorded. A retention test was again conducted 24 h later. The latency to re-enter the dark compartment and the number of electrical shocks (error times) within 5 min were recorded.

The tissues of the cortex of mice were homogenized, and RNA was isolated using the Total RNA Kit (Omega Bio-Tek, Inc., Norcross, GA, USA). The RNA was reverse transcribed into cDNA using the M-MLV reverse transcriptase kit (Invitrogen, Waltham, MA, USA). The sequences of these primers were as follows: TNF-α, 5′-GACAAGGCTGCCCCGACTACG-3′ (forward) and 5′-CTTGGGGCAGGGGCTCTTGAC-3′ (reverse); IL-1β, 5′-AAATCTCGCAGC AGCACATCAA-3′ (forward) and 5′-CCACGGGAAAGACACAGGTAGC-3′ (reverse); IL-6, 5′-AGTTGCCTTCTTGGGACTGA-3′ (forward) and 5′-CCTCC GACTTGTGAAGTGGT-3′ (reverse); IL-8, 5′-CAGCTGCCTTAACCCCATCA-3′ (forward) and 5′-CTTGAGAAGTCCATGGCGAAA-3′ (reverse); TLR2, 5′-CTC CTGAAGCTGTTGCGTT AC-3′ (forward) and 5′-TACTTTACCCAGCTCGCTC ACTAC-3′ (reverse); TLR3, 5′-CAGGATACTTGATCTCGGCCTT-3′ (forward) and 5′-TGGCCGCTGAGTTTTTGTTC-3′ (reverse); TLR4, 5′-CGCTTTCACCT CTGCCTTCACTACAG-3′ (forward) and 5′-ACACTACCACAATAACCTTCCG GCTC-3′ (reverse); CD14, 5′-TGTCGTGGGCAACAAGGGATG-3′ (forward) and 5′-AAGGTGGAGAGGGCAGGGAAGA-3′ (reverse); and actin, 5′-CATCCGTA AAGACCTCTATGCCAAC-3′ (forward) and 5′-ATGGAGCCACCGATCCACA-3′ (reverse). Subsequently, RT-PCR analyses were performed using the SYBR premix EX Taq™ (Takara, Kusatsu, Shiga, Japan) according to the manufacturer’s protocol in a Light Cycler 480 System (Roche Diagnostics, Grenzacherstrasse, Basel, Switzerland). The DNA amplification was performed as follows: the first cycle was maintained at 95 °C for 30 s, followed by 40 cycles consisting of denaturation (95 °C for 10 s), annealing (60 °C for 20 s), and extension (72 °C for 15 s). The TNF-α, IL-1β, and IL-8 were then processed using the 2^−ΔΔCt method, during which a single calibrated sample was compared against the gene expression of every unknown sample.

Radioimmunoprecipitation assay (RIPA) lysis buffer (Beyotime, Beijing, China), 1% protease inhibitor cocktail (Sigma, St. Louis, MO, USA), and 1% phenylmethylsulfonyl fluoride (PMSF; Beyotime, Beijing, China) were used to homogenize samples of the cerebral cortex. Protein quantification was performed using the BCA (bicinchoninic acid) Protein Assay Kit (Thermo Scientific, Vernon Hills, IL, USA). For quantification by ELISA, commercially available ELISA kits were used according to the manufacturer’s instructions to measure levels of TNF-α, IL-6, IL-8 (UBI, Sunnyvale, CA, USA), and IL-1β (Westang, Yangpu, Shanghai, China) in the cortex.

The RIPA lysis buffer, in addition to 1% protease inhibitor cocktail and 1% PMSF, was used to homogenize samples of the cerebral cortex. Following the addition of sodium dodecyl sulfate (SDS) loading buffer, the samples were boiled for 5 min, and proteins were subsequently detected by western blot analysis. The proteins were transferred to a polyvinylidene difluoride (PVDF) membrane after separation. A wide range of protein markers was run in parallel to detect the molecular weight of proteins. Skimmed milk was used for membrane blockage to reduce nonspecific binding. Proteins were probed with anti-TLR4 (1:1000, ab13867; Abcam, Cambridge, UK), anti-CD14 (1:500, ab203294; Abcam), anti-IRAK1 (1:1000, no. 4504; Cell Signaling Technology, Danvers, MA, USA), anti-p65 (1:1000, no. 8242; Cell Signaling Technology), anti-phospho-p65 (1:1000, no. 3033; Cell Signaling Technology), and anti-β-actin (1:1000, abs119600; Abcam). The data were quantified using the Image Studio Lite ver. 5.2 software.

Mice were anesthetized with chloral hydrate and perfused with physiological saline before sacrifice. Immediately after removal of the brain, one hemisphere was placed in 4% paraformaldehyde and left overnight at 4 °C, after which paraffin sections were prepared. The cortex of the other hemisphere was immediately removed and stored at − 80 °C for subsequent ELISA, RT-PCR, and western blot analyses.

Brain sections were incubated with 3% H₂O₂ in methanol for 10 min to quench endogenous peroxidase activity. They were then rinsed with phosphate-buffered saline (PBS), blocked with 10% goat serum for 30 min, and incubated overnight at 4 °C with the following primary antibodies: rabbit anti-glial fibrillary acidic protein (GFAP) (1:400, ab7260; Abcam) and goat Iba1 (1:400, ARG63338; Arigo Biolaboratories, Hsinchu City, Taiwan, China). After being washed, sections were incubated with biotinylated goat anti-rabbit or goat secondary antibody (1:200; Vector Laboratories, Burlingame, CA, USA). After being rinsed with PBS, streptavidin-labeled peroxidase was added and left for 30 min. This was followed by further rinsing, after which newly prepared 3,3′-diaminobenzidine (DAB) solution was added, and the mixture was left for the reaction to develop. The sections were dyed with hematoxylin and dipped in 1% hydrochloric acid in alcohol for differentiation. They were then washed in ammonia and stained blue, after which they were rinsed with water.

For cell counting, images were obtained with a Nikon camera. The numbers of GFAP-positive cells were determined by counting positive cells in two areas of each section in every tenth serial coronal section. At least three coronal sections were analyzed per mouse, and the average of the individual measurements was used to calculate group means [19].

Data were presented as mean ± standard error of the mean (SEM). Statistical analyses were performed using repeated measures ANOVA, one-way ANOVA, and Student’s t test with the GraphPad Prism software. An analysis of variance was performed using Tukey’s post hoc multiple comparison test. A value of p < 0.05 was indicative of statistical significance.

As shown in Fig. 2, the behaviors of mice (total distance covered, percentage of distance covered on the central grid, percentage of time spent on the central grid, number of defecations, number of rearings, and frequency of grooming) showed no significant differences among all groups in the OFT. The behavioral performance indicated that the activity of the mice was not affected by P. gingivalis-LPS.

To assess spatial learning acquisition, mice were trained in the MWM task during four trials per day for five consecutive days. In all groups, the escape latency showed a marked decline over the training days. In the P. gingivalis-LPS group, the escape latency was evidently longer than that in the control group at day 3 (64.25 ± 5.97 versus 47.91 ± 4.93 s, respectively), day 4 (61.81 ± 7.07 versus 35.36 ± 3.90 s, respectively), and day 5 (62.69 ± 6.73 versus 24.06 ± 3.71 s, respectively). Moreover, the P. gingivalis-LPS plus TAK group showed a noticeably shorter latency in comparison to the P. gingivalis-LPS group at day 4 (41.33 ± 6.53 versus 61.81 ± 7.07 s, respectively) and day 5 (36.54 ± 5.83 versus 62.69 ± 6.73 s, respectively). No significant differences were observed between the P. gingivalis-LPS group and E. coli-LPS group (Fig. 3). Results of the probe trial illustrate that the learning ability of mice can be impaired by P. gingivalis-LPS. However, this change was significantly prevented by TAK-242.

During the spatial probe test, the number of platform crossings was significantly reduced in the P. gingivalis-LPS group, in comparison to the control group (Fig. 4c). Furthermore, the P. gingivalis-LPS group traveled a shorter distance (18.05 ± 2.66 versus 32.95 ± 2.85% for the control group) and spent less time (18.36 ± 2.91 versus 36.39 ± 3.12% for the control group) in the target quadrant (Fig. 4a, b). In contrast, mice in the P. gingivalis-LPS plus TAK group traveled a longer distance (34.21 ± 4.02 versus 18.05 ± 2.66% for the P. gingivalis-LPS group) and spent more time (35.24 ± 4.75 versus 18.36 ± 2.91% for the P. gingivalis-LPS group) in the target quadrant than those in the P. gingivalis-LPS group. No significant differences were found between the P. gingivalis-LPS group and E. coli-LPS group (Fig. 4). Results of the probe trial illustrate that the learning ability of mice can be impaired by P. gingivalis-LPS. These effects, however, were significantly prevented by TAK-242.

During the first day of training, no significant differences were observed in the latency of each group to enter the dark compartment. After the electric shock was given, the memory was retained the following day. As shown in Fig. 5, the P. gingivalis-LPS group showed a reduced latency to enter the dark compartment as compared with the control group (54.00 ± 7.67 versus 165.90 ± 14.68 s for the control group) and this was significantly prevented by TAK-242 (134.20 ± 24.42 s). Furthermore, The P. gingivalis-LPS group showed more error times to enter the dark compartment than the control group (2.18 ± 0.23 versus 1.07 ± 0.12 for the control group) and this was significantly prevented by TAK-242 (1.22 ± 0.22). No significant differences were observed between the P. gingivalis-LPS group and E. coli-LPS group. The PAT results indicated that passive avoidance learning in mice could be impaired by P. gingivalis-LPS. These effects, however, were significantly prevented by TAK-242.

The activated microglia, characterized by irregular protrusions and increased volume of the cell bodies, were positively stained with ionized calcium-binding adaptor molecule 1 (Iba1). Activated microglia were observed in the hippocampus and cortex of the P. gingivalis-LPS group and E. coli-LPS group, whereas activated microglia were rarely observed in the control group (Figs. 6 and 7).

The activated astrocytes were positively stained with GFAP, and cell volume showed a relative increase with hypertrophy and irregular protrusions. In comparison to the control group, the P. gingivalis-LPS group and E. coli-LPS group had a greater number of activated astrocytes in both the hippocampus and cortex (Fig. 8a). The number of GFAP-positive cells in the P. gingivalis-LPS group and the E. coli-LPS group was significantly greater in both the cortex and hippocampus, in comparison to the control group. No significant differences were observed between the P. gingivalis-LPS group and E. coli-LPS group (Fig. 8b).

The RT-PCR assays were performed for mRNA expression of TNF-α, IL-1β, IL-6, and IL-8 genes. The P. gingivalis-LPS group showed an approximately fourfold increase in TNF-α mRNA expression, threefold increase in IL-1β mRNA expression, twofold increase in IL-6 mRNA expression, and fourfold increase in IL-8 mRNA expression, in comparison to the control group. The P. gingivalis-LPS plus TAK group showed reduced TNF-α, IL-1β, IL-6, and IL-8 mRNA expression in comparison to the P. gingivalis-LPS group. No significant differences were observed between the P. gingivalis-LPS group and E. coli-LPS group (Fig. 9a, c, e, g).

The concentration levels of TNF-α, IL-1β, IL-6, and IL-8 were detected by ELISA. The expression of TNF-α, IL-1β, IL-6, and IL-8 proteins in the P. gingivalis-LPS group was significantly higher than that in the control group. In comparison to the P. gingivalis-LPS group, the P. gingivalis-LPS plus TAK group showed reduced expression of TNF-α, IL-1β, IL-6, and IL-8 proteins. No significant differences were observed between the P. gingivalis-LPS group and E. coli-LPS group (Fig. 9b, d, f, h).

The RT-PCR assays were performed for TLR2, TLR3, TLR4, and CD14 genes. The P. gingivalis-LPS group induced an approximately twofold increase in TLR4 and CD14 mRNA expression, in comparison to the control group. However, no significant differences were observed in TLR2 and TLR3 mRNA expression in comparison to the control group (Fig. 10a, b). The P. gingivalis-LPS plus TAK group showed reduced TLR4 and CD14 mRNA expression in comparison to the P. gingivalis-LPS group. No significant differences in TLR4 and CD14 mRNA expression were observed between the P. gingivalis-LPS group and E. coli-LPS group (Fig. 10c, d). The E. coli-LPS group showed an approximately 2.5-fold increase in TLR2 mRNA expression in comparison to the control group, whereas no significant differences were observed between the P. gingivalis-LPS group and control group (Fig. 10a). High expression of TLR4 and CD14 genes was promoted by P. gingivalis-LPS. The changes induced by P. gingivalis-LPS were significantly prevented by TAK-242.

As shown in Fig. 11, western blot analysis was performed to study the underlying mechanism of P. gingivalis-LPS-induced neuroinflammation. We detected elevated protein expression of TLR4, CD14, IRAK1, and p-p65/p65 in P. gingivalis-LPS-treated and E. coli-LPS-treated mice, in comparison to the control mice. The elevated expression of TLR4, IRAK1, and p-p65/p65 proteins induced by P. gingivalis-LPS was attenuated by TAK-242 (Fig. 11b, d, e). Overall, these results provide evidence for the involvement of P. gingivalis-LPS in the downstream signal processing of the TLR4/CD14 receptor, through IRAK1 signal processing, and the final activation of the NF-κB signaling pathway.