SCFAs promote intestinal double-negative T cells to regulate the inflammatory response mediated by NLRP3 inflammasome

SCFAs promote DNTs formation and NLRP3 expression in APP/PS1 mice

The animal experiment was approved by the ethics committee of Jiaxing University, Animal experiments conform to animal ethics and animal welfare regulations.

Wild-type C57BL/6J mice (WT mice) and APP/PS1 double-transgenic AD mice (AD mice) were selected. Four-month old AD mice showed senile plaque (SP), cognitive impairment and behavior disorder. Mice were purchased from Beijing Biocytogen Co., Ltd (Beijing, China). Mice were divided into WT, AD, WT + SCFAs and AD + SCFAs groups. Mice in the WT and AD groups were conventionally reared (without SCFAs in the feed). And mice in the WT + SCFAs and AD + SCFAs groups were intervened with SCFAs. Sodium acetate, sodium propionate and sodium butyrate were mixed at a mass ratio of 5%, 4.3% and 3.7%. Four groups of mice were raised in the same environment and measured as follows:

Morris water maze

The Morris water maze and video system were purchased from Feidi Biotechnology Co., Ltd. (Guangzhou, China). Mice were subjected to adaptive training one day before the experiment. In brief, mice entered from the entrance of the water maze and swam freely for 60 s. Mice were allowed to find the platform by themselves and mice were allowed to stand on the platform for 20 s, and mice were later put back into the cage.

Navigation test: The test lasted for 5 days, and the platform was placed in the fourth quadrant. The timer was started when the mouse entered from the entrance. And the time was recorded after the mouse found the platform and boarded the platform. If the mouse cannot find the platform within 60 s, the mouse was guided to the platform and stand for 20 s. between entrance into water and finding the platform was Escape latency (EL).

Space exploration assay: After removing the platform, and the times of crossing the fourth quadrant and retention time on the original platform within 60 s were recorded after entrance.

Detection of the proportion of DNTs in the spleen and intestines

Mice were fed with SCFAs for 30 days and were sacrificed by carbon dioxide asphyxiation, followed by extraction of spleen and colon tissues. Preparation of cell suspension by homogenization and adjust the cell concentration to 10⁷/ml. Afterwards, cell suspension was incubated with 20 μl of PC5-CD3, FITC-CD4 and PE-CD8 monoclonal antibodies (BD, MA, USA) at room temperature in the dark for 15 min, washed with pre-cooled PBS and centrifuged at 1500rpm. After mixing with 0.5 ml of PBS, samples were subjected to EPICS XLII flow cytometer (Beckman Coulter, CA, USA), followed by calculation of CD3⁺CD4⁻CD8⁻ cell ratio using CellQuest software.

The expression of NLRP3 and cleaved-Caspase-1 by immunohistochemistry (IHC)

After sacrificing by asphyxiation, the intestinal tissues of mice were fixed with 4% formaldehyde, embedded in paraffin and sliced, baked at 60°C for 2 h, rinsed in xylene and gradient alcohol. The slices were placed in 0.01 mol/L citrate buffer (PH = 6.0) for antigen retrieval at 98°C using microwave for 20 min. The slices were further incubated with 3% hydrogen peroxide at room temperature for 10 min to eliminate endogenous peroxidase, blocked with 2% bovine serum albumin (BSA) at room temperature for 30 min, incubated with anti-NLRP3 and anti-Caspase-1 and monoclonal antibodies (Abcam, MA, USA) (dilution 1:300). After washing with for 3 times, the slices were incubated with proper secondary antibody (Goat anti rabbit IgG H + L) at 37°C for 15 min, incubated with peroxidase-labeled streptomycin (Maxim Biotechnology Company, Fuzhou, China) for 15 min and wash with PBS for 3 times (5 min each). The slices were further visualized with freshly prepared DAB solution (Dako, Glostrup, Denmark), counterstained with hematoxylin and mounted. For negative control, primary antibody was replaced by TBS. The Olympus-DP72 image acquisition system and the Olympus-BX51 upright microscope of the CRi Nauance multispectral imaging system (Cambridge Research and Instrumentation, MA, USA) were used to acquire pictures and for quantitative analysis.

Enzyme-linked immunosorbent assay (ELISA)

After sacrificing, the peripheral blood of the tail vein from mice was collected and centrifuged to collect the upper serum. The intestinal tissue and brain tissue were ground in liquid nitrogen, added with 1.0 ml of pre-cooled RIPA lysis buffer (Beyotime Biotechnology Co., Ltd., Shanghai, China) on ice for 30 min and centrifuged at 10000g for 15 min. The supernatant was subjected to the protein quantitation bu using BCA kit (Beyotime Biotechnology Co., Ltd., Shanghai, China). After adjusting the protein concentration, the expression of IL-1β, IL-6, and TNF-α in the peripheral blood serum as well as protein from intestinal tissue and brain tissue was measured by ELISA kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer’s instruction.

Western blot (WB)

The protein solution extracted in the ELISA was used for WB. The 8–12% SDS-PAGE gel was prepared. The protein supernatant was mixed with 5x loading buffer (up to 20 μl). After boiling for 8 min, the protein sample was subjected to electrophoresis at 80V, further switching to 120V. The sample was subsequently transferred to membrane at 300mA constant current for 0.5–2 h. The membranes were blocked with 5% skimmed milk for 2 h and incubated with proper primary monoclonal antibodies diluted in TBST. The primary antibodies for intestinal tissue included NLRP3, cleaved-Caspase-1, ASC and Pro-Caspase-1 (dilution 1:500, Abcam, MA, USA). The primary antibodies for brain tissue included IBA-1 and GFAP (dilution 1:500, Abcam, MA, USA). After incubation with primary antibody, the membranes were reacted with HRP-labeled goat anti-rabbit secondary antibody (Abcam, MA, USA). After incubation, chemiluminescence was used for visualization, and Image Pro-Plus 6.0 software was used for optical density analysis.

The effect of Fas and TNFR1 inhibitors on the expression of NLRP3 in intestinal tissue induced by DNTs

In this study, AD mice were used. AD mice were divided into AD, SCFAs, SCFAs + Fas-IN and SCFAs + TNFR1-IN groups. Fas-IN and TNFR1-IN were inhibitors (MCE, Shanghai, China), which were administered by intragastric administration once a day (50 mg/kg). Three groups of mice were fed with SCFAs. After 30 days, the mice were sacrificed by carbon dioxide asphyxiation, followed by extraction of intestinal tissue, spleen, peripheral blood and brain tissue. The following assays were performed accordingly: (1) the proportion of DNTs in mouse intestine and spleen; (2) detection of the expression of NLRP3 and Caspase-1 in intestinal tissue by IHC; (3) the expression of inflammatory factors (IL-1β, IL-6 and TNF-α) in the peripheral blood serum as well as intestinal tissue and brain tissue by ELISA; (4) the protein expression of NLRP3, cleaved-Caspase-1, ASC, Pro-Caspase-1 in intestinal tissue as well as IBA-1 and GFAP in brain tissue by WB.

The influence and mechanism of SCFAs on the differentiation of DNTs in vitro

After sacrificing healthy C57BL/6J mice, the spleen was extracted and ground in liquid nitrogen to prepare cell suspension, followed by isolation of total lymphocytes by using lymphocyte separation solution (Dakewe Biotechnology Co., Ltd., Beijing, China). The 24-well plate was pre-coated with anti-CD3mAb (100 ng/ml) and incubated overnight. Lymphocytes were washed with 2% FBS in PBS, cultured in RPMI-1640 medium (Sigma, MA, USA) containing 20% FBS, IL-2 (50 U/ml) and IL-4 (30 U/ml). The culture medium was changed every three days. Trypan blue was used to detect cell viability. The lymphocytes were divided into control group (Con) and SCFAs group. Cells in the SCFAs group were co-cultured with acetic acid (5 mM), propionic acid (1 mM) and butyric acid (1 mM) for 24 h. Afterwards, cells of the two groups were collected and subjected to the following assays: (1) Detection of the proportion of DNTs by flow cytometry; (2) Detection of secretory protein of DNTs by ELISA; (3) Detection of protein expression by WB.

The role of OX40 signal in promoting the differentiation of DNTs in SCFAs

The lymphocytes isolated from mouse spleen were divided into Con group, SCFAs group and SCFAs + anti-OX40 group. Lymphocytes were cultured in RPMI 1640 medium containing 20% FBS, IL-2 (50 U/ml) and IL-4 (30 U/ml). Lymphocytes in the anti-OX40 group was pre-treated with OX40 antibody to inhibit OX40 activation (BioVisio, CA, USA). Afterwards, lymphocytes were subjected to the following assays accordingly: (1) detection of the proportion of DNTs by flow cytometry; (2) detection of the protein levels of OX40, Bcl-2, Bcl-xl and Survivin.

DNTs induce NLRP3 activation in intestinal macrophages in vitro

The Rosettesep antibody adsorption method was used to separate DNTs. The mouse intestinal macrophages (Procell Biotechnology Co., Ltd., Wuhan, China) were divided into Con group and DNTs group. Cells in the Con group were the intestinal macrophages cultured separately, and cells in the DNTs group were co-cultured macrophages and DNTs in Transwell chambers. After incubation for 24h, the following assays were performed.

Detection of the expression of NLRP3 and Caspase-1 by immunofluorescence (IF)

Intestinal macrophages were fixed with 4% formaldehyde at room temperature for 0.5 h, permeabilized with 0.2% Triton X-100 for 5 min, washed with PBS for 3 times, incubated with NLRP3 or Caspase-1 monoclonal antibody at 4°C overnight, wash with PBS twice, incubated with fluorescent secondary antibody and mounted with 95% glycerol, followed by observation under fluorescent microscope.

Statistical analysis

Measurement data were expressed as mean ± standard deviation (±s). two-way ANOVA was used for comparison between multiple groups, SNK test was utilized for comparison between groups. SPSS 18.0 software was used for statistical analysis. And P < 0.05 was considered to be statistical significance.

Effect of SCFAs on DNTs and NLRP3 inflammasome activation in AD mice

In the behavioral test of mice, we found that the cognitive ability of WT mice was not significantly changed before and after SCFAs intervention. The Morris water maze assay showed that EL, retention time and crossing time were not significantly different between the WT and WT + SCFAs groups. However, the cognitive ability of AD mice was significantly down-regulated after SCFAs intervention, which was significantly different from the AD group (Figure 1A–1C). To detect DNTs, the levels of DNTs in the spleen and intestine of WT and AD mice were relatively low, especially the DNTs in the spleen were not significantly different between WT and AD mice. SCFAs intervention could significantly increase the proportion of DNTs, especially the significantly up-regulated proportion of DNTs in the AD + SCFAs group, which was most remarkable in the intestine (Figure 1D–1F). In terms of inflammatory factors, the expression of inflammatory factors (IL-1β, IL-6 and TNF-α) in the brain tissue, peripheral blood and intestine of the AD group was higher than that in the WT group. After SCFAs intervention, the expression of inflammatory factors in the intestine, brain tissue and peripheral blood of the WT-SCFAs group was significantly higher than that in the WT group, and the expression in the AD-SCFAs group was also significantly higher than the AD group (Figure 1G–1I). IHC showed that NLRP3 and cleaved-Caspase-1 was negatively expressed in the intestine of WT mice and weakly positively expressed in WT + SCFAs, indicating that SCFAs can activate NLRP3 inflammasome. The intestinal expression of NLRP3 and cleaved-Caspase-1 of AD mice was weakly positive, which was positive in AD + SCFAs mice, indicating that SCFAs could promote the high expression of NLRP3 in AD mice (Figure 2A). WB was used to detect the key proteins of NLRP3 inflammasome. As a result, SCFAs could promote the intestinal expression of NLRP3, cleaved-Caspase-1 and ASC. Meanwhile, the expression of microglia activation marker (IBA-1 and GFAP) in brain tissue was also up-regulated, suggesting that SCFAs could activate microglia in brain tissue and promote the release of inflammatory factors (Figure 2B–2E).

Figure 1. Effect of SCFAs on the cognitive ability, DNTs differentiation and expression of inflammatory factor in AD mice. (A–C) Results of Morris water maze in mice ($\overline{x}\pm \text{ s}$, n = 10): Compared with WT group at the same time point, ^*P < 0.05; comparison with AD group at the same time point, ^#P < 0.05. (D–F) Results of DNTs test ($\overline{x}\pm \text{ s}$, n = 10) in vivo: Comparison of proportion of DNTs in the spleen in (E), and comparison of proportion of DNTs in the intestine in (F). Statistical significance between the groups, P < 0.05. (G–I) The expression levels of inflammatory factors (IL-1β, IL-6 and TNF-α) in peripheral blood, intestine and brain of mice ($\overline{x}\pm \text{ s}$, n = 10): Comparison with WT group, ^*P < 0.05; comparison with AD group, ^#P < 0.05.

Figure 2. Effect of SCFAs on the activation of intestinal NLRP3 inflammasome. (A) Detection of mouse intestinal NLRP3 and Caspase-1 expression by IHC (n = 5): The expression of NLRP3 and Caspase-1 was negative in the WT group and weakly positive in the WT-SCFAs group. SCFAs promoted the activation of NLRP3. The expression of NLRP3 and Caspase-1 was weakly positive in the AD group, and was positive in the AD-SCFAs group, which was significantly higher than that in the AD group. (B–C) The expression of NLRP3 inflammasome-related protein in the intestine ($\overline{x}\pm \text{ s}$, n = 10): The expression of NLRP3, cleaved-Caspase-1, ASC and Pro-Caspase-1 in the NLRP3 inflammasome was low in the WT group. Comparison with the WT group, ^*P < 0.05; comparison with the AD group, ^#P < 0.05. (D–E) The protein expression of microglia activation marker (IBA-1 and GFAP) in brain tissue ($\overline{x}\pm \text{ s}$, n = 10): Comparison with the WT group, ^*P < 0.05; comparison with the AD group, ^#P < 0.05.

Effect of Fas/Fasl and TNF-α/TNFR1 signal suppression on DNTs and NLRP3 inflammasome

The Fas/Fas1 and TNF-α/TNFR1 signal was blocked to intervene with SCFAs together. As a result, Fas-IN and TNFR1-IN could inhibit SCFAs-induced formation of DNTs. And the proportion of DNTs in the spleen and intestine was down-regulated, which was significantly lower than that in the SCFAs group (Figure 3A–3C). IHC staining showed that Fas-IN and TNFR1-IN could suppress the intestinal expression of NLRP3 and cleaved-Caspase-1. The expression of NLRP3 and cleaved-Caspase-1 in the SCFAs group was significantly higher than that in the AD group, which was weakly positive in AD. While the expression in the SCFAs-Fas-IN and SCFAs-TNFR1-IN group was significantly reduced (Figure 3D). In the detection of inflammatory factors, we also found that Fas-IN and TNFR1-IN could inhibit the levels of inflammatory factors in the brain, intestine and peripheral blood of mice, and resist the SCFAs-induced inflammatory response, which was significantly different from SCFAs (Figure 4A–4C). WB was used to detect the protein expression. The results showed that SCFAs in the intestines induced the expression of NLRP3, which was inhibited by Fas-IN and TNFR1-IN. Besides, the activation of NLRP3 was significantly down-regulated, and the expression of NLRP3, ASC, cleaved-Caspase-1 and pro-Caspase-1 was down-regulated (Figure 4D–4E). The detection of brain tissue showed that Fas-IN and TNFR1-IN could suppress the expression of IBA-1 and GFAP, indicating that Fas/Fasl and TNF-α/TNFR1 signal blocking could inhibit SCFAs-induced microglia activation (Figure 4F–4G).

Figure 3. Effects of Fas-IN and TNFR1-IN on DNTs and NLRP3 inflammasome activation. (A–C) Results of DNTs detection ($\overline{x}\pm \text{ s}$, n = 10): The proportion of DNTs in the spleen of AD mice was relatively low. Comparison of the proportion of DNTs in the spleen in (B), and comparison of the proportion of DNTs in the intestine in C. Significant difference between the groups, P < 0.05. (D) Expression of intestinal NLRP3 and cleaved-Caspase-1 by IHC (n = 5): SCFAs could activate the expression of NLRP3 and cleaved-Caspase-1.

Figure 4. Effects of Fas-IN and TNFR1-IN on the expression of inflammatory factors and NLRP3 activation. (A–C) Results of expression of inflammatory factors ($\overline{x}\pm \text{ s}$, n = 10): SCFAs could up-regulate the expression of inflammatory factors (IL-1β, IL-6 and TNF-α) in peripheral blood, intestine and brain of mice, comparison with AD group, ^*P < 0.05; comparison with SCFAs group, ^#P < 0.05. (D–E) Result of NLRP3 inflammasome activation in mouse intestine ($\overline{x}\pm \text{ s}$, n = 10): Fas-IN and TNFR1-IN could down-regulate the activation of NLRP3 inflammasome, and down-regulate the expression of NLRP3, ASC, cleaved-Caspase-1 and pro-Caspase-1, comparison with AD group, ^*P < 0.05; comparison with SCFAs group, ^#P < 0.05. (F–G) Expression level of microglia activation marker (IBA-1 and GFAP) in brain tissue ($\overline{x}\pm \text{ s}$, n = 10): Fas-IN and TNFR1-IN could induce the expression of IBA-1 and GFAP. Comparison with AD group, ^*P < 0.05; comparison with SCFAs group, ^#P < 0.05.

SCFAs promoted the formation of DNTs through OX40

The isolated lymphocytes from the spleen were cultured with SCFAs. As a result, the proportion of DNTs was significantly up-regulated, while the proportion of DNTs without induction was low. When OX40 antibody was pre-treated to block OX40 signal, the proportion of SCFAs-induced DNTs was also down-regulated, which was significantly lower than that of the SCFAs group (Figure 5A–5D). The expression of TNF-α and Fas in DNTs was significantly up-regulated, indicating that DNTs could secrete inflammatory factors (Figure 5E). SCFAs could promote the expression of OX40, Bcl-2, Bcl-xl and Survivin, and promote the survival and proliferation of DNTs cells. OX40 antibody pretreatment could significantly suppress the expression of Bcl-2, Bcl-xl and Survivin (Figure 5F–5I).

Figure 5. SCFAs promoted the formation of DNTs through OX40. (A–B) SCFAs promoted the formation of DNTs in vitro ($\overline{x}\pm \text{ s}$, n = 3): Cells in the SCFAs group were co-cultured with acetic acid (5 mM), propionic acid (1 mM) and butyric acid (1 mM) for 24 h, The proportion of DNTs in the SCFAs group was increased, and SCFAs promoted the formation of DNTs. Significant difference between groups, P < 0.05. (C–D) Effect of OX40 inhibition on the formation of DNTs ($\overline{x}\pm \text{ s}$, n = 3): after OX40 inhibition, the proportion of DNTs in the SCFAs-anti-OX40 group was significantly down-regulated than that in the SCFAs group. Significant difference between the groups, P < 0.05. (E) Detection of Fas and TNF-α secretion in DNTs ($\overline{x}\pm \text{ s}$, n = 3): the expression of Fas and TNF-α in DNTs was significantly up-regulated, which was significantly higher than that of the Con group. Comparison between groups, P < 0.05. (F–G) Effect of SCFAs on the activation of OX40 in DNTs ($\overline{x}\pm \text{ s}$, n = 3): the expression of OX40, Bcl-2, Bcl-xl and Survivin was significantly up-regulated in the SCFAs group, which was higher than that of the Con group. Comparison with the Con group, ^*P < 0.05. (H–I) Effect of OX40 inhibition on the expression of related protein ($\overline{x}\pm \text{ s}$, n = 3): anti-OX40 pretreatment could significantly inhibit the expression of Bcl-2, Bcl-xl and Survivin. Comparison with the Con group, ^*P < 0.05; comparison with the SCFAs group, ^#P < 0.05.

DNTs promoted NLRP3 inflammasome activation and inflammatory response in intestinal macrophages

The isolated DNTs from the spleen were co-cultured with intestinal macrophages. IF staining showed no expression of NLRP3 and cleaved-Caspase-1 in cells from the Con group, while co-culture of DNTs could activate the expression of NLRP3 and cleaved-Caspase-1 in macrophages (Figure 6A–6B). The detection of inflammatory factors showed that the co-culture of DNTs significantly up-regulated the expression of IL-6, IL-18 and IL-1β, which was significantly higher than that of the Con group (Figure 6C). Moreover, we also found that DNTs could promote the expression of NLRP3, cleaved-Caspase-1, pro-Caspase-1 and ASC, which was significantly higher than that of the Con group. Meanwhile, the expression of Fasl and TNFR1 in macrophages was also increased (Figure 6D–6E).

Figure 6. DNTs promoted NLRP3 inflammasome activation and inflammatory response in intestinal macrophages. (A–B) Detection of the expression of NLRP3 and Caspase-1 by IF staining (n = 3): NLRP3 and Caspase-1 was not expressed in the Con group, and the expression level was up-regulated after co-culture with DNTs. (C) Effect of DNTs on macrophage activation and the expression of inflammatory factors ($\overline{x}\pm \text{ s}$, n = 3): the level of inflammatory factors in the Con group was low. Comparison with Con group, ^*P < 0.05. (D–E) Effect of NLRP3 inflammasome and related protein expression ($\overline{x}\pm \text{ s}$, n = 3): DNTs co-culture significantly activated the expression of NLRP3 inflammasome related proteins, comparison with the Con group, ^*P < 0.05.