JQ1 treatment inhibits the phosphorylation of BRD4 and prevents LPS-induced sepsis-related symptoms in mice
High dose lipopolysaccharide (LPS) could induce sepsis and sepsis-associated encephalopathy [25]. In the present study, we explore the expression of BRD4 and p-BRD4 in the hippocampus of mice, which is a main region susceptibility to sepsis. Real-Time PCR showed that there was significantly differences among the groups in the mRNA level of BRD4 (F(2,15) = 12.94, p = 0.0008, Figure. 1A). Bonferroni post-test analysis revealed that the mRNA level of BRD4 in DMSO + LPS mice increased when compared with those of the DMSO + NS group (t = 4.927, p < 0.001), which was reversed by JQ1 treatment (t = 3.657, p < 0.01). Western blot was used to detect the protein level of BRD4, and one-way ANOVA suggested that there were significantly differences among the groups in the protein level of BRD4 (F(2,11) = 17.29, p < 0.001, Figure. 1B and 1C). Bonferroni post-test analysis revealed that when compared with the DMSO + NS group, the protein level of BRD4 in DMSO + LPS mice increased significantly (t = 4.927, p < 0.01), which was reversed by JQ1 treatment (t = 5.242, p < 0.01). Furthermore, we investigated the expression of BRD4 and p-BRD4 by immunohistochemistry, and the representative images of BRD4 and p-BRD4-positive cells are shown in Figure. 1D. The analysis results show that there were significantly differences among the groups in the average optical density (OD) of BRD4 (F(2,19) = 8.053, p = 0.0035, Figure. 1E) and p-BRD4 (F(2,44) = 8.952, p = 0.0006, Figure. 1E). Bonferroni post-test analysis revealed that LPS challenge could increase the expression of hippocampal BRD4 (t = 3.764, p < 0.001) and p-BRD4 (t = 2.794, p < 0.05), but JQ1 administration could decrease the expression of BRD4 (t = 2.872, p < 0.05) and p-BRD4 (t = 4.096, p < 0.001). These results suggested that JQ1 treatment not only inhibited the expression of BRD4, but also repressed the phosphorylation of BRD4.
Sepsis exhibited weight loss and splenomegaly. In the present study, the body weight and spleen weight of DMSO + NS, DMSO + LPS and JQ1 + LPS group mice were also determined. Two-way ANOVA with Bonferroni post-test analysis revealed that 24 h after LPS injection, the body weight of mice decreased when compared with the NS group (t = 3.403, p < 0.01), but no significant difference was found between the DMSO + NS groups and JQ1 + LPS mice (t = 1.232, p > 0.05, Figure. 1F). One-way ANOVA indicated there was a significant difference among the groups in the spleen weight/body weight (F(2,20) = 25.99, p < 0.001, Figure. 1G). Bonferroni post-test analysis revealed that the spleen weight/body weight in DMSO + LPS mice increased when compared with those of the DMSO + NS group (t = 5.806, p < 0.001), which was rescued by JQ1 treatment (t = 6.592, p < 0.001). These results suggested that JQ1 could inhibit weight loss and splenomegaly induced by LPS administration.
Procalcitonin (PCT) and D-lactate dehydrogenase (D-LDH) levels are the diagnosis and monitoring markers in sepsis. So we also detected the levels of serum PCT and D-LDH by using ELISA kits. One-way ANOVA suggested that there were significantly differences among the groups in the levels of PCT (F(2,16) = 4.046, p = 0.0076, Figure. 1H) and D-LDH (F(2,18) = 10.55, p = 0.0012, Figure. 1I). Bonferroni post-test analysis revealed that LPS caused increased PCT level (t = 3.694, p < 0.01) and D-LDH level (t = 4.349, p < 0.01) compared to NS mice, these effects were effectively attenuated by JQ1 treatment ((PCT (t = 2.779, p < 0.05) and D-LDH (t = 3.116, p < 0.05) ). These results indicated that JQ1 could prevent the process of sepsis induced by LPS administration.
JQ1 treatment inhibits LPS-induced the activation of NF-κB and inflammasomes
Inflammasomes, mainly including NLRP1, NLRP3 and Aim2, play an important role in triggering pyroptosis and inflammatory response in SAE [28]. To determine the role of JQ1 treatment on the activation of the inflammasomes, we examined the alterations of mRNA levels of NLRP1, NLRP3, Aim2 and ASC in the hippocampus of mice. One-way ANOVA suggested that there were significantly differences among the groups in the mRNA level of NLRP1 (F(2,16) = 26.28, p < 0.0001), NLRP3 (F(2,17) = 11.86, p = 0.0008), Aim2 (F(2,16) = 6.822, p = 0.0085) and ASC (F(2,17) = 14.93, p = 0.0003) ( Figure. 2A ). Bonferroni post-test analysis revealed that LPS increased the mRNA level of NLRP1 (t = 5.312, p < 0.001), NLRP3 (t = 4.855, p < 0.01), Aim2 (t = 3.113, p < 0.05) and ASC (t = 4.346, p < 0.01) compared to NS mice, while JQ1 treatment decreased the levels of NLRP1 (t = 2.791, p < 0.001), NLRP3 (t = 2.758, p < 0.05), Aim2 (t = 3.232, p < 0.05) and ASC (t = 5.047, p < 0.001) compared to those LPS mice.
Nuclear factor-κB (NF-κB) signaling pathway was found to be involved in LPS-induced inflammatory response [29]. Furthermore, western blotting was used to determine the protein levels of NF-κB, NLRP1, NLRP3, Aim2 and ASC in the hippocampus of mice. One-way ANOVA suggested that there were significantly differences among the groups in the protein level of NF-κB (F(2,11) = 12.89, p = 0.0023), NLRP1 (F(2,11) = 18.14, p = 0.0007), NLRP3 (F(2,11) = 10.21, p = 0.0048), Aim2 (F(2,11) = 22.04, p = 0.0003) and ASC (F(2,11) = 11.95, p = 0.0029) ( Figure. 2B and 2C ). Bonferroni post-test analysis revealed that LPS increased the protein level of NF-κB (t = 3.540, p < 0.05), NLRP1 (t = 4.468, p < 0.01), NLRP3 (t = 4.188, p < 0.01), Aim2 (t = 5.807, p < 0.001) and ASC (t = 4.582, p < 0.01) compared to NS mice, while JQ1 treatment decreased the levels of NF-κB (t = 4.922, p < 0.01), NLRP1 (t = 5.732, p < 0.001), NLRP3 (t = 3.566, p < 0.05), Aim2 (t = 5.690, p < 0.001) and ASC (t = 3.766, p < 0.05) compared to those LPS mice. These results suggested that JQ1 treatment could inhibit the activation of NF-κB signaling pathway and NLRP1, NLRP3 and Aim2 inflammasomes induced by LPS injection.
JQ1 treatment ameliorates pyroptosis in the hippocampus of SAE mice exposed to LPS
Accordingly, a feature of pyroptosis is the pore formation in cell membranes, which could be triggered by gasdermins (GSDMs) family protein [30]. Researches showed that GSDMD [10] and GSDMA [11] were cleaved by Caspase-1/11 upon inflammasome activation and mediated canonical pyroptosis. As displayed in the Real-Time PCR analysis, one-way ANOVA suggested that there were significantly differences among the groups in the mRNA level of Caspase-1 (F(2,17) = 20.84, p < 0.0001), Caspase-11 (F(2,16) = 7.816, p = 0.0053), GSDMD (F(2,17) = 17.86, p = 0.0001) and GSDMA (F(2,17) = 20.27, p < 0.0001) (Figure. 3A). Bonferroni post-test analysis revealed that LPS stimulated up-regulation of the mRNA level of Caspase-1 (t = 3.994, p < 0.01), Caspase-11 (t = 3.590, p < 0.01), GSDMD (t = 3.445, p < 0.05) and GSDMA (t = 3.410, p < 0.05) compared to NS mice, and these effects were effectively attenuated by JQ1 treatment ((Caspase-1 (t = 6.390, p < 0.001), Caspase-11 (t = 3.167, p < 0.05), GSDMD (t = 5.952, p < 0.001) and GSDMA (t = 6.361, p < 0.001)). Western blotting further confirmed this fluctuating expression pattern of cleaved Caspase-1, cleaved Caspase-11, GSDMD-N and GSDMA-N. One-way ANOVA revealed that there were significantly differences among the groups in the protein levels of cleaved Caspase-1 (F(2,11) = 7.331, p = 0.0129), cleaved Caspase-11 (F(2,11) = 7.146, p = 0.0139), GSDMD-N (F(2,11) = 20.19, p = 0.0005) and GSDMA-N (F(2,11) = 14.74, p = 0.0014) ( Figure.3B and 3C). Bonferroni post-test analysis revealed that LPS stimulated up-regulation of the protein level of cleaved Caspase-1 (t = 2.985, p < 0.05), cleaved Caspase-11 (t = 3.036, p < 0.05), GSDMD-N (t = 6.348, p < 0.001) and GSDMA-N (t = 3.618, p < 0.05) compared to NS mice, these effects were effectively attenuated by JQ1 treatment ((cleaved Caspase-1 (t = 3.569, p < 0.05), cleaved Caspase-11 (t = 3.469, p < 0.05), GSDMD-N (t = 3.413, p < 0.05) and GSDMA-N (t = 5.316, p < 0.01)). The results indicated JQ1 treatment inhibited LPS-induced the mRNA and protein levels of canonical pyroptosis pathway.
Caspase-8 could cleave GSDMC, and then mediated non-canonical pyroptosis pathway [12]. In order to determine the role of JQ1 treatment on the non-canonical pyroptosis pathway, we examined the alterations of mRNA and protein levels of Caspase-8 and GSDMC. The results of Real-Time PCR showed the mRNA level of Caspase-8 (F(2,17) = 0.2848, p = 0.7561) and GSDMC (F(2,16) = 0.2408, p = 0.7892) had no significant difference among DMSO + NS, DMSO + LPS and JQ1 + LPS group (Figure. 3D). And that western blotting further certificated there were no significant difference among these groups in the protein levels of cleaved Caspase-8 (F(2,11) = 0.04195, p = 0.9591) and GSDMC-N (F(2,11) = 1.797, p = 0.2205) (Figure. 3E and 3F).
GSDME has been demonstrated to be cleaved by Caspase-3, and then converting apoptosis to pyroptosis [10]. In our present study, we found the mRNA level of Caspase-3 (F(2,17) = 0.3112, p = 0.7372) and GSDME (F(2,17) = 0.01089, p = 0.9892) (Figure. 3G), and the protein level of cleaved Caspase-3 (F(2,11) = 2.405, p = 0.1456) and GSDME-N (F(2,11) = 0.001761, p = 0.9982) (Figure. 3H and 3I) had no significantly difference among DMSO + NS, DMSO + LPS and JQ1 + LPS group. Because GSDMB wasn’t expressed in mice [31], so we didn’t detect the expression of GSDMB. All these results indicated JQ1 treatment inhibited LPS-induced the canonical pyroptosis, but not non-canonical pyroptosis or the apoptosis converting to the pyroptosis.
JQ1 suppresses microglia activity and the release of inflammatory cytokines in the hippocampus of SAE mice exposed to LPS
Microglia are innate immune cells in the brain, it could be activated after LPS injection [32]. Here, we investigated the role of JQ1 treatment on the activation of microglia. One way ANOVA showed there was significantly difference among the groups in the mRNA level of CD68 (F(2,17) = 13.32, p = 0.0005, Figure. 4A ), which was a marker of activated microglia. Bonferroni post-test analysis revealed that LPS stimulated up-regulation of the mRNA level of CD68 (t = 3.469, p < 0.05), this effect was effectively attenuated by JQ1 treatment (t = 5.045, p < 0.001). Western blotting further demonstrated that the protein level of CD68 (F(2,11) = 7.869, p = 0.0106, Figure. 4B and 4C ) was significantly different in DMSO + NS, DMSO + LPS and JQ1 + LPS group. The following post-test analysis revealed that when compared with DMSO + NS mice, LPS stimulation up-regulated the protein level of CD68 (t = 3.625, p < 0.05), but JQ1 treatment could reverse the expression of CD68 (t = 3.209, p < 0.05). Furthermore, we investigated the expression of IBA1, a marker of activated microglia, by using immunohistochemistry, and the representative images of IBA1-positive cells were shown in Figure. 4D. The results showed that there were significantly differences among the groups in the average optical density (OD) of IBA1 (F(2,46) = 33.89, p < 0.0001, Figure. 4E). Bonferroni post-test analysis revealed that LPS stimulation could increase the expression of IBA1 (t = 8.011, p < 0.001), but JQ1 administration could decrease the expression of IBA1 (t = 5.582, p < 0.001). These results indicated that JQ1 could suppress microglia activity induced by LPS injection.
The main inflammatory cytokines that produced by pyroptosis were the IL-1β and IL-18 [33, 34], which could induce the release of downstream inflammatory factors, such as IL-6. And the inflammatory cytokines over-expression has been associated with impairments in hippocampal-dependent memory [35]. Here, we examined alterations in the mRNA expression levels of IL-1β, IL-18 and IL-6 in the hippocampus of SAE. One-way ANOVA suggested that there were significantly differences among the groups in the mRNA levels of IL-1β (F(2,16) = 7.729, p = 0.0055), IL-18 (F(2,17) = 16.70, p = 0.0002) and IL-6 (F(2,17) = 26.13, p < 0.0001) (Figure. 4F). Bonferroni post-test analysis revealed that expressions of IL-1β, IL-18 and IL-6 were significantly increased in DMSO + LPS mice compared with the DMSO + NS mice (IL-1β:t = 3.457, p < 0.05; IL-18:t = 3.051, p < 0.05; IL-6:t = 6.783, p < 0.001), while JQ1 treatment remarkably decreased the expression of IL-1β, IL-18 and IL-6 (IL-1β:t = 3.455, p < 0.05; IL-18: t = 5.776, p < 0.001; IL-6:t = 5.556, p < 0.001 ). We also examined alterations in the protein levels of IL-1β, IL-18 and IL-6 in the hippocampus of mice. One-way ANOVA revealed that there were significantly differences among the groups in the protein level of IL-1β (F(2,11) = 37.66, p < 0.0001), IL-18 (F(2,11) = 16.86, p = 0.0009) and IL-6 (F(2,11) = 7.486, p = 0.0122)(Figure. 4G and 4H). Bonferroni post-test analysis revealed that LPS stimulated up-regulation of the protein level of IL-1β (t = 3.315, p < 0.05), IL-18 (t = 3.614, p < 0.05) and IL-6 (t = 2.978, p < 0.05) compared to NS mice, while JQ1 treatment remarkably decreased expression of IL-1β, IL-18 and IL-6 (IL-1β: t = 8.603, p < 0.01; IL-18: t = 5.743, p < 0.001; IL-6: t = 3.629, p < 0.05). These results suggested JQ1 treatment inhibited the upregulated expression of IL-1β, IL-18 and IL-6 induced by LPS.
JQ1 prevents tight junction and reverses neuronal morphological damage in the hippocampus of SAE mice exposed to LPS
Hippocampus is one of the important region of blood-brain barrier, high dose of LPS could damage the blood brain barrier (BBB) [36]. In our current study, tight junction proteins occludin and ZO-1 were detected to assess the effect of JQ1 on the blood-brain barrier of hippocampus. And the results showed that there were significantly differences among the groups in the protein level of occludin (F(2,11) = 24.41, p = 0.0002) and ZO-1 (F(2,11) = 35.03, p < 0.0001) (Figure. 5A and 5B). Bonferroni post-test analysis revealed that LPS stimulated down-regulation of the protein level of occludin (t = 5.776, p < 0.001) and ZO-1(t = 4.964, p < 0.01) compared to NS mice, while JQ1 treatment remarkably increased the expression of occludin (t = 6.293, p < 0.001) and ZO-1 (t = 8.319, p < 0.001).
To determine whether neurons were the main cell type that participated in LPS-induced pyroptosis. Western blotting was used to assess the relative expression of the NeuN protein, which is a marker of neuron, and one-way ANOVA revealed that there were significantly differences among the groups in the protein level of NeuN (F(2,11) = 11.89, p = 0.003) ( Figure. 5A and 5B). Bonferroni post-test analysis revealed that LPS stimulated down-regulation of the protein level of NeuN (t = 4.674, p < 0.01) compared to NS mice, while JQ1 treatment remarkably increased the expression of NeuN (t = 3.540, p < 0.05). Furthermore, we used immunohistochemistry to estimate the expression of NeuN, and the newborn neurons maker, Doublecortin (DCX). The representative images of NeuN and DCX-positive cells are shown in Figure. 5C. The results showed that there were significantly differences among the average optical density (OD) of NeuN (F(2,22) = 47.81, p < 0.0001) and DCX (F(2,50) = 34.48, p < 0.0001) (Figure. 5D). Bonferroni post-test analysis revealed that LPS stimulated could increase the expression of NeuN (t = 6.443, p < 0.001) and DCX (t = 6.540, p < 0.001), but JQ1 administration could decrease the expression of NeuN (t = 9.659, p < 0.001) and DCX (t = 7.606, p < 0.001). These results suggested that JQ1 administration could enhance tight junction of hippocampus and inhibit neuronal death, as well as protect neurogenesis.