Captopril treatment prevents kainate-induced epilepsy
We established a KA (ip injection of 14 mg/kg KA) induced chronic epilepsy model to evaluate the effects of captopril on preventing the development of epilepsy. Captopril at a dose of 50 mg/kg/day was administered intraperitoneally from the 3rd day after the KA induction (Fig. 1A). Electroencephalogram (EEG) was recorded from week 2 to 7, and the typical records of seizures in KA and KA + Cap groups in the 7th week are shown in Fig. 1B. The EEG results indicated that 50 mg/kg captopril treatment significantly reduced the frequency of recurrent seizures from weeks 2 to 7, as shown in Fig. 1C (*p < 0.05 and **p < 0.01, with an average of 0.94 vs. 2.53, KA + Cap vs. KA group). Furthermore, captopril treatment reduced the seizure duration in both 2 and 7 weeks following the KA induction (2 weeks: **p < 0.01, 169.40 ± 73.07 vs. 519.00 ± 81.34; 7 weeks: *p < 0.01, 19.50 ± 12.69 vs. 337.00 ± 93.10, KA + Cap vs. KA group, Fig. 1D).
Captopril treatment ameliorates the epilepsy-associated cognitive impairment
To determine the effects of captopril on the epilepsy-associated cognitive deficits, we assessed the hippocampal-dependent short-term learning and memory performance by Y-maze and NOR test (Fig. 2A, D). The KA group exhibited memory deficits in Y-maze and NOR test, compared to the control group, which were evident from the significant reduction of the ratio of discrimination index in the NOR test (##p < 0.01, 0.43 ± 0.04 vs. 0.60 ± 0.02, KA vs. control group, Fig. 2B, C) and the alternation score in the Y-maze test (#p < 0.05, 63.22 ± 4.46% vs. 79.65 ± 4.16%, KA vs. control group, Fig. 2D, E). Moreover, captopril treatment markedly improved the ratio of discrimination index to the control level in the NOR test (**p < 0.01, 0.61 ± 0.02 vs. 0.43 ± 0.04, KA + Cap vs. KA group, Fig. 2B, C) and the alternation score in the Y-maze test, compared to the KA group (*p < 0.05, 83.35 ± 4.20% vs. 63.22 ± 4.46%, KA + Cap vs. KA group, Fig. 2D, E). Thus, captopril attenuated the KA-induced impairment in short-term memory.
Spatial learning and memory were tested using MWM 7 weeks after the KA induction. An average of three trials held each day for each group was recorded. The time escape latency in control, KA and KA + Cap groups were decreased over successive days, which showed no obvious difference on the training day 1–4 (control: 28.71 ± 3.57 s; KA: 33.22 ± 5.48 s; KA + Cap: 27.15 ± 5.20 s, Fig. 2F). On the probe test day, KA group revealed a significant decrease in the time spent in the target quadrant, where the platform was previously located, compared to the control group. Typical swimming route paths for each group are shown in Fig. 2G. The KA + Cap group spent more time in the target quadrant compared to the KA group (#p < 0.05, 0.23 ± 0.01 s vs. 0.33 ± 0.04 s, KA vs. control group; *p < 0.05, 0.33 ± 0.03 s vs. 0.23 ± 0.01 s, KA + Cap vs. KA group, Fig. 2H). These data indicate that the KA-induced cognitive decline is ameliorated by captopril treatment.
Captopril treatment suppresses phagocytosis and inflammatory responses in the hippocampus 9 weeks after the KA induction
To gain a deeper understanding of the molecular mechanisms underlying the effects of captopril treatment, we performed RNA sequencing (RNA-seq) analysis using the hippocampal tissues from the control, KA and KA + Cap groups 9 weeks after the KA induction. We identified 692 differentially expressed genes (DEGs, adjusted p < 0.05, 517 up and 175 down) in the KA group, compared to the control group, and 371 DEGs (130 up and 241 down) in the KA + Cap group, compared to the KA group (Fig. 3A, B). Further cross-genotype comparisons showed that captopril treatment reversed about one-third of the DEGs compared to the KA and control group (Fig. 3C). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways analysis of the DEGs revealed a significant down-regulation of various immune pathways in the KA + Cap group, including inflammatory response, immune response, phagocytosis, complement and coagulation cascades, etc. (Fig. 3D). To explore potential mechanisms related to the effects of captopril treatment, we analyzed the known and predicted protein interactions of these different genes. The results showed that the central regulatory proteins of these genes include Cd68, Cd44, Ccl2, matrix metallopeptidase 9 (Mmp9) and tissue inhibitor of metalloproteinases 1 (Timp1), all of which are associated with the glia activation-related immune and inflammatory responses (Fig. 3E).
To investigate the enrichment of DEGs between the KA and KA + Cap group, we performed GSEA assay, a robust computational method that determines whether an a priori-defined set of genes shows statistical significance, concordant difference between the 2 groups. In the KA group, the genes related to the phagocytosis signaling pathway (rno 04145), the complement system (rno 04610) and cytokine–cytokine receptor interaction (rno 04060) were up-regulated compared to the control group (NES = 1.90, FDR = 0.01; NES = 2.01, FDR = 0.01; NES = 1.80, FDR = 0.13), which were significantly down-regulated by captopril treatment (NES = − 1.53, FDR = 0.25; NES = − 1.92, FDR = 0.16; NES = − 1.52, FDR = 0.21). Noticeably, the changes in phagocytosis were more significant than that of inflammatory cytokines (Fig. 4A).
Most significantly down-regulated genes by captopril treatment were immune response-related genes. These DEGs can be divided explicitly into phagocytosis related genes, such as C1r, C3, Itga5, Itgβ2, etc. (Fig. 4B); complement system signaling pathway genes, such as C1q, C3, C3ar1, Cd68, etc. (Fig. 4C) and chemokine genes, such as Ccl2, Ccl3, Ccl7, Cxcl13, etc.; cytokine genes, such as Il-18, Il-1β and cytokine receptor genes, such as Il-1r, etc. (Fig. 4D).
To confirm the reliability of the expression profiles generated by the RNA-seq and DEGs analysis, qRT-PCR was performed to validate the expression of the typical significantly up-regulated genes. Microglia and astrocytes activation is associated with increased proinflammatory cytokines, including Il-1β, Il-18 and Tnf-α, and chemokines, such as Ccl2, Ccl3 and Cxcl13, all of which were elevated in the hippocampus of the KA group. Captopril treatment significantly lowered the KA-induced transcriptional levels of cytokine, consistent with the results from the RNA-seq experiments (Il-1β: ***p < 0.001, 1.74 ± 0.15 vs. 3.09 ± 0.42; Il-18: *p < 0.05, 2.23 ± 0.54 vs. 3.95 ± 0.38; Tnf-α: *p < 0.05, 1.24 ± 0.12 vs. 1.66 ± 0.05; Ccl2: **p < 0.01, 1.61 ± 0.34 vs. 13.80 ± 3.77; Ccl3: ***p < 0.001, 1.93 ± 0.45 vs. 5.19 ± 0.50, KA + Cap vs. KA group, Fig. 4E).
Captopril treatment reduces the activation of astrocytes and microglia in the hippocampus 9 weeks after the KA induction
Given that microglia and astrocytes are the main cell types exerting inflammatory and immune responses in the brain, we performed qRT-PCR and immunofluorescence staining for microglia and astrocyte marker genes to investigate the effects of captopril on glial activation in the hippocampus 9 weeks after the KA induction when recurrent seizures were supposedly formed. We found that the mRNA expression of Iba1 and Gfap was significantly increased in the hippocampus in the KA group, which were almost totally reversed by captopril treatment (Iba1: ##p < 0.01, 1.65 ± 0.09 vs. 1.00 ± 0.13, KA vs. control group; **p < 0.01, 0.99 ± 0.14 vs. 1.65 ± 0.09, KA + Cap vs. KA group; Gfap: ###p < 0.001, 4.04 ± 0.46 vs. 1.00 ± 0.10, KA vs. control group; ***p < 0.001, 1.51 ± 0.36 vs. 4.04 ± 0.46, KA + Cap vs. KA group, Fig. 5A). The quantitative analysis of immunostaining in the hippocampal CA1 area revealed that captopril treatment markedly decreased GFAP+ and Iba1+ cell numbers and area compared to the KA group, which was consistent with the results from the qRT-PCR analysis. (Gfap: *p < 0.05, 49.33 ± 3.38 vs. 72.33 ± 5.23; **p < 0.01, 4.11 ± 0.30 vs. 17.74 ± 1.19, KA + Cap vs. KA group; Iba1: **p < 0.01, 51.33 ± 1.45 vs. 72.00 ± 3.51; **p < 0.01, 2.75 ± 0.34 vs. 5.40 ± 0.41, KA + Cap vs. KA group, respectively, Fig. 5B–D).
In addition, we found that microglia in the KA group underwent a morphological change from a “resting” ramified phenotype to an “activated” bushy phenotype. Sholl analysis of astrocyte microglia morphologies using the ImageJ software showed reduced branch number and branch length in the KA group. Noticeably, these phenotypes were reversed by captopril treatment (***p < 0.001, 33.83 ± 1.49 vs. 17.08 ± 0.77; 70.80 ± 4.69 vs. 27.78 ± 2.78 KA + Cap vs. KA group, Fig. 5E, F).
Double immunostaining of astrocytes and microglia showed that they had a tendency to approach each other in the KA group, indicated by the fluorescence intensity of Gfap at different distances of 10, 30 and 50 μm from the center of microglia, which was significantly restored by the captopril treatment (10 μm: ***p < 0.001, 608.10 ± 21.17 vs. 1323.00 ± 111.70; 30 μm: ***p < 0.001, 5570.00 ± 254.30 vs. 11,513.00 ± 637.5; 50 μm: ***p < 0.001, 15,584.00 ± 619.8 vs. 33,367.00 ± 2269.00, KA + Cap vs. KA group, Fig. 6A, B). Taken together, these data suggest a pivotal role for captopril in ameliorating the KA-induced glial cell activation and contact in the hippocampus.
Captopril treatment attenuates the KA-induced glial activation through C3–C3ar negative expression in the hippocampus 9 weeks after the KA induction
To investigate the role of C3–C3ar signaling in immune regulation and glial activation, we examined the mRNA expression of C3 and C3ar in the hippocampus 9 weeks after the KA induction. The results of RT-qPCR showed that mRNA expression of C3 and C3ar was increased in the KA group, which was almost completely abolished by captopril treatment (C3: ###p < 0.001, 11.02 ± 2.53 vs. 1.00 ± 0.19 KA vs. control group; **p < 0.01, 2.81 ± 0.59 vs.11.02 ± 2.53, KA + Cap vs. KA group; C3ar: ##p < 0.01, 1.71 ± 0.11 vs. 1.00 ± 0.67, KA vs. control group; *p < 0.05, 1.16 ± 0.14 vs. 1.71 ± 0.11, KA + Cap vs. KA group, Fig. 7A).
Consistent with the elevated C3 and C3ar mRNA in the hippocampus, we detected a drastic increase of C3 intensity in the CA1, which was concentrated in the Gfap+ astrocytes in the KA group (Fig. 7B). In addition, the C3ar intensity in the CA1 area was significantly increased, concentrated in Iba1+ microglia (Fig. 7D). Noticeably, both C3 and C3ar co-localized signals with astrocytes or microglia were significantly attenuated after captopril treatment (***p < 0.001, C3 intensity: 43.16 ± 6.56 vs.113.80 ± 9.25; C3 in Gfap: 8.38 ± 1.28 vs. 53.98 ± 5.42; C3ar intensity: ***p < 0.01, 54.43 ± 4.60 vs. 96.64 ± 8.27; C3ar in Iba1: 1.91 ± 0.24 vs. 8.82 ± 0.71, KA + Cap vs. KA group, Fig. 7E, F). Co-immunostaining of C3 and Iba1 revealed no positive C3 protein signal in Iba1-positive microglia (Fig. 7 C). These results suggest that captopril attenuates epilepsy-related astrocyte and microglia activation and the production of proinflammatory cytokines through the C3–C3ar signaling pathway.
Captopril treatment reduces microglia-mediated synaptic phagocytosis in the hippocampus 9 weeks after the KA induction
The complement pathway has been implicated in microglia-mediated synapse pruning. The activation of C3 receptors on microglia triggers their activation and synaptic elimination by phagocytosis [42]. To further characterize the effects of captopril treatment on microglia phenotype, Cd68 and Iba1 double staining were conducted to identify phagocytic microglia (Fig. 8A). Quantification of Cd68 immunoreactivity revealed that captopril treatment dampened the proportion of Cd68-positive microglia in the hippocampal CA1 area (***p < 0.001, 21.48 ± 6.13% vs. 74.90 ± 3.57%, KA + Cap vs. KA group; ###p < 0.001, 74.90 ± 3.57% vs. 30.74 ± 2.03%, KA vs. control group, respectively, Fig. 8B, C). To investigate whether captopril treatment attenuated total synaptic loss in the model, we quantified the synapsin immunoreactivity in the hippocampal CA1 (Fig. 8D), where it showed a significant reduction of synapsin positive signals in the hippocampal CA1 area in the KA group, compared to the control group (###p < 0.001, pyramidale layer: 58.42 ± 5.51 vs. 93.88 ± 2.68; radiatum layer: 69.73 ± 6.96 vs. 124.00 ± 5.17, KA vs. control group, Fig. 8E). Captopril treatment significantly ameliorated the KA-induced synaptic loss indicated by an elevated synapsin immunoreactivity in both the pyramidale and radiatum layers, compared to the KA group (***p < 0.001, pyramidale layer: 101.30 ± 19.58 vs. 58.42 ± 5.51; radiatum layer: 132.70 ± 8.32 vs. 69.73 ± 6.96, KA + Cap vs. KA, Fig. 8E). Together, these results indicate that captopril treatment attenuates the KA-induced synaptic phagocytosis by microglia in the hippocampal CA1, which may further prevent synaptic loss in the KA-induced rat model of epilepsy.
Intranasal C3a treatment leads to cognitive impairment in the KA-induced model of epilepsy after captopril treatment
The above results demonstrated a prominent role of C3–C3ar signaling in the therapeutic effects of captopril on the KA-induced model of epilepsy. To further verify that captopril treatment acted through the suppression of C3–C3ar signaling, C3a was administered intranasally to the captopril-treated rats after the KA induction, then EEG and cognition-related behavioral performance were assessed (Fig. 9A). The KA + Cap + C3a group exhibited more severe memory deficits in both NOR and Y-maze tests compared to the KA + Cap group, evident from a significant reduction of the ratio of discrimination index in NOR test (&p < 0.05, 0.42 ± 0.05 vs. 0.61 ± 0.04, KA + Cap + C3a vs. KA + Cap group, Fig. 9B, C) as well as a reduction of the alternation score in Y-maze test (&&p < 0.01, 56.46 ± 5.42% vs. 83.35 ± 4.20%, KA + Cap + C3a vs. KA + Cap group, Fig. 9D, E). The mean time escape latency of 4 groups were all reduced over successive days, which showed no obvious difference (control: 28.71 ± 3.57 s; KA: 33.22 ± 5.48 s; KA + Cap: 27.15 ± 5.20 s; KA + Cap + C3a: 40.61 ± 5.39 s, Fig. 9F). For the MWM test, typical swimming route paths for each group are shown in Fig. 9G. Furthermore, on the probe test day of the MWM task, the KA + Cap + C3a group spent less time in the target quadrant compared to the KA + Cap group (&p < 0.05, 0.20 ± 0.03 s vs. 0.33 ± 0.03 s, KA + Cap + C3a vs. KA + Cap group, Fig. 9H). Taken together, intranasal C3a worsened the cognitive deficits after captopril treatment in epileptic rats.
Intranasal C3a treatment leads to the development of epilepsy and synaptic phagocytosis after captopril treatment
We next assessed the effects of intranasal C3a treatment on the therapeutic effects of captopril on preventing epileptogenesis. We found that daily intranasal administration of 1.3 μg/kg C3a partially reversed the therapeutic effects of captopril on the prevention of epileptogenesis, indicated by more frequent seizures and total seizure duration 7 weeks after the KA induction (Seizures frequency: &p < 0.05, 2.00 ± 0.44 vs. 0.33 ± 0.20; Total seizure duration: 244.50 ± 79.13 vs. 19.50 ± 12.69, KA + Cap + C3a vs. KA + Cap group, respectively, Fig. 10A, B).
Next, we assessed the effects of intranasal C3a treatment on the therapeutic effects of captopril on glial activation-mediated synaptic remodeling. Co-immunostaining of C3ar and Iba1 revealed a higher level of expression and a higher degree of co-localization in the KA + Cap + C3a group, compared to the KA + Cap group (C3ar intensity: &p < 0.05, 80.22 ± 6.56 vs. 54.05 ± 4.72; C3ar within Iba1: &&&p < 0.001, 57.45 ± 6.40 vs. 8.40 ± 1.22, KA + Cap + C3a vs. KA + Cap group, Fig. 11A, B). Co-immunostaining of Cd68 with Iba1 revealed that C3a treatment partially blocked the effects of captopril treatment on inhibiting microglia phagocytosis in the hippocampal CA1 (&&&p < 0.001, 72.33 ± 2.84% vs. 29.26 ± 3.62%, KA + Cap + C3a vs. KA + Cap group, Fig. 11C, D). Moreover, C3a treatment also reversed the therapeutic effects of captopril treatment on synaptic loss in the hippocampal CA1 to a similar level of the KA group (&&&p < 0.001, pyramidale layer: 55.20 ± 3.37 vs. 107.90 ± 3.72; radiatum layer: 74.82 ± 4.48 vs. 150.70 ± 3.43, KA + Cap + C3a vs. KA + Cap group, Fig. 11E, F). These findings indicate that intranasal C3a treatment starting on the 3rd day following the KA induction contributes to cognitive deficits, epileptogenesis, C3-mediated glial activation and synaptic phagocytosis after the captopril treatment.