2.1. Microarray-based profiling

The SAE-related murine brain mRNA dataset GSE167610, which contains eight control and eight disease samples, was downloaded from the GEO database. The sequencing platform annotation file GPL6885 was used for gene ID conversion. R “limma” package was used for differential analysis to identify the differentially expressed mRNAs with |log₂FC| >1 and p-value <0.05 as the threshold.

2.2. Multiplex cytokine bead array assay

Cytokine and chemokine levels were measured in mouse serum (20× diluted) and brain samples (10× diluted) using the MILLIPLEX MAP Mouse Cytokine/Chemokine Magnetic Bead Panel (MCYTMAG-70 K-PX32; Millipore), following the manufacturer’s instructions.

2.3. Animals

Male 8–12-week-old wild-type C57BL/6 J mice were purchased from the Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). The mice were housed in a specific pathogen-free (SPF) facility at the Shandong Provincial Hospital Affiliated to Shandong First Medical University for at least one week before inclusion in the experiments.

2.4. Development of mouse cecal ligation and puncture (CLP) model and treatment protocol

A mouse model of sepsis was induced using the CLP method, as previously described [12]. Briefly, mice were anesthetized by intraoperitoneal (i.p.) administration of 1.5% isoflurane, and the head and limbs were fixed. The cecum was exposed by a small midline in the lower abdomen, ligated using 3.0 silk and punctured using a 21-gauge needle. A small amount of feces was then extruded using forceps. The cecum was relocated into the abdominal cavity and the abdomen was closed using sutures. Sham-operated mice were subjected to the same operation procedure but without CLP.

Animal experiments were approved by the Animal Ethics Committee of Shandong Provincial Hospital Affiliated to Shandong First Medical University and performed in strict accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health.

Mice were randomly assigned to three groups: sham (sham-treated with corn oil), CLP (CLP treated with corn oil), and cortistatin groups (CLP treated with cortistatin-14, GlpBio, Montclair, CA). The reagents were administered intraperitoneally 30 min after the operation. The mice were euthanized at 7d after the operation.

2.5. Open-Field Test (OFT)

The OFT was used to measure the exploratory activity and anxiety-like behavior of the mice. Each mouse was placed at the center of an open-field device (50 cm ×50 cm ×50 cm). The central zone was defined as a square that was 10 cm from the wall. The time spent exploring the central zone was recorded for 5 min and quantified using the SMART system (Panlab Harvard Apparatus; Barcelona, Spain). The central duration was calculated using the following formula: central duration (%) = (time spent exploring the central zone/300 s) ×100%.

2.6. Tail-suspension test (TST)

TST was conducted according to previous studies with some modifications [13, 14]. The tail of each mouse was suspended at the height of 50 cm using adhesive tape. The duration of immobility over the 5 min test period was recorded and quantified using the SMART system (Panlab Harvard Apparatus) [15].

2.7. Elevated plus maze test (EPMT)

Each mouse was placed in the center zone (10 cm × 10 cm) of the elevated plus maze, facing one of the open arms (40 cm above the floor). The number of entries and time spent in both open arms (30 cm × 5 cm) and enclosed arms (30 cm × 5 cm, with 15 cm-high walls) were recorded using the SMART system (Panlab Harvard Apparatus) for 5 min [15]. The percentages of open arm entries and time spent in the open arms were calculated as the number of open arm entries divided by the total number of arm entries and the time spent in the open arms divided by the total time, respectively [13].

2.8. Novel object recognition test (NORT)

NORT was used to investigate the working and memory of rodents. The procedure was performed according to a previously reported protocol with some modifications [16]. The experimental procedure consisted of three sessions: habituation, training, and testing. Each mouse was placed in a Plexiglas cage (50 cm ×50 cm ×50 cm) with an exchangeable floor. Habituation was carried out for 15 min on the first three consecutive days when the mice were positioned into empty chambers without any object. Twenty-four hours after habituation, mice entered the training session. In this session, the mice were placed in the same chamber facing the wall opposite to two identical objects (A1 and A2) and allowed to explore the objects for 5 min. The time spent exploring each object was recorded. After a delay of 3 h (short-term memory), the mouse was returned to the chamber with one of the familiar objects (A2) replaced by a novel object (A3). In the test session, mice were allowed to freely explore the familiar object (A1) and the new object (A3) for 5 min. The exploration time for each object during the test session was recorded using a SMART system (Panlab Harvard Apparatus). The discrimination ratio was calculated as follows: time of exploration of the novel object/(time of exploration of the familiar object + time of exploration of the novel object) ×100. An index greater than 0.5 indicates a preference for the novel object. The apparatus and the objects were thoroughly cleaned with 75% ethanol between each trial to control olfactory cues.

2.9. Transmission electron microscopy

The 1 mm³ fragments of prefrontal cortex tissue were fixed by immersing in 3% glutaraldehyde in 0.2 M phosphate buffer (0.874 g NaH₂PO₄, 5.158 g Na₂HPO₄ in 100 mL distilled water) for more than 24 h, post-fixed with osmic acid for 2 h and processed for transmission electron microscopy. The ultrathin sections were stained with uranyl acetate and lead citrate. Finally, the ultrastructure of the endothelial cells, neuronal mitochondria and synapses was examined under a transmission electron microscope (JEOL-1011; JEOL, Tokyo, Japan).

2.10. Evans blue staining

The integrity of the mouse BBB was evaluated using Evans blue staining. The leakage of Evans blue dye in the brain tissues of five mice in each group was analyzed one day after surgery. Three hours before euthanasia, the mice received an intravenous injection of 2% Evans blue dye (E2129, EBD, Sigma-Aldrich, St. Louis, MO; 6 mL/kg saline). The concentration of Evans Blue in the brain was quantified at 610 nm using spectrophotometry, and a quantification plot was plotted.

2.11. Immunofluorescence

Mouse brain sections were preincubated with 0.3% Triton X-100 (HFH10, Thermo Fisher Scientific) in PBS for 10 min and blocked with 1% BSA (37525, Thermo Fisher Scientific) and 0.1% Triton X-100 for 1 h. The sections were then immunostained overnight at 4°C with primary antibodies against mouse Iba1 (ab178846, rabbit, monoclonal, 1 : 500, Abcam, Cambridge, UK). The sections were then incubated with AF488-conjugated secondary antibody for 1 h at 25°C. Images were acquired using a confocal microscope (Leica Microsystems GmbH, Wetzlar, Germany) in five different fields selected from each section. Fluorescent cells were counted using the ImageJ software.

2.12. Statistical analysis

The measurement data were described as mean ± standard deviation, and SPSS 21.0 software and GraphPad Prism 5.0 were used to analyze the data. Statistical significance was measured using the paired t-test (two-group paired data), unpaired t-test (two-group unpaired data), one-way ANOVA (multi-group data), or repeated measures ANOVA (multi-group data). The Kaplan-Meier method was used to assess survival, and the results were compared between groups with a log-rank test. Differences were considered statistically significant at P < 0.05.

3.1. Cortistatin is decreased in SAE mice

Analysis of the GSE167610 dataset suggested that 1313 mRNA were expressed in various ways (Figure 1(a)). Among these, decreased expression of cortistatin mRNA in the brains of SAE mice compared to sham mice in the acute phase of sepsis was identified (three days after peritoneal contamination and infection [PCI]; fold change = 0.482, t = 11.92, P = 2.33 × 10^-7, adjusted P = 1.67 × 10^-4) (Figure 1(b)).

3.2. Cortistatin-14 improved the survival and behavioral performance of CLP mice

To investigate the potential therapeutic effect of cortistatin-14 in SAE, we treated a CLP septic murine model by intraperitoneal administration of cortistatin at various doses (100–500 μg/kg). The results indicated that cortistatin exerted a protective effect on the total survival of CLP mice (Figure 2(a)), and the effect was dose-dependent (Figure S1).

We then examined the therapeutic effect of cortistatin-14 (500 μg/kg) on behavioral performance in post-septic mice. The tests were performed one day and seven days after CLP operation.

NORT was used to investigate the hippocampal mediated learning and memory in all the surviving mice. During the training session, all groups showed similar exploration times for two identical objects (data not shown). However, during the test session, CLP mice showed decreased exploratory preference and spent less time exploring the novel object both at days 1 and 7 after CLP surgery when compared to the sham mice. In contrast, cortistatin-treated CLP mice displayed improved recognition memory compared with the CLP group (Figures 2(b)–2(d). Consequently, we performed EPMT, OFT and TST to investigate the emotional behaviors of each group one day and seven days after the CLP surgery. In the EPMT, the number of entries and the amount of time spent in the open arm were comparable between the groups (Figures 2(e) and 2(f). Additionally, the TST, an anxiety-related behavioral test, revealed that the mean immobility time was significantly shorter in the cortistatin-treated mice than in septic mice (Figure 2(g)). Similarly, the OFT was used to examine the locomotor activitiy and anxiety-like behaviors. CLP mice spent lesser time in the central zone of the field than that of the sham mice, but spent more time than the cortistatin-treated mice (Figure 2(h)).

In general, these results indicate that the onset of sepsis significantly impaired the the cognitive function and emotion of the mice, while cortistatin-14 treatment prevented these adverse events.

3.3. Cortistatin treatment mitigates local and systemic inflammatory responses and inhibits microglia activation in septic mice

Next, we estimated the levels of inflammatory cytokines in the serum and brains of the animals. The serum levels of IL-1β (Figure 3(a)), IL-6 (Figure 3(b)), TNF-α (Figure 3(c)), and IFN-γ (Figure 3(d)) levels were increased in the acute phase of sepsis (one day after the surgery, CLP group) as compared with sham mice, suggesting an enhanced systemic inflammatory reaction. In contrast, cortistatins significantly decreased the levels of inflammatory cytokines. Similarly, in the late phase of sepsis (seven days after the surgery), the levels of serum IFN-γ, TNFα, and IL-1β were upregulated in the CLP group compared to the sham group, while the levels of these cytokines were downregulated with cortistatin treatment. Interestingly, the level of IL-6 in the serum returned to normal seven days after the surgery, and cortistatin showed no impact on IL-6 levels at this time point (Figure 3(b)). Similar trends were observed in the cortex (Figure 3(b)).

Microglia are the major local macrophages in the CNS that orchestrate the neuroinflammation. Therefore, we investigated the activation of microglia in the brain. Immunofluorescence was used to evaluate the morphology and distribution of microglia, as investigated by Iba1 immunoreactivity. At 1 and 7 days after surgery, the number of microglia was increased in the frontal cortex. Moreover, as shown in Figure 4(a), the cell body of microglia in CLP mice was enlarged with a thick, shrunken process and aligned with the ameboid morphology of activated microglia. In contrast, microglia had small cell bodies with fine and long processes in the sham and cortistatin groups in the mouse brain, consistent with the amplified morphology of resting microglia. In septic mice, the number of active microglia in the brain was higher than that in the sham group. Treatment with cortistatin significantly decreased the number of microglia (Figure 4(b)). Compared to the CLP group, the area of the cell body was decreased in the cortistatin group (Figure 4(c)).

3.4. Cortistatin ameliorated sepsis-induced BBB disruption in the brain of the CLP -induced mice

To explain the neuroprotective effect of cortistatin on the cognition and memory in septic mice, we explored the ultrastructure of the brain using transmission electron microscopy. One day after the CLP surgery, BBB integrity was investigated using transmission electron microscopy (Figure 4(d)). In the sham group, the morphology of the endothelial cells remained normal with no deformation or swelling, the tight junctions of the plasma membrane of adjacent endothelial cells were intact, and no gaps were observed. In the CLP group, disrupted tight junctions and swollen mitochondria in the endothelial cells were observed, and these injuries were clearly attenuated by cortistatin treatment. Similarly, Evans Blue staining also confirmed that, in the acute phase (1 day after septic onset), BBB integrity was disturbed in the brains of septic mice compared with sham mice, and cortistatin treatment could improve Evans Blue extravasation (Figure 4(e)).