Exploring Molecular Mechanisms Involved in the Development of the Depression-Like Phenotype in Interleukin-18-Deficient Mice

Interleukin-18 (IL-18) is an inflammatory cytokine that has been linked to energy homeostasis and psychiatric symptoms such as depression and cognitive impairment. We previously revealed that deficiency in IL-18 led to hippocampal abnormalities and resulted in depression-like symptoms. However, the impact of IL-18 deficiency on other brain regions remains to be clarified. In this study, we first sought to confirm that IL-18 expression in neural cells can be found in human brain tissue. Subsequently, we examined the expression of genes in the prefrontal cortex of Il18−/− mice and compared it with gene expression in mice subjected to a chronic mild stress model of depression. Extracted genes were further analyzed using Ingenuity® Pathway Analysis, in which 18 genes common to both the chronic mild stressed model and Il18−/− mice were identified. Of those, 16 were significantly differentially expressed between Il18+/+ and Il18−/− mice. We additionally measured protein expression of α-2-HS-glycoprotein (AHSG) and transthyretin (TTR) in serum and the brain. In the prefrontal cortex of Il18−/− mice, TTR but not AHSG was significantly decreased. Conversely, in the serum of Il18−/− mice, AHSG was significantly increased but not TTR. Therefore, our results suggest that in IL-18-deficit conditions, TTR in the brain is one of the mediators causally related to depression, and AHSG in peripheral organs is one of the regulators inducing energy imbalance. Moreover, this study suggests a possible “signpost” to clarify the molecular mechanisms commonly underlying the immune system, energy metabolism, neural function, and depressive disorders.


Introduction
Major depressive disorder (MDD) is common and increasing in prevalence worldwide, which makes it an urgent health issue. MDD is known to interfere with people's dayto-day activities such as going to work, and these patients report more negative social interactions [1]. In addition, depressive symptoms might make patients more sensitive in everyday experiences of both social acceptance and rejec-tion [2]. Therefore, MDD affects patients' daily life and social interactions.
In clinical studies, reduced volumes of the PFC and hippocampus were revealed by neuroimaging [7,9], and abnormal activity of the hypothalamic-pituitary-adrenal (HPA) axis in depressed patients has also been reported [10]. Moreover, the HPA axis and hippocampal homeostasis were shown to be important to prevent the progression of depression [11]. Even though these findings seem to support the possible biological mechanisms of MDD, including the principal brain regions involved [12], the etiological mechanism underlying the development of MDD remains unclear.
IL-18 was initially purified as a proinflammatory cytokine that induces interferon-(IFN-) γ [13]. IL-18 is secreted in an active 18 kDa mature form, metabolized by cleaved caspase-1 from a nonactive 24 kDa precursor [14][15][16][17]. IL-18 is associated with energy metabolism and psychiatric disorders such as depression [18][19][20]. Evidence for a role in energy metabolism has been demonstrated in IL-18-deficient mice, which presented diabetes mellitus and dyslipidemia, which leads to nonalcoholic fatty liver disease and steatohepatitis [18,19,21]. In the central nervous system, a deficit in IL-18 has been shown to induce hippocampal abnormalities and result in a depressive-like behavioral change [20]. Thus, IL-18 is closely associated with the immune system, energy balance, neural function, and depressive disorders.
In our previous study, these biological roles of IL-18 were intensively examined. In the central nervous system, IL-18-deficient mice showed significantly impaired learning and memory function and exhibited lower motivation, which resulted in depressive-like behaviors. However, only the hippocampal function was evaluated [20].
Hence, we herein sought to determine the molecular expression based on the development of depression in the condition of IL-18 deficiency, especially focusing on the PFC. To do so, we designed a four-step study: (1) confirming IL-18 protein expression in human neurons, (2) comparing gene expression in IL-18-deficient mice with that of the mouse chronic mild stress (CMS) model of depression, (3) evaluating differences in gene expression between the PFC of Il18 −/− and Il18 +/+ mice by reverse transcription quantitative polymerase chain reaction (RT-qPCR), and (4) measuring the protein expression encoded by identified genes in serum, the PFC, hypothalamus, and hippocampus of Il18 +/+ and Il18 −/− mice.

Materials and Methods
2.1. Animals. Il18 −/− male mice were generated on a C57Bl/6 background as previously described [22]. As controls, we used C57Bl/6 Il18 +/+ male littermates. Mice were housed in groups of three to five in polycarbonate cages with free access to water and standard food in a colony room maintained under controlled conditions as follows: temperature (22 ± 1°C), humidity (50%-60%), and a 12 h light/dark cycle (lights on at 08:00 am). Mice were sacrificed at 12 weeks of age at 10:00 am. Twelve and six Il18 +/+ and Il18 −/− mice, respectively, were used for RT-qPCR and western blotting, and an enzyme-linked immunosorbent assay (ELISA) was performed using serum samples (n = 18 for each). In total, eighteen Il18 +/+ and Il18 −/− mice were used in the present study. The numbers of mice in this study were decided based on our previous study [20,23], and the PFC extracted in our previous research was used in the current study to minimize the number of mice sacrificed. All animal experiments were conducted according to the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health, and all protocols were approved by the Animal Care Committee of Hyogo College of Medicine (Hyogo, Japan; approval nos. A09-206, #28041, #16-013, and #19-046).

Sample
Collection. Mice were euthanized by deep anesthesia using 5% isoflurane inhalation (overdose) for at least 5 min. After confirming shallow or absent breathing or a lack of response to pain stimulation, blood from their heart was collected immediately. The serum and brain regions (PFC, hypothalamus, and hippocampus) were removed as previously described, immediately placed in liquid nitrogen, and stored at −80°C until required [24,25] 2.4. Immunofluorescence Staining. Immunofluorescence staining was performed to assess the expression of IL-18 and microtubule-associated protein 2 (MAP2), a mature neuron marker, in the human PFC. Human brain sections were deparaffinized and incubated in HistoVT One antigen retrieval buffer at 90°C for 40 min (cat. no. 06380-76; Nacalai Tesque Inc. Kyoto, Japan). The sections were blocked with 5% bovine serum albumin in phosphate-buffered saline (PBS) for 1 h at room temperature. They were then incubated in Can Get Signal™ immunoreaction enhancer solution (cat. no. NKB-601; Toyobo Co. Ltd., Tokyo, Japan) with primary antibodies overnight at 4°C. The primary antibodies included anti-mouse IL-18 (dilution 1 : 50, cat. no. 60070-1-lg; Proteintech Japan, Tokyo, Japan) and antirabbit MAP2 (dilution 1 : 500, cat. no. AB5622; Merck Millipore, Burlington, MA, USA). Anti-mouse IL-18 antibody specificity was first validated by western blot using HeLa whole cell lysates (cat. no. sc-2200; Santa Cruz Biotechnology, Inc., Heidelberg, Germany) as a positive control in accordance with the manufacturer's protocol and previous studies [26,27]. For immunofluorescence, fluorochromeconjugated goat anti-mouse IgG H&L secondary antibody (dilution 1 : 200, Alexa Fluor® 488, cat. no. A-11011, Thermo Fisher Scientific K.K., Tokyo, Japan) or goat anti-rabbit IgG H&L secondary antibody (dilution 1 : 1000, Alexa Fluor® 555, cat. no. ab150078; Abcam, Cambridge, UK) was added and incubated for 2 h at room temperature. After washing with PBS, sections were cover-slipped with Vectashield mounting medium with 4,6-diamidino-2-phenylindole (DAPI) (cat. no. H-1800; Vector Laboratories, Inc., Burlingame, CA, USA). The sections were visualized with a fluorescence microscope using red and green filters and blue light (LSM780, Carl Zeiss Co. Ltd., Tokyo, Japan). The images were scanned using ZEN Imaging Software (Carl Zeiss Co. Ltd.) for analysis.

Molecular Analysis.
Microarray data from the PFC of CMS model mice [28] and from the hippocampus of Il18 −/− mice were obtained in our previous studies [20,24].
Their series entries are GSE49867 and GSE108485, respectively. Molecules showing expression increases of more than twofold or decreases of more than half were identified and then further narrowed down to those expressed in common between the mouse models.
2.6. Ingenuity Pathway Analysis (IPA). IPA (QIAGEN Digital Insights, Redwood City, CA, USA) was used to search for functions of extracted genes as previously described [23]. The molecules identified in the microarray results were further limited to those related to psychiatric diseases (depression, dementia, Alzheimer's disease (AD), and cognitive impairment) as suggested by observing the behavior of "CMS mice" and "Il18 −/− mice" represented fold change by log 2 ratio of CMS/control and Il18 −/− /Il18 +/+ , respectively.

BioMed Research International
Il18 −/− mice. Core analysis was performed to analyze the functions of these molecules as follows: tissue, brain; all other settings, default. The network explorer of IPA was run with default settings to detect molecule-molecule interactions and to reveal pathways between molecules.

RNA Purification.
Total RNA was extracted and purified as previously described, with minor modifications [19]. Briefly, RNA was extracted using ISOGEN (cat. no. 311-02501, Wako Pure Chemical Industries, Ltd., Osaka, Japan), following the manufacturer's protocol, and then incubated with 5 U of RNase inhibitor and 5 U of DNase I at 37°C for 30 min. After phenol/chloroform purification and ethanol precipitation, RNA was dissolved in sterilized and deionized distilled water.
2.8. RT-qPCR. RT-qPCR was performed as previously described [19,23]. Briefly, total RNA (10 ng/reaction) was used in the RNA-direct SYBR Green Real-Time PCR Master Mix: One-step qPCR kit (Toyobo Co. Ltd.) according to the manufacturer's protocol. Samples were examined in duplicate reactions in 384-well plates. The median threshold cycle values were used to calculate fold changes between the groups. The following cycling conditions were used: 30 s at 90°C and 20 min at 61°C for reverse transcription, followed by 45 cycles of 98°C for 1 s, 67°C for 15 s, and 74°C for 35 s. Fold-change values were normalized to β-actin (Actb) levels using the relative standard curve method. Primer sequences for RT-qPCR are shown in Table 1.

Identification of Molecules Commonly Expressed in CMS
and Il18 −/− Mice. We previously isolated a total of 494 and 294 genes whose expression was increased more than twofold or decreased by more than half from CMS and Il18 −/− mice, respectively [20,24]. To examine which genes were responsible for causing depression under IL-18-deficit conditions, we searched for molecules expressed in both CMS and Il18 −/− mice. We detected 42 such genes, of which 18 encode molecules shown by IPA to be associated with psychiatric disorders. The summarized results are shown in Figure 2(a), and detailed gene information including primer sequences is shown in Table 1.

Disease Association, Functional Annotation, and
Interaction among 18 Extracted Genes. We performed IPA core analysis to further analyze the disease associations and functional annotations of the extracted genes ( Table 2).
Pathway analysis was then performed to elucidate interactions among these genes (Figure 2(b)). Several molecules were shown to be related to each other; however, we observed no interactions between Il18 and these molecules.     18 showing that no association was found. (c) mRNA expression of these genes by quantitative reverse transcription polymerase chain reaction (n = 12 per group); * P < 0:05 and * * P < 0:01. The genes are listed in Table 1.

Protein Expression of AHSG and TTR in the PFC,
Hypothalamus, and Hippocampus of Il18 −/− and Il18 +/+ Mice. RT-qPCR showed that mRNA expression decreased significantly for 16 of 18 molecules in Il18 −/− mice. In our previous study [25], we proposed that Ahsg and S100 calcium binding protein A9 (S100a9) were key mediators not only of depression but also of physiological dysregulation, for example, diabetes mellitus. Additionally, Ttr was suggested to be an inducer of MDD. Therefore, we focused on the protein expression of AHSG and TTR, which showed significant differences in mRNA expression between Il18 −/− and Il18 +/+ mice. Contrary to our expectations, only TTR expression in the PFC of Il18 −/− mice was significantly decreased compared with Il18 +/+ mice, whereas AHSG expression was not changed (n = 6 per group, P < 0:05, Figures 3(a) and 3(b)).

Protein Expression of AHSG and TTR in the Serum of
Il18 −/− and Il18 +/+ Mice. The serum level of IL-18 was not detected in Il18 −/− mice (n = 18 per group, Figure 4). To test our hypothesis that AHSG and TTR could be markers linked to MDD and the dysregulation of physiological homeostasis, we measured serum AHSG and TTR levels in Il18 −/− and Il18 +/+ mice. Serum AHSG, but not TTR, was significantly higher in Il18 −/− mice than in Il18 +/+ mice (n = 10 per group, P < 0:05, Figure 4).

Discussion
In this study, we profiled the function of IL-18 in the central nervous system as follows: (1) IL-18 was found in the cytoplasm of human neurons; (2) 16 genes expressed in the PFC of Il18 −/− mice were shared with a CMS model of depression; (3) TTR in the brain, including the PFC, was a mediator associated with depression in IL-18 deficiency; and (4) AHSG in the serum was associated with the dysregulation of physiological homeostasis, such as energy imbalance or immune impairment.
IL-18 is a proinflammatory cytokine promoting IFN-γ production that is secreted by Th1 immune cells when exposed to lipopolysaccharide stimulation [13]. Therefore, in the central nervous system, microglia can secrete IL-18  [29]. In the human brain, IL18 gene expression in the PFC has been examined previously; however, its protein expression has not [30]. Figure 1 shows the costaining of IL-18 and MAP2 in the PFC of both a healthy control and an MDD patient. Thus, our results showed the existence of a precursor form of IL-18 with a molecular weight of approximately 24 kDa in neural cells, suggesting that IL-18 might play a role in neurons of the PFC. From the results of the core analysis shown in Table 2, we identified C3, C5, Kng1, and Ttr as genes showing decreased expression in the PFC of Il18 −/− mice compared with Il18 +/+ mice (Figure 2(c)). C3 is one of the main factors of inflammatory signaling pathways that promotes IL-1β secretion through activation of the NLR family pyrin domain-containing 3 inflammasome, including IL-18 [31]. The complement system exerts some effect on the central nervous system, and its impairment might induce the progression of neurodegenerative disorders such as AD [32]. Moreover, clinical studies have found reduced expression of C3 in both the serum and cerebrospinal fluid (CSF) of MDD patients [33,34]. In an animal model of AD with C3 deficiency, increased amyloid-β (Aβ) plaque and neurodegeneration in the PFC and hippocampus were observed, suggesting that cerebral C3 may play an important role in inhibiting the progression of AD [35]. As shown in Table 1 and Figure 2(c), C3 expression was decreased in the PFC of Il18 −/− mice but increased in the hippocampus.
With regard to C5, a clinical study revealed that CSF C5 expression was significantly increased in MDD patients, resulting in microglial neuroinflammation [36]. Furthermore, bradykinin, a final product of Kng1, may be associated with the emergence of depression [37]. Decreased bradykinin is thought to reduce brain-derived neurotrophic factor release, inhibit synaptogenesis, and enhance the progression of depression [37]. Even though the role of these two molecules in the central nervous system remains unclear, they might connect depression with neuroinflammation. In sum, these immune-related molecular changes suggest that an IL-18 deficiency causes the abnormal activation of immune pathways and affects neural plasticity, leading to psychiatric disorders such as MDD and AD, though other inflammatory factors should be considered.
Disruption of the blood-brain barrier (BBB) might also be a crucial factor in depression with some immune impairment [38,39]. C3a and bradykinin are vasoactive molecules that disrupt the BBB [40], while IL-18 helps to mitigate such disruption [41]. AHSG might inhibit the brain inflammatory response in the early stage of cerebral ischemic injury, resulting in the disruption of the BBB [42]. TTR has an ability to cross the BBB from the brain to the blood; however, it cannot cross in the opposite direction [43]. Therefore, TTR might play a role in removing the peptide from the brain and inhibiting the reverse flow [43]. We detected reduced C3 and Kng1 mRNA expression, decreased TTR protein in the PFC, and increased AHSG expression in the serum in Il18 −/− mice compared with Il18 +/+ mice (Figure 2(c)). These results indicate that IL-18 deficiency might influence the BBB function via C3 and Kng1, leading to the development of a depressive phenotype, although AHSG might protect against neuroinflammation under a depressive state.
In a clinical study, serum IL-18 levels of depressed patients were increased compared with healthy controls [44], while decreased C3 in the CSF of depressed patients suppressed inflammasome activation and decreased IL-18 [31]. Therefore, our results suggest that increased IL-18 may be a compensatory response in MDD patients, and that IL-18 may have protective roles against inflammation in the brain of MDD patients. Furthermore, these findings also indicate that neuroinflammation is a major mechanism involved in the progression of MDD.
AHSG is secreted into the blood mainly by the liver and is also expressed in the cerebral cortex, thalamus, and hippocampus [45,46]. Based on previous studies, AHSG functions in the induction of inflammation and energy balance [47,48]. During the inflammatory response, AHSG synthesis is downregulated by IL-6, TNF-α, and IFN-γ [49,50]. IL-18 has a role in IFN-γ induction; therefore, increased serum AHSG levels might be caused by reduced IFN-γ following IL-18 deficiency, as shown in Figure 4.

BioMed Research International
In energy homeostasis, several studies have reported a relationship between AHSG and the glucose and lipid balance. Indeed, higher plasma AHSG levels were linked to higher insulin resistance and fat accumulation in the liver [51]. Moreover, increased serum AHSG was correlated with the risk of metabolic syndrome, including higher triglycerides [48]. Il18 −/− mice have been reported to show higher insulin resistance, diabetes mellitus, obesity, and dyslipidemia [18,19,21], and their serum AHSG levels were higher than those of Il18 +/+ mice ( Figure 4); hence, our results are consistent with previous studies. Furthermore, higher AHSG was reported in male MDD patients compared with controls [52]. While AHSG expression in the PFC, but not the thalamus and hippocampus, was increased in our depressive model mice, its expression in Il18 −/− mice was unchanged [25] (Figure 3(b)). Thus, these results suggest that AHSG is a key factor connecting depression, energy homeostasis, and the immune system in IL-18 deficiency.
According to the Allen Brain Atlas Human (https:// human.brain-map.org/), TTR is expressed in the PFC, hypothalamus, and hippocampus [53,54]. In depressed patients, TTR was significantly decreased in the CSF compared with healthy controls [55]. Similar results were obtained in AD studies [56]. TTR was suggested to be involved in Aβ protein degradation and clearance; thus, reduced TTR in the brain may increase Aβ levels and cause AD progression [56]. Moreover, TTR-deficient mice exhibited fewer depressive behavioral changes through increased levels of noradrenaline, but not dopamine or serotonin, in the limbic forebrain, and it was suggested that TTR has an important role in resilience to stress exposure [55,57]. In our previous study, Il18 −/− mice showed less motivation and had impaired learning and memory, which are clinically included in depression or dementia [20]. Furthermore, TTR expression in the PFC of Il18 −/− mice was significantly decreased compared with Il18 +/+ mice in this study (Figure 3(b)). However, serum TTR levels were unchanged (Figure 4). No correlation was observed between serum TTR and CSF in a human study, indicating that TTR expression in the brain may differ from that in peripheral organs [58]. Thus, lower TTR of the PFC in Il18 −/− mice might be a mediating factor of depression.
Regarding the limitations of our study, because of the limited number of human samples, the immunofluorescence results were insufficient. To validate our hypothesis, additional human samples are required. Second, we discussed the relationship among BBB, depression, and our extracted genes; however, we did not analyze the function of the BBB. Additional experiments will be indispensable to clarify the function of the BBB in IL-18 deficiency. Third, we focused on AHSG and TTR in this study; however, the details of their functions are still unclear. Therefore, mechanistic analysis is needed to explore their functions. Moreover, our study was limited by the fact that we only measured AHSG and TTR in the serum and not in CSF of Il18 +/+ and Il18 −/− mice. To further examine the validity of our hypothesis, measurements in CSF are warranted.

Conclusion
In conclusion, we investigated the effects of IL-18 deficiency on the central nervous system and its association with psychiatric disorders. First, we demonstrated the presence of IL-18 in the cytoplasm of human PFC neurons. Then, we identified 16 molecules that might cause the depressive characteristics of Il18 −/− mice; in particular, decreased TTR in the PFC might be a mediator for developing depression. Finally, AHSG in peripheral organs might also be an inducer of immune and energy imbalance in Il18 −/− mice. Although these identified molecules are related to several psychiatric disorders such as MDD and AD, the mechanism by which abnormal immune function affects the central nervous system and psychiatric diseases remains unclear. IL-18 might be associated with a potential strategy for elucidating this mechanism and novel treatments for MDD and AD. It might also be a potential "signpost" to clarify the molecular links among the immune system, energy metabolism, neural function, and psychiatric diseases, though further study is required for confirmation. . Serum AHSG and TTR protein expression was measured to compare the central nervous system and peripheral response. In Il18 −/− mice, serum AHSG levels, but not TTR, were significantly higher than in Il18 +/+ mice. n = 10 per group; * P < 0:05.

Data Availability
The datasets shown and/or analyzed in the present study are available from the corresponding author upon reasonable request.

Ethical Approval
Our animal experiments were conducted according to the "Guide for Care and Use of Laboratory Animals" published by the National Institutes of Health and approved by the Animal Care Committee of Hyogo College of Medicine (Hyogo, Japan) (approval nos. A09-206, #28041, #16-013, and #19-046). All efforts were made to minimize the number of mice used and the suffering of animals.

Consent
BioChain Institute Inc. disclosed that all human tissue samples were collected with informed consent from the donors and their relatives.

Disclosure
The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.