Morinda officinalis oligosaccharides alleviate depressive‐like behaviors in post‐stroke rats via suppressing NLRP3 inflammasome to inhibit hippocampal inflammation

Abstract Aims Morinda officinalis oligosaccharides (MOOs), a traditional Chinese medicine, have been used to treat mild and moderate depressive episodes. In this study, we investigated whether MOOs can ameliorate depressive‐like behaviors in post‐stroke depression (PSD) rats and further explored its mechanism by suppressing microglial NLRP3 inflammasome activation to inhibit hippocampal inflammation. Methods Behavioral tests were performed to evaluate the effect of MOOs on depressive‐like behaviors in PSD rats. The effects of MOOs on the expression of IL‐18, IL‐1β, and nucleotide‐binding domain leucine‐rich repeat (NLR) family pyrin domain containing 3 (NLRP3) inflammasome were measured in both PSD rats and lipopolysaccharide (LPS) and adenosine triphosphate (ATP) stimulated primary rat microglia by reverse transcription polymerase chain reaction (RT‐PCR), immunofluorescence and Western blot analysis. Adeno‐associated virus (AAV) was injected into the hippocampus to regulate NLRP3 inflammasome expression. The detailed molecular mechanism underlying the effects of MOOs was analyzed by Western blot and immunofluorescence. Results MOOs can alleviate depressive‐like behaviors in PSD rats. PSD rats showed increased expression of IL‐18, IL‐1β, and NLRP3 inflammasome in the ischemic hippocampus, while MOOs reversed the elevation. NLRP3 downregulation ameliorated depressive‐like behaviors and hippocampal inflammation response in PSD rats, while NLRP3 upregulation inhibited the effect of MOOs on depressive‐like behaviors and hippocampal inflammation response in PSD rats. Moreover, we found that NLRP3 was mainly expressed on microglia. In vitro, MOOs effectively inhibited the expression of IL‐18, IL‐1β, and NLRP3 inflammasome in LPS + ATP treated primary rat microglia. We also showed that modulation of NLRP3 inflammasome by MOOs was associated with the IκB/NF‐κB p65 signaling pathway. Conclusion Overall, our study reveals the antidepressive effect of MOOs on PSD rats through modulation of microglial NLRP3 inflammasome. We also provide a novel insight into hippocampal inflammation response in PSD pathology and put forward NLRP3 inflammasome as a potential therapeutic target for PSD.


| INTRODUC TI ON
Post-stroke depression (PSD) refers to depression after cerebrovascular incident. 1 The incidence of PSD is almost one-third in stroke survivals. 2 A large number of studies have indicated that PSD is positively associated with poor functional outcomes and increased mortality after stroke, 3,4 which gains public concerns. Currently, the treatment of PSD mainly relies on antidepressants such as selective serotonin reuptake inhibitors, 5 while side effects remain a problem. 6 Thus, it is urgent to find novel targets for therapeutic approaches to PSD.
In our previous study, we found that PSD rats showed suppressed glucose metabolism due to decreased glucose transporter-3 expression in the neurons of the medial prefrontal cortex but not in the hippocampus, which is responsible for the occurrence of PSD. 7 However, the hippocampus has been demonstrated to be an important regulator of emotion and cognition in the brain, 8,9 and evidence has also suggested the correlation between depression, hippocampal inflammation, and stroke. Studies have reported high serum levels of pro-inflammatory biomarkers in depressive patients, 10,11 while several anti-inflammatory agents such as celecoxib have been found to alleviate depressive symptoms in clinical trials. 12 Inhibition of hippocampal inflammation has been proven to ameliorate depressive behaviors in mice. 13 Moreover, cerebral ischemia can lead to a widespread inflammatory response in the brain, which can further affect the structure of the hippocampus. 14 In clinical trials, stroke patients with higher serum IL-18 or IL-1β levels at admission have a higher risk of developing PSD. 15,16 Thus, excessive hippocampal inflammation may induce depression in stroke survivals.
Nucleotide-binding domain leucine-rich repeat family pyrin domain containing 3 (NLRP3) inflammasome, which is comprised of NLRP3, caspase-1, and apoptosis-associated speck-like protein (ASC), plays an important role in inflammatory response via regulating the maturation of IL-1β and IL-18. 17 In the central nervous system, NLRP3 inflammasome is mainly expressed on microglia 18 and has been extensively reported to participate in brain disorders. In ischemic stroke, NLRP3 inflammasome activation can drive inflammation, 19 increase BBB permeability 20 and exacerbate recurrent strokes. 21 In Alzheimer's disease and vascular dementia, NLRP3 inflammasome is activated 22 and associated with cognitive impairment. 23 Selective NLRP3 inflammasome inhibitor has been shown to alleviate neurological deficits and reduce IL-1β production in intracerebral hemorrhage mice. 24 Moreover, microglial NLRP3 inflammasome activation in the hippocampus has been found to mediate chronic stress-induced depressive-like behaviors in rats. 25 These findings strongly imply that microglial NLRP3 inflammasome activation is firmly associated with depression. However, whether it is involved in PSD remains unknown.
Morinda officinalis oligosaccharides (MOOs), the traditional Chinese medicine extracted from Morinda officinalis roots, have long been used as a tonic to nourish the kidneys and strengthen immunity in humans. 26 The Chinese Food and Drug Administration (CFDA) approved MOOs as a prescribed traditional herbal medicine for mild and moderate depressive episodes in 2012. There have been reports that oral administration of MOOs to depressed rodent animals can increase the monoamine and brain-derived neurotrophic factor (BDNF) levels, which suggests the alleviation of depression. 26,27 In addition, studies have also reported that extracts from Morinda officinalis have an effect on colitis by regulating inflammation and cell apoptosis. 28,29 In this study, we found that MOOs could ameliorate depressivelike behaviors and hippocampal inflammation in PSD rat models via suppressing microglial NLRP3 inflammasome activation. We further demonstrated the mechanism by which MOOs influence the IκB/NF-κB p65 pathway to downregulate NLRP3 inflammasome expression in PSD rats. We also propose NLRP3 inflammasome as a potential therapeutic target for PSD and provide a novel insight into hippocampal microglial inflammation response in PSD pathology. Conclusion: Overall, our study reveals the antidepressive effect of MOOs on PSD rats through modulation of microglial NLRP3 inflammasome. We also provide a novel insight into hippocampal inflammation response in PSD pathology and put forward NLRP3 inflammasome as a potential therapeutic target for PSD.

K E Y W O R D S
hippocampus, Morinda officinalis oligosaccharides, NLRP3 inflammasome, post-stroke depression described procedures. 7 The filament was withdrawn for reperfu-

| Drugs
MOOs were provided by Beijing Tongrentang Ltd. Co., Beijing, China and dissolved at a concentration of 10 mg/mL in distilled water.
During the last 2 weeks of PSD model establishment, the PSD rats were orally administered with MOOs at a dose of 0.1 mg/g/d. The control groups were orally administered with vehicle of the same volume.

| Tail suspension test
Generally, adhesive tape was attached to the tail of the rats and they were then suspended 50 cm above the floor and observed for 6 min.
The climbing time and immobility time were recorded as previously described. 31

| Forced swimming test
On the first day, rats were placed in a transparent cylinder (50 cm, height; 30 cm, diameter) filled with water (30 cm, depth; 25°C, water temperature) for 6 min to adapt to the experiment environment.
On the second day, rats were exposed to the same cylinder and observed for 6 min. The climbing time and immobility time were recorded as previously described. 32

| Open field test
The Open field test (OFT) was performed to determine the motor ability of the rats before other behavioral tests to exclude motor deficiency as previously described. 33 Briefly, rats were placed in a square wooden box (100 cm × 100 cm × 50 cm) for 5 min. Anilab software (Anilab Scientific Instruments Ltd. Co, Ningbo, China.) was used to analyze the traveling distance and average speed of the rats.

| Differential expression analysis and functional enrichment analysis
R software and the limma package were used to identify the Differentially Expressed Genes (DEGs) of each GEO dataset. DAVID 6.8 (https://david.ncifc rf.gov/tools.jsp) was applied for functional enrichment analysis of DEGs and the top ten biological process terms were selected according to the P-value of each term.

| Viral injection
After 2 weeks of the tMCAO procedure, the rats were aseptically injected with 10% pentobarbital to induce anesthesia. Then, stereotaxic frames were applied to fix the anesthetized rats. Adenoassociated virus (AAV) solution (1 × 10 13 vg/ml) was injected into the hippocampus at a rate of 0.05 μL/min (total volume 2 μL).

| Quantitative real-time PCR
Total RNA was isolated using TRIzol reagent from Invitrogen. RNA extraction was reverse-transcribed using the cDNA Synthesis Kit (Takara, Kyoto, Japan) according to the manufacturer's protocol, followed by the amplification of the cDNA using the SYBR Premix Ex Taq TM Kit (Takara, Kyoto, Japan). Then, quantitative real-time PCR was performed using the thermocycler (Bio-Rad, Hercules, CA, USA). The primers are shown below.

| Statistical analyses
All values were expressed as mean ± SEM. Shapiro-Wilk test was used to evaluate the distribution of the data. Data that did not exhibit a Gaussian distribution was analyzed via a non-parametric equivalent. For parametric analysis, differences between groups were evaluated by the one-way ANOVA of Fisher's least significant difference test or the two-tailed Student's t-test using GraphPad Prism 7 software. Values of p < 0.05 were considered significant, and values of p < 0.01 were considered markedly significant.

| MOOs alleviate depressive-like behaviors in PSD rats
To study the antidepressive effect of MOOs, we used SD rats to establish the PSD model. We randomly divided the SD rats into 4 groups: sham groups administered with vehicle or MOOs (Ctrl + v group and Ctrl + M group) and PSD groups administered with vehicle or MOOs (PSD + v group and PSD + M group). As shown in Figure 1A, we first performed the tMACO operation on PSD rats.
After 1 week's recovery from surgery, the PSD rats then underwent individual housing and CUMS stimulation for up to 5 weeks. We administered MOOs or vehicle to the PSD rats orally in the last two weeks of PSD model establishment. The sham groups only experienced the same surgical operation, without tMACO, CUMS stimulation or individual housing. Finally, we performed behavioral tests to assess their depressive-like behaviors ( Figure 1A).
From the behavioral tests, we found that depressive-like behaviors were more obvious in PSD rats. The immobility time of PSD + v rats was dramatically increased, while the climbing time was significantly decreased in the Tail suspension test (TST) and Forced swim- In SPT test, PSD + v rats showed reduced sucrose preference ( Figure 1F). Orally administration of MOOs to PSD rats can significantly ameliorate depressive-like behaviors. In the TST and FST test, PSD + M rats showed more climbing time and less immobility time than PSD + v rats ( Figure 1B-E), while the sucrose preference did not increase in PSD + M rats ( Figure 1F), which means that MOOs treatment did not alleviate anhedonia in PSD rats. To exclude the possibility that motor dysfunction caused by stroke may have influenced the behavioral tests, we performed OFT and weighted the rats. No significant difference in traveling distance and average speed was found among the four groups in the OFT and rats' weight in the four groups was the same (Figure 1G-I). From these results, it can be concluded that MOOs can alleviate depressive-like behaviors in PSD rats.

| MOOs inhibit the inflammatory response in the ischemic hippocampus of PSD rats
Cerebral ischemia can lead to the inflammatory response in the brain, and excessive inflammation, especially in the hippocampus, has been found to be associated with depression. 13 Extracts of Morinda officinalis have also been found to regulate intestinal inflammation in mice. 28 Thus, we speculated that MOOs might ameliorate depressive-like behaviors via suppressing inflammatory response in the ischemic hippocampus of PSD rats.
We downloaded the stroke-related gene expression profiling dataset GSE80681 and depression-related gene expression profiling dataset GSE12 4387 from the GEO database and analyzed the DEGs using R software. Totally, 7956 DEGs were identified in GSE12 4387 and 3644 DEGs were identified in GSE80681. As shown in Figure 2A, 819 DEGs were detected in both GSE80681 and GSE12 4387. Using DAVID 6.8, we found that the overlapped 819 DEGs were enriched in several inflammation-related biological processes ( Figure 2B). The genes involved in the inflammation-related biological processes are shown in Table 1. Among them, the levels of IL-1β, IL-18 as well as TNFα in serum have been reported to be elevated in PSD patients. 15,16 Next, we validated the mRNA expression of IL-1β, IL-18, and TNFα in the rat hippocampus using RT-PCR analysis.
As shown in Figure 2C-  Figure 3D and 3H). We further performed immunofluorescence staining and showed that NLRP3 inflammasome was mainly expressed in Iba-1 + cells, the normally used marker for microglia ( Figure 3I). PSD + v rats exhibited increased microglial NLRP3 inflammasome expression in the ischemic hippocampus, while MOOs can reverse the elevation ( Figure 3I). In contrast, there was rare colocalization of GFAP, a marker of astrocytes, and NLRP3 in the hippocampus in all four groups ( Figure S1A). These results suggested that MOOs can suppress the microglial NLRP3 inflammasome activation in the hippocampus of PSD rats.

| NLRP3 downregulation ameliorates depressive-like behaviors and hippocampal inflammation response in PSD rats
In order to further demonstrate the participation of NLRP3 inflam- Then, we performed behavioral tests to evaluate the effect of NLRP3 downregulation on depressive-like behaviors of PSD rats.
Compared with PSD + sh-NC rats, the immobility time was decreased and the climbing time was increased in PSD + sh-NLRP3 rats in both TST and FST tests ( Figure 4D-G), while Ctrl + sh-NLRP3 rats showed no changed behavior phenotype ( Figure 4D-G). Our data showed that NLRP3 downregulation can ameliorate depressive-like behavior in PSD rats. In accordance to Figure 1F, no differences were detected between PSD + sh-NC and PSD + sh-NLRP3 on SPT ( Figure 4H). Moreover, the traveling distance and average speed in the OFT and weight showed no difference ( Figure 4I-K), excluding the motor disability among the 4 groups.
To further investigate whether NLRP3 downregulation is adequate to alleviate hippocampal inflammation in PSD rats, we performed Western blot to measure the protein expression of pro-Caspase1, Caspase1 p20, IL-1β, and IL-18 in the hippocampus. As shown in Figure 4L-P, shRNA NLRP3 transduction to PSD rats significantly decreased the Caspase1 p20, IL-1β, and IL-18 expression in the hippocampus compared with PSD + sh-NC rats without affecting expression of pro-Caspase1. Transduction of shRNA NLRP3 in control rats did not influence the Caspase1 p20, IL-1β, and IL-18 protein expression ( Figure 4L-P). Taken together, these results confirmed that downregulating NLRP3 was adequate to abrogate hippocampal inflammation response in PSD rats.

| NLRP3 upregulation abrogates the effect of MOOs on depressive-like behaviors and hippocampal inflammation response in PSD rats
Next, to fully confirm whether MOOs function through NLRP3 inflammasome in PSD rats, we microinjected MOOs-treated PSD rats with overexpression NLRP3 AAV (PSD +MOOs + OE-NLRP3) or negative control AAV (PSD +MOOs + OE-NC) using stereotaxic frames 2 weeks after tMCAO ( Figure 5A). After overexpression NLRP3 AAV injection, we found that the NLRP3 expression in the hippocampus of the MOOs-treated PSD rats was successfully increased ( Figure 5J-K).
Then, behavioral tests were performed to evaluate the effect of Hippocampal inflammation response was also tested in all 4 groups. As shown in Figure 5J-N, overexpression NLRP3 AAV transduction to MOOs-treated PSD rats significantly increased the Caspase1 p20, IL-1β, and IL-18 expression in the hippocampus compared with PSD + MOOs + OE-NC rats. In conclusion, it is confirmed that NLRP3 upregulation could abrogate the effect of MOOs on inhibiting the hippocampal inflammation response in PSD rats. Previous studies have demonstrated that NLRP3 inflammasome activation can be regulated by transcriptional factor NF-κB. 17 Thus, we measured the pIκB, IκB, pNF-κB p65, and NF-κB p65 protein expression in LPS + ATP treated primary rat microglia in the presence or absence of MOOs. We found that levels of phosphorylated IκB (p-IκB), p-IκB/IκB, phosphorylated NF-κB p65 (p-p65) and p-p65/ p65 were significantly increased in LPS + ATP treated primary rat microglia, which indicated the activation of IκB/NF-κB signaling pathway. MOOs treatment can reverse these changes, while total

| MOOs inhibit the IκB/NF-κB/NLRP3 signaling pathway in LPS+ATP treated primary rat microglia
IκB and NF-κB p65 levels were not influenced ( Figure 7A-G). The activated NF-κB p65 can translocate into the nucleus initiating the transcription. Accordingly, we found that nucleus p65 (np65) level was increased after LPS + ATP treatment, while MOOs significantly abrogated the effect ( Figure 7A, H). Confocal microscopy also showed that MOOs treatment can lessen the nuclear translocation of NF-κB p65 in LPS + ATP treated primary rat microglia ( Figure 7I).
In conclusion, MOOs may inhibit the IκB/NF-κB signaling pathway to negatively regulate NLRP3 inflammasome in primary rat microglia.

| MOOs suppress the IκB/NF-κB signaling pathway in PSD rats
To further confirm the effect of MOOs on the tIκB/NF-κB signaling pathway in vivo, we measured the protein expression of p-IκB, IκB, p-p65, p65, and np65 in the ischemic hippocampus of PSD rats.
We found that compared with the control group, the levels of p-IκB, p-IκB/IκB, p-p65, p-p65/p65, and np65 were increased in PSD + v rats. MOOs can reverse the elevation without affecting total IκB and p65 levels ( Figure 8A-H). Further confocal microscopy was also performed to assess the nuclear translocation of p65 in microglia in PSD rats. We found that PSD +v rats showed higher nuclear translocation of p65 in microglia than Ctrl +v rats and MOOs can reduce this translocation ( Figure 8I). Taken together, MOOs may inhibit NLRP3 inflammasome activation of PSD rats by suppressing the IκB/NF-κB signaling pathway.

Terms Counts Genes
In our study, we have further demonstrated the effect of MOOs in alleviating depressive-like behaviors in PSD rats, which may provide evidence for its use in the clinical treatment of PSD patients. Of note, the blood brain barrier is injured with high permeability after cerebral ischemic insult, even up to 1 month after stroke in rats, 38 thus the orally administered MOOs can finally reach the brain to function.
Hippocampal inflammation has been extensively investigated in depression. Inhibition of hippocampal inflammation can alleviate LPS-induced depressive behaviors in mice. 13 Moreover, several clinical trials have illustrated that stroke patients with higher serum IL-1β or IL-18 levels at admission have a higher risk of developing PSD. 15,16 Here, we further demonstrated that inflammatory response was activated in the hippocampus of PSD rats and the microglial NLRP3 inflammasome activation was responsible for the excessive inflammatory response. However, it is interesting to note that the role of NLRP3 inflammasome in ischemic stroke is still controversial. In the study of Denes et al 39 , NLRP3 −/− and WT mice showed no difference on ischemic brain injury after MCAO, while absent in melanoma 2 (AIM2) and NLR family, CARD domain containing 4 (NLRC4) rather than NLRP3 inflammasome may contribute to brain injury. 40,41 This result is in contrast to a number of studies that have demonstrated the detrimental effect of NLRP3 inflammasome in ischemic stroke. 42 Notwithstanding, the inflammation induced by NLRP3 inflammasome is a key factor in the pathogenesis of depression, which has been validated by both clinical and preclinical findings. 43 45 Additionally, the NF-κB signaling pathway is activated in both depression 46 and ischemic situations. 47 In this study, MOOs suppressed the increase of phosphorylated IκB and NF-κB p65 in both PSD rats and primary rat microglia, which could be involved in inhibited NLRP3 inflammasome activation. Notably, the regulation of NLRP3 inflammasome activation is a complex process and many other factors may also be involved, such as trans-Golgi disassembly, lysosomal disruption, and K + efflux. 17 Thus, further investigation is needed to fully clarify the mechanism of MOOs on NLRP3 inflammasome activation inhibition.
One of the limitations of our studies is that the AAV-NLRP3 we designed to regulate hippocampal NLRP3 inflammasome was not specifically targeted to microglia. Other cell types such as astrocytes and neurons may also be transfected with AAV-NLRP3, resulting in decreased reliability of the results. However, consistent with other studies, 18   is interesting to note that sex differences exist in microglial activation and neuroinflammation response after stroke. 54 Sex differences in depressive-like behavior may also be associated with the imbalance of microglia-induced neuroinflammation in the hippocampus. 55 Thus, it is possible that the effect of MOOs on microglia NLRP3 may be different in male and female rats. To fully elucidate this, both male and female PSD rats should be applied in further studies.
In conclusion, this study revealed the antidepressive effect and pharmacological mechanism of MOOs and emphasized NLRP3 inflammasome as the possible therapeutic target of PSD. We further illustrated that MOOs alleviate depressive-like behaviors in PSD rats through the IκB/NF-κB p65/NLRP3 inflammasome signaling pathway, which is an inflammatory pathway and may also offer insights for PSD treatment. Moreover, we proposed a correlation between stroke, hippocampal inflammation, and PSD. We further proposed the viewpoint that hippocampal inflammatory response activation followed by stroke could play a crucial role in the pathogenesis of PSD and that alleviating hippocampal inflammation through downregulating NLRP3 inflammasome is a potential novel target for PSD.

| CON CLUS ION
In this study, we confirmed the antidepressive effect of MOOs in PSD rat models. We also identified that MOOs can suppress microglial NLRP3 inflammasome activation to alleviate hippocampal inflammation. Moreover, we further illustrated that the IκB/NF-κB