Curzerene Ameliorates Depression-Like Behaviors and Cognitive Impairment by Modulating the Gut Microbiota and HMGB1/RAGE/TLR4 Pathway

Abstract

Curzerene, a sesquiterpene compound isolated from Curcuma Radix, exhibits various therapeutic effects, such as anti-tumor and anti-hyperlipidemic properties. However, its neuroprotective effects have not yet been reported. This study focused on exploring the neuroprotective effect of curzerene and elucidating its potential mechanism by combining molecular biotechnology with multi-omics approaches. Curzerene was orally administered to LPS-induced depressive-like behaviors and cognitive impairment in mice for 14 days, and the related biochemical parameters were evaluated. The possible mechanisms were elucidated using qRT-PCR, Western Blot, immunofluorescence, untargeted metabolomics, GC-MS and 16S rDNA comprehensively. Curzerene ameliorated depression symptoms and cognitive impairment by increasing the preference for sucrose in SPT and the central area and total distance traveled in OFT, reducing the immobility time in TST and FST, as well as rising the spontaneous alternation ratio in Y maze. Multiple molecular biology techniques analyses indicated the ameliorative effect of curzerene via HMGB1/RAGE/TLR4 pathway. Moreover, curzerene primarily regulates purine metabolism, phenylalanine, tyrosine and tryptophan biosynthesis, phenylalanine metabolism, pyrimidine metabolism, etc. Furthermore, intervention increased the relative abundance of Parabacteroides, Clostridia_UCG-014_unclassified, and Rhodospirillales_unclassified, and enhanced the production of SCFAs. This work demonstrated that curzerene effectively protects against LPS-induced neurological damage, potentially by inhibiting the HMGB1/RAGE/TLR4 pathway through the restoration of gut microbiota homeostasis, modulation of metabolites, and enhancement of SCFAs. In conclusion, this study offers new perspectives on the therapeutic possibilities of curzerene in mitigating depressive-like behaviors and cognitive impairment.

INTRODUCTION

Depression is a serious, chronic, and frequently recurring psychological disorder characterized by symptoms such as anhedonia, sleep disturbances, persistent low mood, and impaired concentration. In extreme cases, individuals with depression may exhibit suicidal behaviors, including self-harm and suicidal ideation. According to an epidemiological survey released in 2017, the global prevalence of depression was 4.4%, representing a 50% increase over the past three decades (Chekroud et al., 2017). Currently, selective serotonin reuptake inhibitors (SSRIs) and serotonin-norepinephrine reuptake inhibitors (SNRIs) are widely used in the treatment of depression. However, these medications may cause significant adverse effects, such as weight gain, disrupted sleep patterns, dizziness, and headaches (Gill et al., 2020). Moreover, these findings may suggest suboptimal clinical pharmacodynamics and a high risk of recurrence after discontinuation, thereby highlighting the challenges in meeting patients’ clinical needs. (Gerhard et al., 2016; Henssler et al., 2019). As a result, there is an urgent requirement to develop antidepressant medications or therapies that simultaneously reduce side effects and enhance therapeutic efficacy.

The exact pathogenesis of depression in clinical practice remains unclear, and accumulating evidence suggests that gut microbiota disorders may be involved in the onset, progression, and prognosis of the disease (Malan-Muller et al., 2023; Okdeh et al., 2023). Adequate clinical evidence confirms the existence of a relationship between the gut bacteria and brain neurons and characterized by bidirectional regulation of the Microbiota-Gut-Brain (MGB) axis as a critical mechanistic pathway (Chakrabarti et al., 2022). The disordered microbiota could activate micro-associated molecular patterns (MAMPs) and are recognized by host nucleotide-binding oligomerization domain (NOD)-like receptors (NLRs) and toll-like receptors (TLRs) such as TLR4 (Frieri and Stampfl, 2016; Pellegrini et al., 2020). Serving as the entry point to the innate immune system, these receptors initiate the cascade reaction of producing inflammatory cytokines. These cytokines could disrupt the blood-brain barrier (BBB), thereby affecting neuro-inflammation in the central nervous system (CNS) via serotonin, and tryptophan pathways (Herselman et al., 2022). However, the gut microbiota and their beneficial metabolites, such as short-chain fatty acids (SCFAs), which can regulate the intestinal barrier, immune and neuroinflammation by crossing the BBB (Corrêa-Oliveira et al., 2016). Thus, reshaping the composition of gut microbiota and its metabolic byproducts may represent an effective strategy for treating depression.

Traditional Chinese Medicine (TCM) contains various potential antidepressant phytoconstituents with excellent therapeutic effects and minimal side effects, which can be explored as novel therapeutic agents for treating depressive disorders, particularly in patients who exhibit poor response or experience significant side effects from current antidepressants (Wang et al., 2019a). Curcumae Radix, also named wen-yu-jin, originated from the dried radix of Curcuma wenyujin Y. H. Chen et C. Ling. Studies indicate that Curcumae Radix exhibits neuroprotective effects (Chen et al., 2022; Li et al., 2021; Zhao et al., 2011). Additionally, our prior studies have demonstrated that Curcumae Radix possesses anti-neuroinflammatory properties and can alleviate Alzheimer’s disease-related pathologies (Qi et al., 2025). Curzerene, a sesquiterpene compound isolated from Curcuma Radix, exhibits a range of therapeutic effects, including anti-tumor and anti-hyperlipidemic properties (Tahir et al., 2025; Zhou et al., 2025). However, the neuroprotective effect and precise mechanism of curzerene remain unclear. Therefore, this research aimed to assess the therapeutic impact of curzerene on depression and explore its possible mechanisms of action.

MATERIALS AND METHODS

Materials

Curzerene with purity of 98% was sourced from Yuanye Biotechnology Co., Ltd. (Shanghai, China). The antibodies were obtained from Cell Signaling Technology (USA). The lipopolysaccharide (LPS) was purchased from (Sigma Chemical Co., St. Louis, MO, USA). The ELISA kits used in this study were bought from MultiSciences Biotech Co., Ltd., China (Hangzhou, China).

Experimental design for drug treatment

The C57BL/6 mice (aged 6-8 week) were provided with Beijing Weitong Lihua Laboratory Animal Technology Co., LTD. The experimental mice were kept in a controlled temperature (20 ± 2°C) and humidity (40%-60%) environment with a 12-h light/dark rhythm. The project was subjected to supervision and inspection by the Ethics Committee of Wenzhou Medical University and with ethics approval number WYDW 2023-0663.

After one week of adaptation, mice were divided into five groups. (1) The control group (Physiological saline intraperitoneal injection (i.p.) group); (2) The model group (0.83 mg/kg lipopolysaccharide (LPS) i.p. group); (3) The curzerene (15 mg/kg)+LPS group; (4) The curzerene (30 mg/kg)+LPS group; (5) The Fluoxetine (12 mg/kg Flu)+LPS group. Mice were pretreated oral administration (o.p.) with vehicle, curzerene or Flu for 14 consecutive days. One hour after the last curzerene or Flu administration, the mice were treated i.p. with 0.83 mg/kg LPS dissolved in normal saline (Wang et al., 2019b). For each experimental group, fecal samples and tissues were collected from a subset of mice, whereas the remaining mice were subjected to behavioral testing.

Behavioral testing

The sucrose preference test (SPT), forced swim test (FST), Y maze, tail suspension test (TST) and open field test (OFT) were performed as previously described (Li et al., 2021). The detailed procedures were described in Supplementary File 1.

Hematoxylin and eosin (HE) staining

The tissues were fixed with paraffin after dehydration. Then, the embedded tissue was sliced into 3-5 μm thick sections and dewaxed, stained with hematoxylin (3 min) and eosin (3 min). The slices were sealed with neutral gum and observed under a Nikon Eclipse E100 microscope.

Nissl’s staining

The slices were sequentially processed with xylene, anhydrous ethanol, 95% alcohol, 80% alcohol, 70% alcohol, and distilled water, each for two rounds of 5 min. Subsequently, the sections were stained with 0.5% toluidine blue solution for 5 min. The slices were dehydrated by alcohol and anhydrous ethanol, and then removed with xylene. Images were observed using a Nikon Eclipse E100 microscope.

Sample preparation and ELISA assay

The hippocampus was dissected into pieces and then transferred to PBS containing protease inhibitors. Then, the tissues were mechanically lysed for ELISA detection. The IL-6 (EK206), IL-1β (EK201B) and TNF-α (EK282) were quantified according to the protocol of the ELISA manufacturer.

Western blotting analysis

The 20 μg denatured protein sample from hippocampus tissues was loaded to the 8-12% SDS-PAGE gels and separated at 90 V. The protein on the gel was transferred onto the PVDF membrane (Millipore, USA). Next, the membrane was blocked with 7.5% no-fat milk for 1.5 h at 25°C and then incubated with primary antibodies against HMGB1 (#3935, CST), NF-κB p65 (#8242, CST), RAGE (ab30381, Abcam), IKK (ab124957), IκB (ab32518), TLR4 (19811-1-AP, Proteintech), β-actin (20536-1-AP, Proteintech) and GAPDH (10494-1-AP, Proteintech) in a refrigerator at 4°C overnight. Next day, the membrane was incubated with the secondary antibody for 2 h at 25°C. The bands of required internal reference and target were observed using an enhanced chemiluminescence kit. The quantifications of the bands were evaluated by Image J software.

Quantitative real-time PCR analysis

Hippocampal RNA was extracted using the RNeasy Mini Kit (Qiagen, Hilden, Germany), and subsequently, its purity and concentration were determined. The cDNA was synthesized by reverse transcription kit (Shenggong Bioengineering Co., Ltd, Shanghai, China). Gene expression was quantified by qPCR kit (Vazyme Biotech, Nanjing, China) employing SYBR Green (Vazyme Biotech, Nanjing, China). The related gene expression was normalized against GADPH and determined by the 2^-ΔΔCt technique. Primer design and purchase from Beijing Soleibao Co., Ltd. The primer are shown in Table 1.

Table 1.

The primer sequences used for qPCR assay

Gene	Forward primer (5′-3′)	Reverse primer (5′-3′)
HMGB1	CCCAGCGAAGGCTATCACA	TTCATAAAGGGACAAACCACAG
NF-κB p65	TGCACCTGTTCCAAAGAGCA	GGTCTGTGAACACTCCTGGG
RAGE	CCTGAAGGTGGAATAGTCGCT	AGCTATAGGTGCCCTCATCCTC
IKK	GAGGAGATGGCTGTGATCGG	TCCCTCATTAGTTGCGGTGT
IκB	TGCAGGCCACCAACTACAAT	AAGAGCGAAACCAGGTCAGG
TNF-α	TTAGAAAGGGGATTATGGCTCA	TTTGCAGAACTCAGGAATGGAC
IL-1β	GGGCTGGACTGTTTCTAATGC	CTTGTGACCCTGAGCGACC
IL-6	ACAACCACGGCCTTCCCTA	CATTTCCACGATTTCCCAGA
TLR4	CGCTCTGGCATCATCTTCA	TTTTCCATCCAATAGGGCAT
GAPDH	AAGAAGGTGGTGAAGCAGG	GAAGGTGGAAGAGTGGGAGT

Immunofluorescence analysis

The slices were treated with 3% H2O2 for 20 min, followed by blocking of nonspecific binding in a solution of 5% bovine serum albumin (BSA) with 0.3% Triton X-100 for 30 min at 37°C. The washed slices were treated with primary antibodies anti-HMGB1 (1:500) for 12 h at 4°C, followed by incubation with the secondary antibodies at 25°C for 90 min. Lastly, the nuclei were stained with DAPI for 10 min. The slices were then examined and imaged using a microscope (ECLPSE 80i, Nikon, Tokyo, Japan).

Untargeted metabolomics analysis

The untarget metabolomics analysis was detected by a high-resolution tandem mass spectrometer Q-Exactive. The detailed procedures were described in Supplementary File 1.

GC-MS detection of short-chain fatty acids (SCFAs) in intestinal contents

The extraction of SCFAs, Instrumentation and GC-MS conditions, GC-MS method validation, Sample analysis and 16S rDNA analysis were described in Supplementary File 1.

Statistical analysis

The original data was shown as means ± SD. One-way ANOVA was used to evaluate datasets involving two or more groups, followed by Tukey’s multiple comparison test and the student’s t-test was utilized for difference analysis between two groups. The statistical analysis of data was performed using the SPSS 22.0 software. The difference was defined as p<0.05.

RESULTS

Curzerene improved LPS-induced anxiety and depression-like behaviors

LPS is a recognized pro-inflammatory agent commonly used to induce depression-like behaviors in rodent model (Liu et al., 2024; Wang et al., 2020). As described in Fig. 1A, the SPT, FST, TST, OFT, and Y maze were employed to evaluate the effect of curzerene on behavioral phenotypes in mice. In the OFT, mice administered with LPS alone demonstrated a marked decrease in both the time spent in the central zone and the frequency of entries into this area compared to normal mice. Conversely, pretreatment with different concentrations of curzerene for 2 weeks, especially at the concentration of 30 mg/kg significantly improved the time spent in the central area and total distance traveled (Fig. 1B). In the SPT, mice exposed to LPS alone exhibited less preference for sucrose compared to the control group, while pretreatment with curzerene could reversed this trend (Fig. 1C). Although the control mice showed a consistent with the LPS group alone in immobility time, a significant decrease in immobility time was observed pretreat with curzerene at 15 or 30 mg/kg for 2 weeks in the FST (Fig. 1D). As demonstrated in the TST, a markedly increase in the duration of immobility was recorded in the LPS-treated group, which was alleviated by curzerene (Fig. 1E). The Y maze test also demonstrated that pretreat with curzerene for 2 weeks reversed the spontaneous alternation ratio with no significant difference in the number of arm entries (Fig. 1F, Supplementary Fig. 1).

Fig. 1.

Curzerene ameliorate anxiety and depression-like behaviors on behavioral experiments in LPS-induced mice. The experimental design for curzerene treatments and positive control fluoxetine (A). Line crossing and distance in central area in the OFT (B). SPT of the indicated mouse encompassed analyzing the overall preference for sucrose (C). The immobility time were analyzed in FST (D) and TST (E). The spontaneous alternation ratio was analyzed in Y maze test (F). Data were showed as means ± SD (n=6). ^##p<0.01, ^###p<0.001 vs the control group; *p<0.05, **p<0.01, ***p<0.001 vs the model group.

Curzerene attenuated LPS-induced inflammatory cytokines and neural damage in vivo

The levels of inflammatory cytokines in the hippocampus were measured in mice. The results showed that LPS induction significantly increased the levels of IL-6, IL-1β, and TNF-α in the hippocampus (Fig. 2A-2C). These increases were attenuated by treatment with either 15 or 30 mg/kg curzerene or fluoxetine. These findings provide evidence, to some extent, of the beneficial effects of curzerene and fluoxetine on the regulation of inflammatory cytokine production. Further assessment of hippocampal damage revealed that LPS induction caused alterations in cell morphology and arrangement (Fig. 2D), as well as a significant reduction in the number of Nissl-positive cells in hippocampal tissue compared to the control group (Fig. 2E-2H). These pathological changes were ameliorated by treatment with curzerene and fluoxetine. Therefore, curzerene demonstrates the potential to alleviate both LPS-induced inflammation and neuronal damage in LPS-treated mouse models.

Fig. 2.

Curzerene ameliorate inflammatory cytokines and neural damage in the different region of mice. (A-C) IL-1β, IL-6 and TNF-α in hippocampal homogenates. (D, E) HE staining and Nissl images, scale bar=100 μm. (F-H) Quantitative analysis nissl-stained cell per field in CA1, DG, and Cortex. Data were showed as means ± SD (n=6). ^##p<0.01, ^###p<0.001 vs the control group; *p<0.05, **p<0.01, ***p<0.001 vs the model group.

Curzerene attenuated neuroinflammation-related signaling in LPS-treated mice

LPS induced neuroinflammatory response is one of the typical markers for inducing depressive-like behavior in mice (Lian et al., 2017). In present study, we evaluated the expressions of mRNA related to the HMGB1-RAGE inflammatory signaling pathway in the hippocampus (Fig. 3). LPS significantly reduced the mRNA expressions of IL-1β, TNF-α, IL-6, NF-κB p65, IKK, TLR4, RAGE and HMGB1 in the hippocampus compared to those in the control mice, all of which were reversed post treatment with curzerene at the dose of 30 mg/kg for 14 d. On the other hand, LPS treatment down-regulated the mRNA level of IκB, which was subsequently up-regulated by curzerene administration.

Fig. 3.

Curzerene reduce the mRNA expression of IL-1β (A), TNF-α (B), IL-6 (C), NF-κB p65 (D), IKK (E), IκB (F), TLR4 (G), RAGE (H) and HMGB1 (I) in the hippocampus of LPS-treated mice. Data were showed as means ± SD (n=3). ^#p<0.05, ^##p<0.01 and ^###p<0.001 vs the control group; *p<0.05, **p<0.01 and ***p<0.001 vs the model group.

Curzerene attenuated neuroinflammation by regulating the dysregulation of the HMGB1/RAGE/TLR4 signaling pathway in LPS-treated mice

Under LPS stimulation, HMGB1 translocates from the nucleus to the cytoplasm, and is subsequently released into the extracellular space, where it binds to the receptor RAGE, leading to activation of the NF-κB signaling pathway and the subsequent promotion of inflammatory cytokines such as TNF-α and IL-1β. Western blot analysis demonstrated that curzerene reduced the protein expressions of cytoplasm-HMGB1, RAGE, TLR4, and IKK, while up-regulating IκB expression (Fig. 4A-4F). Nuclear NF-κB expression was lowered while nuclear HMGB1 was elevated by curzerene treatment (Fig. 4G-4I). Immunofluorescence results showed that curzerene promoted HMGB1 nuclear translocation (Fig. 5A, 5B). Collectively, the neuroprotective effects of curzerene may be mediated by curtailing the HMGB1/RAGE/TLR4 signaling pathway.

Fig. 4.

Effects of curzerene on neuroinflammatory response and HMGB1/RAGE/TLR4 signaling pathways activation in LPS-induced mice. (A) Western blots of HMGB1, RAGE, TLR4, IKK and IκB in the hippocampus. (B-F) Relative expression of HMGB1, RAGE, TLR4, IKK and IκB protein. (G) Western blots of nuclear HMGB1 and NF-κB in the hippocampus. (H-I) Relative expression of nuclear HMGB1 and NF-κB protein. Data were showed as means ± SD (n=3). ^#p<0.05, ^###p<0.001 vs the control group; *p<0.05, **p<0.01, ***p<0.001 vs the model group.

Fig. 5.

Curzerene promotes HMGB1 nuclear translocation. (A, B) Immunofluorescence of nuclear HMGB1 in the hippocampus. Data were showed as means ± SD (n=3). ^#p<0.05 vs the control group; *p<0.05 vs the model group.

Metabolomics analysis

A serum-based untargeted metabolomics approach was developed to assess the metabolic variations between the control group and the group exhibiting LPS-induced depression-like behavior. This study aims to investigate biochemical changes associated with metabolites in the depression-like state and to clarify the possible therapeutic effects of curzerene. The UPLC-QE-Orbitrap-MS approach was employed to identify specific metabolites. A model containing total metabolic information was constructed using Simca-P14.0 software and the PCA scatter plots of each group were shown in Fig. 6A and 6B.

Fig. 6.

Multivariate statistical analysis of the serum samples (n=6) intervened by curzerene in positive ion and negative ion. (A) PCA score plot of all the experimental groups in positive ion; (B) PCA score plot of all the experimental groups in negative ion; (C) OPLS-DA score plot of all the experimental groups in positive ion; (D) OPLS-DA score plot of all the experimental groups in negative ion; (E) Heatmap of metabolites with significant changes; (F) Bubbles map of metabolism pathway of the potential pharmacodynamic serum biomarker of curzerene in intervention of depression mice.

To investigate the metabolic changes in LPS-induced depression model treated with curzerene intervention, supervised OPLS-DA was used to search for endogenous foreign bodies between groups (Fig. 6C, 6D). The OPLS-DA scores of the control group and model group, S-plot diagrams and 200 permutations test, as well as the model group and curzerene group, S-plot diagrams and 200 permutations test were shown in Supplementary Fig. 2 and Supplementary Fig. 3. The result shown there are had a good fitness and prediction. Furthermore, the deviation of six QC samples was within ± 3 std, which showed that the built method was reliable and stable (Supplementary Fig. 4). The potential serum metabolites with marker contribution were screened out on VIP values >1 and p values <0.05. Subsequently, 93 serum differential metabolites, including 4-Pyridoxic acid, acetic acid, deoxyinosine and others (Supplementary Table 3) were identified based on their MS analysis results, accurate molecular weight, and matching with HMDB database in control and model group. 234 serum differential metabolites, including glyceric acid, capric acid, cholesterol and others (Supplementary Table 4) were identified based on their MS analysis results, accurate molecular weight, and matching with HMDB database in model and curzerene group. Finally, 31 serum differential metabolites with changes had been screened out (Table 2). Among these, 25 serum metabolites levels were significantly upregulated, while 6 serum metabolites levels were significantly decreased after curzerene intervene for 14 d. The visualized heatmap of the relative content of 31 metabolic foreign substances was shown in Fig. 6E. The metabolic pathway enrichment analysis of 31 serum biomarkers were preformed using MetaboAnalyst 6.0 software. The results showed that curzerene improve depression mice mainly involving Purine metabolism, Phenylalanine, tyrosine and tryptophan biosynthesis, Phenylalanine metabolism, Pyrimidine metabolism, Histidine metabolism, Pantothenate and CoA biosynthesis, β-Alanine metabolism, etc (Fig. 6F).

Table 2.

Identified differential metabolites in serum among each group

Name	Formula	MZ	RT (min)	Scan mode	P(C/M)	VIP (C/M)	P(CR/M)	VIP (Cur/M)	Trend of C/M	Trend of Cur/M
5-Aminopentanoic acid	C5H11NO2	116.0707	0.73	-	0.0132	1.09	0.0232	5.67	↓	↑
L-Threonine	C4H9NO3	118.0499	0.70	-	0.0149	1.49	0.0373	2.03	↓	↑
L-Aspartic acid	C4H7NO4	132.0293	0.64	-	3.2415	0.01	0.0242	2.04	↑	↓
D-Malic acid	C4H6O5	133.0134	0.64	-	0.0494	1.97	0.0330	3.35	↑	↓
Xanthine	C5H4N4O2	151.0253	0.74	-	0.0002	4.42	0.0176	3.64	↓	↑
Protocatechuic acid	C7H6O4	153.0185	1.30	-	0.0083	1.36	0.0221	1.32	↑	↓
Uric acid	C5H4N4O3	167.0196	0.73	-	0.0353	1.32	0.0187	1.84	↓	↑
Capric acid	C10H20O2	171.1382	6.23	-	0.0366	1.14	0.0190	1.10	↓	↑
Uridine	C9H12N2O6	243.0620	1.02	-	0.0182	1.63	0.0035	3.01	↓	↑
Inosine	C10H12N4O5	267.0734	1.03	-	0.0384	1.77	0.0055	2.29	↓	↑
Guanosine	C10H13N5O5	282.0843	1.03	-	1.2492	0.01	0.0068	1.38	↓	↑
Arachidonic acid	C20H32O2	303.2328	7.70	-	0.0022	18.66	0.0259	17.24	↓	↑
Adrenic acid	C22H36O2	331.2641	8.25	-	0.0019	6.03	0.0140	6.58	↓	↑
Thymine	C5H6N2O2	127.0504	1.15	+	0.0018	1.87	0.0197	2.63	↓	↑
L-5-Oxoproline	C5H7NO3	130.0500	0.64	+	0.0019	1.12	0.0008	3.38	↑	↓
Adenine	C5H5N5	136.0756	0.74	+	0.0010	2.22	0.0028	2.83	↓	↑
Hypoxanthine	C5H4N4O	137.0458	1.02	+	0.0034	8.78	0.0447	3.86	↓	↑
Methionine	C5H11NO2S	150.0571	1.02	+	0.0037	1.36	0.0018	4.73	↓	↑
Phenylalanine	C9H11NO2	166.0862	1.19	+	0.0021	4.51	0.0099	17.52	↓	↑
Tyrosine	C9H11NO3	182.0811	1.02	+	0.0006	1.79	0.0096	4.37	↓	↑
4-Pyridoxic acid	C8H9NO4	184.0606	1.07	+	0.0014	1.76	0.0033	1.73	↑	↓
Tryptophan	C11H12N2O2	205.0972	2.13	+	0.0016	6.05	0.0411	6.14	↓	↑
Thiamine	C12H17N4OS	265.1120	1.03	+	0.0020	3.77	0.0036	5.23	↓	↑
Uracil	C4H4N2O2	111.0190	1.02	-	0.0006	3.84	0.0108	1.77	↓	↑
4-Imidazoleacetic acid	C5H6N2O2	125.0348	1.04	-	0.0003	3.43	0.0490	1.46	↓	↑
Valyl-Leucine	C11H22N2O3	229.1550	1.07	-	0.0219	1.68	0.0057	2.12	↓	↑
Docosahexaenoic acid	C22H32O2	327.2328	7.54	-	0.0216	10.64	0.0269	12.54	↓	↑
Propionic acid	C3H6O2	73.0280	1.29	-	0.0425	1.61	0.0138	1.51	↓	↑
PE(18:2(9Z,12Z)/18:0)	C41H78NO8P	742.5388	7.99	-	0.0412	1.53	0.0008	1.00	↑	↓
Toluene	C7H8	91.0542	4.63	-	0.0203	1.35	0.0044	2.35	↓	↑
1-Stearoyl-2-linoleoyl-sn-glycero-3-phosphocholine	C44H84NO8P	786.5992	9.67	+	0.0011	1.15	0.0013	1.23	↓	↑

Curzerene increased SCFAs Levels in LPS-induced mice

SCFAs, a class of metabolites with distinct characteristics of gut microbiota, are closely associated with inflammation, neuronal damage, anxiety and depression-like behaviors. As depicted in Fig. 7A-7F, relative to the control mice, the treatment with LPS mice displayed markedly decreases in fecal concentrations of acetic acid, butyric acid, propionic acid, valeric acid, isobutyric acid, isovaleric acid and valeric acid by 56.06%, 53.48%, 54.59%, 50.64%, 63.85%, and 75.19%, respectively. In contrast, treatment with curzerene at the dose of 30 mg/kg significantly increased the levels of acetic acid, butyric acid, propionic acid, valeric acid, isobutyric acid, isovaleric acid and valeric acid by 72.77%, 107.8%, 80.71%, 94.21%, 190.0% and 191.2%, respectively.

Fig. 7.

Curzerene increased SCFAs level in LPS-induced mice. (A) Fecal levels of acetic acid; (B) Fecal levels of butyric acid; (C) Fecal levels of propionic acid; (D) Fecal levels of isobutyric acid; (E) Fecal levels of isovaleric acid; (F) Fecal levels of valeric acid. Data are presented as the means ± SD; n=6 mice per group. ^##p<0.01and ^###p<0.001 vs the control group; *p<0.05, **p<0.01 and ***p<0.001 vs the model group.

Curzerene reshaped gut microbiota structure of LPS-induced mice

Growing evidence shows that disordered gut microbiota structure is a critical factor in the occurrence and development of depression. Therefore, we further studied the effects after curzerene intervention on the gut microbiota of LPS-induced mice. A significant separation was observed between the gut microbiota of the three groups in β-Diversity analysis measured by Bray-Curtis distance (Fig. 8A). Although α-Diversity analysis showed a decreasing trend in LPS induced mice, the difference was not statistically significant. Curzerene could increase the α-Diversity compare with the LPS-induce mice (Fig. 8B). Gut microbiota structural analysis revealed twenty-one prominent phyla and 41 dominant genera from the three groups (Fig. 8C, 8D). The data showed that treatment with LPS and 30 mg/kg curzerene intervention remarkably reshaped the gut microbiota composition in mice. linear discriminant analysis effect size (LDFSe) displayed that o_Enterobacterales and f_Enterobacteriaceae were the dominant taxa in treatment with LPS, while s_Parabacteroides_sp, o_Rhodospirillales, p_Proteobacteria and c_Alphaproteobacteria were the predominant taxa in curzerene -fed mice (Fig. 8E). Wilcoxon analysis shown that 16 abundant and statistically different bacterial groups between LPS and 30 mg/kg curzerene -treated in genus and species level, and (Fig. 8F, 8G). In summary, ecological disturbance of gut microbiota was reshaped by treatment with curzerene could potentially contribute to protect mice from LPS-induced depressive behaviors.

Fig. 8.

Curzerene reshaped gut microbiota structure in LPS-induced mice. (A) Principal component analysis (PCA) based on Bray-Curtis distances. (B) Shannon index, stand for the richness and the diversity of each group. (C) Relative abundance at the phylum level. (D) Relative abundance at the genus level. (E) Linear discriminative analysis (LDA) score of differentially enriched bacterial genera obtained from LEfSe analysis in each group. (F) Wilecoxon analysis at the genus level. (G) Wilcoxon analysis at the species level. Data are presented as the means ± SD; n=6 mice per group.

DISCUSSION

This experiment investigated the therapeutic effects of curzerene on LPS-induced depression-like behavior in mice. The results of behavioral evaluations demonstrated that curzerene significantly ameliorated depression-like behaviors in LPS-treated (i.p.) mice across multiple tests, including the SPT, FST, TST, OFT, and Y-maze. Pathological analyses revealed that curzerene mitigated abnormal tissue organization. Furthermore, 16S rDNA sequencing, untargeted metabolomics analysis, and rescue experiments indicated that gut microbiota dysregulation and its associated metabolites played a critical regulatory role. Specifically, reshaping the gut microbiota structure enhanced the antidepressant activity and anti-inflammatory effects of curzerene.

Depression is widely recognized as a global public health issue due to its high prevalence and associated mortality rates. The LPS-induced depression-like behavior model is a well-established paradigm for studying depression, as previous studies have demonstrated that it causes significant hippocampal damage and structural disorganization in rats (Wang et al., 2020). In this research, curzerene treatment increased sucrose preference in the SPT and enhanced locomotor activity, as evidenced by increased distance traveled in the OFT, in LPS-induced mice. Additionally, curzerene reduced immobility time in both the TST and FST. Furthermore, the Y-maze test revealed that two weeks of curzerene intervention restored the spontaneous alternation ratio without significantly altering the number of arm entries, indicating an alleviation of depression-like behaviors in rats. These results align with those reported by Cao et al. (2022).

In both rodent models and human depression cases, activated microglia within the hippocampus have been observed, accompanied by excessive production of proinflammatory cytokines by these cells. These proinflammatory cytokines further initiate downstream cascade reactions, leading to synaptic dysfunction and ultimately contributing to depressive symptoms. (Ye et al., 2023). Our findings demonstrate that treatment with curzerene significantly reduced the levels of IL-6, IL-1β, and TNF-α in LPS-treated mice. HMGB1, released from microglia and neurons in the hippocampus, interacts with receptors to trigger neuroinflammatory responses, which are closely implicated in the emergence of behaviors resembling depression and anxiety in animal models (Cheng et al., 2016; Wu et al., 2015). Under LPS stimulation, HMGB1 translocates from the nucleus to the cytoplasm, and is subsequently released into the extracellular space, where it binds to the receptor RAGE, leading to activation of the NF-κB signaling pathway and the subsequent promotion of inflammatory cytokines such as TNF-α and IL-1β. Caryophyllene, similar to curzerene, is derived from Curcumae Radix and has been shown to exert beneficial effects in the management of ischemic stroke and neuroinflammation by regulating the HMGB1 signaling pathway in mice model (Wang et al., 2025). Similarly, in our study, curzerene significantly attenuated neuroinflammation and inhibited the HMGB1/RAGE/TLR4 signaling pathway in LPS-treated mice, thereby alleviating depression-like and anxiety-related behaviors. This effect may be at least partially attributed to the following factors, alterations in the gut microbiota and subsequent specific metabolites, as well as SCFAs production. The imbalance of metabolites leads to increased serotonin depletion, promoting the translocation of luminal molecules, toxins, and thereby causing neuroinflammation and neuronal damage. Reduced levels of 5-aminopentanoic acid, arachidonic acid, hypoxanthine and tryptophan are positively correlated with the risk of depression. The key metabolic pathways involved were phenylalanine, tyrosine and tryptophan biosynthesis, purine metabolism and phenylalanine metabolism. Tryptophan, a tyrosine metabolite, exhibits natural anti-inflammatory properties by inhibiting the activity of TLR4/NF-κB signaling (Marx et al., 2021). In this study, curzerene significantly reversed the decreasing trend in tryptophan levels, indicating enhanced anti-inflammatory activity in the hippocampus.

In addition to specific metabolites, SCFAs, a type of small molecule, which mainly produced by gut microbiota, plays key roles in regulating brain function and behavior. SCFAs supplementation improves the neuroprotective effects through astrocyte-neuron glutamate-glutamine shuttle (Sun et al., 2023). It has also been reported that SCFAs alleviate anxiety- and depression-like behaviors in rat model of refractory depression by modulating the level of neurotransmitter, tryptophan metabolism, inflammation and gut microbial homeostasis (Palepu et al., 2024). Furthermore, SCFAs supplementation repairs the colonic inflammatory damage, promotes intestinal homeostasis, and mitigates anxiety- and depression-like behaviors in methamphetamine-induced mice (Zhang et al., 2023). In this study, we found that curzerene administration markedly enhanced the production of SCFAs in LPS-induced mice, potentially linking to the alleviation of neuro-inflammation.

The disordered gut microbiota has been confirmed to deeply involvement in the occurrence and pathogenesis of depression-like behaviors. It has been reported that LPS-induced decreased microbial diversity and an increase in harmful bacteria such as NK4A214_group, Odoribacter, and Ileibacterium, which have been proved to participate in gut damage, systemic infections, inflammatory responses and depression. (Matsuzaki et al., 2024; Wang et al., 2017; Zheng et al., 2024). A recent investigation suggested that gut dysbiosis contributes to the emergence of anxiety- and depression-like behaviors via abnormal synapse pruning mediated by microglia through complement C3 in LPS and CUMS-induced mice (Hao et al., 2024). Furthermore, LPS administration triggers systemic inflammation, alters the composition of gut microbiota, and causes spleen nerve denervation in mice exhibiting depression-like phenotypes (Ma et al., 2022). Here, curzerene intervention significantly increased gut microbiota diversity in LPS-treated mice and altered the structure and composition of the microbial community. It was reported that Parabacteroides, Clostridia_UCG-014_unclassified, and Rhodospirillales_unclassified were associated with neuroinflammation, SCFAs production, and gut health (Guo et al., 2021; Zhu et al., 2023). Consistently, in present study, curzerene intervention distinctly increased the relative abundance of Parabacteroides, Clostridia_UCG-014_unclassified, and Rhodospirillales_unclassified. Furthermore, Akkermansia muciniphila plays a significant role in reducing inflammation through the regulation of immune responses and maintaining gut barrier function (Cani et al., 2022). Similarly, curzerene intervention could reverse the decrease of Akkermansia muciniphila in LPS-induced mice.

In summary (Fig. 9), this study identified the content of curzerene in Curcumae Radix and evaluated its impact on LPS-induced acute depression-like behaviors and cognitive impairment. This work demonstrated that curzerene effectively protects against LPS-induced depression-like behaviors and cognitive impairment, possibly by inhibiting neuroinflammatory responses in the brain through improving gut microbiota homeostasis, regulating metabolites, and increasing SCFAs. Collectively, these results highlight the significant role of the microbiota-metabolites-brain axis in the treatment of depression. Our findings provide robust evidence to support the application of curzerene intervention for managing and preventing depression.

Fig. 9.

The summary of curzerene exerts neuroprotective effect against depression.