Sensitization of GSH synthesis by curcumin curtails acrolein-induced alveolar epithelial apoptosis via Keap1 cysteine conjugation: A randomized controlled trial and experimental animal model of pneumonitis

Graphical abstract

Highlights

• •Curcumin improves systemic inflammation in subjects residing in an air-polluted area.
• •Curcumin mitigates acrolein-induced pneumonitis in a preclinical animal model.
• •Curcumin inhibits acrolein-induced alveolar epithelial apoptosis in a ROS-dependent manner.
• •Curcumin increases redox sensing of Nrf2 and thus sensitizes de novo GSH synthesis.
• •Curcumin-Keap1 cysteine conjugation enhances redox sensing of Nrf2.

Abstract

Introduction

Epidemiological studies have reported an association between exposures to ambient air pollution and respiratory diseases, including chronic obstructive pulmonary disease (COPD). Pneumonitis is a critical driving factor of COPD and exposure to air pollutants (e.g., acrolein) is associated with increased incidence of pneumonitis.

Objectives

Currently available anti-inflammatory therapies provide little benefit against respiratory diseases. To this end, we investigated the preventive role of curcumin against air pollutant-associated pneumonitis and its underlying mechanism.

Methods

A total of 40 subjects was recruited from Chengdu, China which is among the top three cities in terms of respiratory mortality related to air pollution. The participants were randomly provided either placebo or curcumin supplements for 2 weeks and blood samples were collected at the baseline and at the end of the intervention to monitor systemic markers. In our follow up mechanistic study, C57BL/6 mice (n = 40) were randomly allocated into 4 groups: Control group (saline + no acrolein), Curcumin only group (curcumin + no acrolein), Acrolein only group (saline + acrolein), and Acrolein + Curcumin group (curcumin + acrolein). Curcumin was orally administered at 100 mg/kg body weight once a day for 10 days, and then the mice were subjected to nasal instillation of acrolein (5 mg/kg body weight). Twelve hours after single acrolein exposure, all mice were euthanized.

Results

Curcumin supplementation, with no noticeable adverse responses, reduced circulating pro-inflammatory cytokines in association with clinical pneumonitis as positive predictive while improving those of anti-inflammatory cytokines. In the pre-clinical study, curcumin reduced pneumonitis manifestations by suppression of intrinsic and extrinsic apoptotic signaling, which is attributed to enhanced redox sensing of Nrf2 and thus sensitized synthesis and restoration of GSH, at least in part, through curcumin-Keap1 conjugation.

Conclusions

Our study collectively suggests that curcumin could provide an effective preventive measure against air pollutant-enhanced pneumonitis and thus COPD.

Introduction

Chronic exposure to ambient air pollution, including environmental and household pollution, represents a considerable threat to public health and is particularly serious in low- and middle-income countries [1]. Specifically, chronic obstructive pulmonary disease (COPD) is mediated by inflammatory responses following inhalation of cigarette smoke or other noxious external particles such as air pollution and biomass fuel [2]. The highly-cited, influential Global Burden of Disease (also known as GBD) studies estimated that COPD is the third leading cause of death worldwide, causing 3.23 million deaths, 43% of which are attributed to elevated levels of air pollutants; China presents a high COPD prevalence leading to more than 1.8 million deaths annually [3].

Multiple studies have demonstrated that pneumonitis is a key driver of COPD pathogenesis as well as a consistent feature in the progression and exacerbation of COPD [2], [4]. The hallmark of pneumonitis is enhanced pulmonary inflammatory responses (e.g., increased neutrophils and macrophages) due to, at least in part, inhaled particles and gases, usually from environmental exposures (e.g., second-hand cigarette smoke) [5]. Pneumonitis is also characterized by systemic inflammatory responses: activation and mobilization of inflammatory cells into the systemic circulation, and production of acute-phase proteins along with circulating inflammatory mediators (i.e., cytokines and chemokines) [6], [7]. Despite the inflammatory nature of COPD, current anti-inflammatory therapies, such as Roflumilast, an orally active phosphodiesterase 4 inhibitor, provide little therapeutic benefits in COPD patients and may present detrimental effects [2], [8]. As a consequence, there is an urgent need to discover effective and safe preventive measures that target the root of pathogenic development thereby an effective intervention of COPD [9].

Among air pollutants, acrolein is a highly reactive aldehyde, present in cigarette smoke and mobile exhaust, and which has been recognized as causing respiratory diseases including COPD [10]. In the detailed mechanism, acrolein not only damages macromolecules such as nucleic acids, lipids, and proteins by forming adducts but also causes glutathione (γ-glutamyl-cysteinyl-glycine; GSH) depletion. GSH, an intracellular antioxidant, is a cellular guard against reactive oxygen species (ROS) as a substrate in the antioxidant enzyme system. Acrolein conjugates the thiol group of GSH, resulting in capitulated antioxidant property of GSH, referred to as GSH depletion [11]. In context, restoration of the GSH pool is a critical point for pneumonitis. Indeed, the administration of N-acetyl cysteine, a precursor of GSH, has been shown to effectively prevent the pathogenesis of COPD by suppressing lung inflammation [12]. Related, transactivation of nuclear factor erythroid-2-related factor 2 (Nrf2)-Kelch-like ECH-associated protein 1 (Keap1) signaling protected cells against acrolein-induced pulmonary cell death by restoring GSH level [13], collectively suggesting a critical role of GSH in pulmonary inflammatory responses and thus, pneumonitis.

Phytochemicals are secondary metabolites produced by plants, hence found in various foods including fruits and vegetables. Due to their therapeutic potentials in various disease models, there are successful cases where plant-based drugs were developed (e.g., Paclitaxel) [14]. Curcumin, a linear diarylheptanoid, is most abundant in turmeric and has a long history of therapeutic uses for various diseases [15]. Curcumin is one of the phytochemicals capable of transactivating the Nrf2-Keap1 signaling pathway and thus is protective in disease models related to oxidative stress; at the molecular level, it was reported that curcumin acts as an electrophile to modify cysteine residues of Keap1 protein [16]. Hitherto, no study has reported the exact underlying mechanism by which curcumin provides preventive potential against acrolein-induced respiratory dysfunction including in pneumonitis. To this end, we hypothesized that curcumin will alleviate pneumonitis markers through transactivation of Nrf2 and restoring the depleted GSH pool. In order to test our hypothesis, we first conducted a randomized controlled trial (RCT) using a curcumin supplement with human subjects who lived in Chengdu, an urban area in China with poor air quality than Beijing [17]. In a preclinical study, we used an acrolein-induced pneumonitis mouse model to verify the protective effect of curcumin against pneumonitis phenotypes and validate the hypothesized mechanism of action. In addition, molecular docking analysis was conducted to demonstrate the direct physical binding of curcumin to Keap1.

Materials and methods

Ethics statement

All studies involving human participants were approved by the Ethical Committee of Foshan University, Foshan, China (IRB Approval number: Fosu2019012). All experiments involving animals were conducted according to the ethical policies and procedures approved by the Animal Care and Use Committee of Korea University (Approval number: KUIACUC-2019–0060). All animal care and experimental procedures in this work were by Guide for the Care and Use of Laboratory Animals (National Institutes of Health).

Study population

Healthy adults who had lived in Chengdu, China for more than 1 year were recruited for the study. We first assessed inclusion and exclusion criteria through an extensive review of the participants’ questionnaires about their health status and other information. Once qualified and enrolled, participants engaged in an initial 30-min meeting, which included a discussion with the project staff member. Inclusion criteria were as follows: (1) resident of Chengdu, China for longer than 1 year, (2) 18–40 years old both male and female adults, (3) non-smoker (or smoking-cessation for at least 1 year; smoking includes traditional or e-cigarettes, and chewing tobacco), (4) willingness to avoid using over-the-counter or prescription drugs for the entirety of study period, (5) willingness to avoid using any Chinese medicine for the entirety of study period, (6) non-alcohol drinker for the entirety of study period, (7) willingness to avoid eating ginger, chili and curry (including curry-products) for the entirety of study period, (8) willingness to eat controlled meals and take intervention capsules (curcumin or placebo samples) on time every day, and cooperate with sampling every week during the study. Exclusion criteria include (1) previous disease conditions including lung (e.g., asthma or COPD), liver (e.g., non-alcoholic liver disease), kidney (e.g., dialysis) or intestinal diseases (e.g., Crohn’s disease), (2) allergies to cumin, ginger, and/or curry, (3) plan to move out of Chengdu area during the study period, (4) smoking-cessation for<1 year, (5) history of cancer therapy/treatment, (6) record of lung transplantation, (7) diagnosis of hypercapnia, (8) any fatal disease conditions, and (9) any health conditions that may affect the results of the study. Reasons for withdrawal from the study included: (1) concurrent diseases or a serious medical risk to continue participating in the study, (2) non-compliance with the study protocol, (3) use of antibiotics or probiotics, and (4) pregnancy.

Study design and sample collection

A total of 40 participants who met the criteria were enrolled in this study. The subjects were randomly divided into 2 groups (stratified by gender and BMI), and then the subjects were provided either placebo or curcumin supplements for 2 weeks. The curcumin supplement was purchased from Integrative Therapeutics (Theracurmin HP; Green Bay, WI, USA). Visually identical placebo samples were produced at a local pharmacy using inactive ingredients of the curcumin supplement (e.g., cellulose, maltose, and dextran). Each subject took 2 capsules/day (i.e., 1,200 mg curcumin/day or placebo; per os) as instructed on the label. During the study, the subjects were provided with controlled lunch and dinner (detailed information is provided in Supplementary Table 1), and they were asked to record any foods consumed other than the provided meals using a food questionnaire. Blood samples were collected at two-time points throughout the intervention: 1 week after consuming the basal diet (i.e., baseline) and at the end of the intervention. Six mL of fasting venous blood was collected at 8:00 am and, stored at −80 °C for further analyses.

Table 1.

Baseline characteristics of the subjects and immune-related markers in plasma of subjects^1.

Characteristics	All (n = 40)	CON (n = 20)	CUR (n = 20)	p-value2
Sex
Male	22	12	10	–
Female	18	8	10	–
Age, y	21.90 ± 2.21	22.35 ± 2.01	21.75 ± 2.31	0.508
Weight, kg	60.64 ± 11.61	60.50 ± 13.44	56.38 ± 9.20	0.671
Height, m	1.63 ± 0.27	1.67 ± 0.08	1.66 ± 0.09	0.387
BMI, kg/m2	21.58 ± 2.34	22.10 ± 2.40	20.75 ± 1.86	0.192
WBC, 109/L	5.76 ± 0.18	6.15 ± 0.27	5.34 ± 0.20	0.024
Neutrophils, 109/L	3.06 ± 0.13	3.32 ± 0.18	2.77 ± 0.16	0.026
Lymphocytes, 109/L	2.20 ± 0.09	2.30 ± 0.15	2.08 ± 0.10	0.236
Monocytes, 109/L	0.32 ± 0.01	0.33 ± 0.02	0.31 ± 0.02	0.428
Eosinophils, 109/L	0.16 ± 0.02	0.16 ± 0.03	0.16 ± 0.02	0.931
Basophils, 109/L	0.02 ± 0.002	0.02 ± 0.002	0.03 ± 0.002	0.052
RBC, 1012/L	4.80 ± 0.07	4.94 ± 0.11	4.64 ± 0.09	0.040
Hemoglobin, g/L	140 ± 2.5	143 ± 3.4	136 ± 3.6	0.137
Hematocrit, %	44.9 ± 0.72	45.9 ± 1.00	43.8 ± 0.98	0.155
MCV, fL	93.83 ± 0.94	93.2 ± 1.5	94.6 ± 1.1	0.446
MCH, pg	29.20 ± 0.36	29.13 ± 0.59	29.26 ± 0.41	0.858
MCHC, g/L	310.83 ± 1.40	312.29 ± 1.87	309.21 ± 2.04	0.279
RDW-CV, %	12.1 ± 0.1	11.8 ± 0.12	12.4 ± 0.24	0.058
RDW-SD, fL	47.59 ± 0.57	46.16 ± 0.53	49.17 ± 0.90	0.008
PLT, 109/L	218.83 ± 8.37	228.43 ± 12.34	208.21 ± 10.71	0.229
MPV, fL	9.45 ± 0.19	9.40 ± 0.17	9.50 ± 0.35	0.811
PDW	16.15 ± 0.05	16.1 ± 0.05	16.2 ± 0.08	0.715
PCT, %	0.20 ± 0.01	0.21 ± 0.01	0.2 ± 0.01	0.115

MCH, mean corpuscular hemoglobin (average amount of hemoglobin); MCHC, MCH concentration (average concentration of hemoglobin in red blood cells); MCV, mean corpuscular volume (average size of red blood cells); PLT, platelet count; MPV, mean platelet volume; PDW, platelet distribution width; PCT, procalcitonin; RBC, red blood cells; RDW-CV, red cell distribution width-coefficient of variation; RDW-SD, RDW-standard deviation, WBC, white blood cells.

¹Values are mean ± standard deviation.

²p-values indicate statistical difference between CON and CUR.

Clinical measurements

An automated hematology analyzer (BC5100; Mindray, Shenzhen, China) was used to conduct routine blood examinations; the analyses were done at the Huimin Hospital of Chengdu. An electrochemiluminometer (Meso QuickPlex SQ 120; Meso Scale Discovery, Rockville, MD, USA) was used to detect plasma inflammatory cytokines using V-PLEX Proinflammatory Panel 1 Human Kit (K15049D-1; Meso Scale Discovery) according to manufacturer’s instruction.

Preparation of curcumin emulsion

Curcumin powder was purchased from Sigma-Aldrich (St. Lois, MO, USA). Curcumin emulsion was prepared according to previous reports with sight modifications [18]. Briefly, curcumin powder was dissolved in 1% tween-80 in normal saline and mixed under ultrasonication. While sonication, 70 °C of heat was applied for 30 min, followed by immediate cooling in ice, which leads to phase inversion.

Animals and study design

An animal study was conducted according to the institutional guidelines for the care and use of laboratory animals and permission was granted by Korea University (IACUC approval number: KUIACUC-2019–0060). Animals used in this study were 8-week old male C57BL/6 mice obtained from Samtako Inc. (Gyeonggi, Republic of Korea). Upon delivery, the mice were acclimated in conventional polycarbonate mice cages with corncob bedding and adapted to a pelleted purified diet (AIN-93G) for the first 5 days. The animals had free access to food and water and were maintained under conditions of a consistent 22 °C temperature and 12 hr light/dark cycles. After acclimation, mice were randomly allocated into 4 groups: Control group [CON group (n = 10); 1% tween-80 in normal saline (hereafter 1% tween-80) + no acrolein (ACR)], Curcumin only group [CUR group (n = 10); CUR in 1% tween-80 + no ACR], Acrolein group [ACR group (n = 10); 1% tween-80 + ACR], and Acrolein + Curcumin group [ACR + CUR group (n = 10); CUR in 1% tween-80 and ACR]. Curcumin was orally administered to the mice at 100 mg/kg body weight once a day for 10 days, and then the mice were subjected to acute nasal instillation of acrolein (5 mg/kg body weight); the exact volume of acrolein solution was intranasally delivered. Twelve hours after single acrolein exposure, all mice were euthanized through isoflurane inhalation (3–5%), followed by cardiac puncture. After, tissues were harvested and then either snap-frozen in liquid nitrogen or submerged in 4% paraformaldehyde for further analyses.

Plasma inflammatory cytokines analysis

Plasma inflammatory cytokines [i.e., interleukin (IL)-1β, IL-6, and tumor necrosis factor (TNF)-α] were quantified using Mouse IL-1β/IL-6/ TNF-α Quantikine ELISA Kits (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions. Absorbances were measured at 490 nm using a microplate reader (Thermo Fisher Scientific, Waltham, MA, USA).

Morphological assessments

For morphological assessments, 5 μm paraffin lung sections were stained sequentially with hematoxylin and eosin, and Masson’s trichrome. First, to determine air space enlargement in the lungs [19], air space areas (alveoli diameter) were measured and quantified using the ImageJ software (National Institute for Health; Bethesda, MD, USA). Afterward, lung injury was graded from 0 (normal) to 4 (severe) in four categories: interstitial inflammation, neutrophil infiltration, congestion, and edema [20]. The degree of lung injury was calculated by adding the individual scores for each category and all histology analyses were performed by three pathologists in a blinded manner.

Immunoblot analysis

Total protein extracts were separated using sodium dodecyl sulfate–polyacrylamide gel electrophoresis (8–15% depending upon protein molecular weights), and then proteins were transferred onto nitrocellulose membranes, followed by blocking the membrane with 3% bovine serum albumin in tris-buffered saline with 0.05% Tween-20. Once probed with appropriate primary antibodies (Supplementary Table 2), the proteins were visualized by using horseradish peroxidase-labeled secondary antibodies and an enhanced chemiluminescence detection kit (Amersham Pharmacia Biotech, Buckinghamshire, UK). Antibody-specific bands were detected using a chemiluminometer (ImageQuant^TM LAS4000, GE Healthcare, Chicago, IL, USA). Quantification analyses of protein bands were performed using ImageJ software.

Immunofluorescence

Lungs were fixed with 4% paraformaldehyde solution overnight, embedded in paraffin, and processed for antibody probing. After deparaffinization, the lung sections were rehydrated by washing with a gradual decrease in ethanol and rinsed in deionized water. Antigens were retrieved by heating the slides in sodium citrate buffer for 10 min in a microwave oven followed by cooling for 20 min. Slides were rinsed in 50 mM Tris-HCl with 150 mM NaCl at pH 7.5 and blocked with 3% BSA in tris-buffered saline with 0.05% Tween 20. Immunofluorescence analyses were performed using secondary antibodies conjugated with fluorophores (e.g., Alexa Fluor 488; Invitrogen, Carlsbad, CA, USA) and analyzed using a Zeiss Axiovert 200 inverted microscope (Oberkochen, Germany). The average fluorescence intensity was calculated with ImageJ software.

Terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling-(TUNEL) assay

Lung apoptosis was assessed by measuring TUNEL positive signals using an In Situ Cell Death Detection Kit (Roche Diagnostics, Mannheim, Germany). Deparaffinized lung tissue sections were permeabilized using 0.1% Triton X-100 solution (v/v), followed by incubation with a TUNEL labeling reaction mixture for 1 hr at 37 °C. Nuclei were localized by counterstaining with 4,6-diamidino-2-phenylindole and visualized by fluorescence microscopy (Axiovert 200, Carl Zeiss AG, Oberkochen, Germany). The fluorescein isothiocyanate-labeled tissues undergoing apoptosis were recognized by their green fluorescent nuclei.

In silico protein structure prediction and covalent docking

We aligned the amino acid sequence of mouse Keap1 protein (UniProt ID: Q9Z2X8) against that of human Keap1 BTB domain (PDB ID: 4CXI) and BACK domain (PDB ID: 313 N) using BLOSUM62 scoring matrix. Different modules of the Schrödinger suite (Schrödinger; New York, NY, USA) were utilized for the following analyses. The homology model of mouse Keap1 protein was built with the “Biologics” module of the Schrödinger suite. The predicted model and the Kelch domain (PDB ID: 2DYH), which includes Cys613 residue were processed using the “Protein Preparation Wizard” to remove all water molecules and to add proton [21]. Minimization was performed on all atoms of the model structure with OPLS3e force field using the “Prime” module. The grids of size 15 Å each were generated for the docking centered at Cys257, Cys297, and Cys613 residues, respectively. The secondary structure of curcumin obtained from PubChem (CID: 969516) was optimized using the “LigPrep” module, and the keto form structure with the lowest stable penalty was selected. Curcumin was docked using the grids centered at 3 target residues. Flexible ligand sampling with the standard precision mode of the “Glide” module was used to pick the top-ranked binding pose by “Glidescore”. Subsequently, the “Michael Addition” reaction mode was employed to covalently dock the curcumin molecule to the SH functional group of each target residue [22].

Detection of curcumin-Keap1 binding

Lung tissue obtained from C57BL/6 mice were lysed in TBSN buffer (50 mM Tris-HCl, 120 mM NaCl, 0.5% MP-40, 1.5 mM EDTA). Thoroughly resuspend the magnetic beads and transfer 50 μl of magnetic beads to 1.5 ml tubes. Magnetize beads and discard supernatant, followed by washing with TBSN. Add Keap1 antibody (1 μg) in final volume of 200ul and resuspend the magnetic bead at RT for 10 min. Magnetize and wash with TBSN, followed by adding total protein (800 μg) obtained from lung lysate at RT for 1 hr. Keap1 protein-magnetic beads complex isolated by immunoprecipitation was incubated with curcumin solution (600 nM dissolved in 1% DMSO) for 1 hr at RT [16], followed by magnetizing and washing three times with TBSN. curcumin is excited at wavelength of 430 nm (Emission: 520 nm) and was detected by fluorescence multi-mode reader (Cytation 7, Agilent, California, United States) [23].

Statistical analysis

Clinical data were expressed as mean ± standard error of the mean (SEM). Data for plasma inflammatory cytokines from the clinical study were subjected to the two-tailed Welch’s t-test for unequal variance correction using GraphPad Prism Software Ver 9.3.1 (GraphPad, San Diego, CA, USA). Differences between the groups for the rest of the data were calculated using Student’s t-test for comparison and the results were shown as the mean ± standard deviation. A p-value of 0.05 or less was considered statistically significant.

Results and discussions

Curcumin reduces pneumonitis-related circulating inflammatory markers in subjects in an urban area with poor air quality

Chengdu is one of the most important economic hubs in China, even worse than Beijing [17]. Numerous studies reported that the air pollution level in cities is strongly correlated with the decrease in pulmonary function associated with the respiratory disease among residents [24]. Related, the air pollution in Chengdu was reported to have a clear correlation with COPD mortality [25]. In addition, 21.34% of the population over 40 years old in Chengdu was found to have a high risk of lung cancer [26]. It is well known that the high mortality of respiratory diseases is developed by pneumonitis, enhanced or abnormal inflammatory responses of the lungs. Exposure to air pollutants is associated with an increased incidence of pneumonitis [24]. Since pneumonitis-induced systemic inflammatory responses play a pivotal role in COPD development and progression, suppression of systemic inflammation is an effective approach to the prevention of COPD. The strong anti-inflammatory properties of curcumin have been well reported in various disease models such as dextran sulfate sodium-induced colitis, acetaminophen-induced hepatitis, and cigarette smoke-induced COPD [27], [28], [29]. Therefore, to confirm the preventive efficacy of curcumin supplementation against pneumonitis, we recruited human subjects who have lived in Chengdu for more than 1 year.

Baseline characteristics of the recruited subjects (n = 40) were summarized in Table 1. The recruited cohort was 55% male and 45% female, with an average age of 21.90 ± 2.21 years, and a BMI of 21.58 ± 2.34. Using collected plasma samples, conventional serological markers were analyzed. Interestingly, total white blood cells (WBC) and neutrophils were lower in the curcumin group compared to the placebo group (Table 1). Of note, pneumonitis is systemic inflammation characterized by activation and migration of circulating blood cells. Many studies have reported that neutrophils increased in the blood and lungs of patients with pneumonitis [30]. Related, studies demonstrated that neutrophils play a major role in the pathogenesis of COPD, and the count of neutrophils of COPD patients changes in response to air quality [31]. Also, neutrophil–lymphocyte ratio (also known as NLR) is clinically used as a diagnostic marker for lung function impairment and COPD [32]; in the study, we noted that the NLR was decreased by 8% in the curcumin fed subjects (p < 0.05; Supplementary Fig. 1A). However, platelet to lymphocyte ratio (PLR, another inflammatory response) was not different between groups (Supplementary Fig. 1B). Overall, our clinical serological data showed that curcumin attenuates circulation of immune cells, a key player in pneumonitis while not presenting any noticeable adverse responses in the subjects (e.g., allergic reaction or liver toxicity; data not shown).

The migration and activity of WBCs, such as neutrophils, are regulated by cytokines released into the blood. Also, systemic pro-and anti-inflammatory cytokines are associated with pulmonary disease risk and are widely accepted as biomarkers for acute lung injuries [33]. To explore the implications of curcumin on circulatory cytokines, we measured a total of 10 inflammatory cytokines in plasma, these included: interferon (IFN)-γ, IL-1β, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p70, IL-13, and TNF-α. The cytokines were quantified and compared individually and/or combined between groups. Interestingly, a recent study reported on 12 plasma cytokine levels in healthy Chinese subjects [34] which allowed us to indirectly compare the circulatory cytokines between different regions of China. Compared with the published data, pro-inflammatory cytokine levels were much higher in the Chengdu subjects recruited from our study (IFN-γ, 2-fold; IL-1β, 10-fold; IL-8, 2.5-fold). Within our dataset, next, we analyzed the intervention effects of curcumin. Among the markers, pro-inflammatory IL-1β and IL-8 were lower in the curcumin group than the placebo group (p < 0.05 and p < 0.01, respectively; Fig. 1A-D). Also, when combined, protein expression of pro-inflammatory cytokines was significantly lower in the curcumin group than those of the placebo group (Fig. 1G). This result is noteworthy given that IL-1β is one of the most biologically active cytokines in the early phases of acute lung injury [35]. Other studies also demonstrated that plasma IL-1β and IL-8 levels showed positive predictive of clinical outcomes such as increased permeability of the alveolar barrier or recruitment of inflammatory cells such as neutrophils [36], [37]. Similarly, anti-inflammatory cytokines were examined and compared between groups; there was a trend of increase in IL-4 (p = 0.08) in the curcumin group while IL-10 was significantly higher in the curcumin group than in that of the placebo group (Fig. 1E and F). When combined, again, curcumin significantly increased anti-inflammatory cytokines levels in plasma (Fig. 1G); to note, IL-12p70 (pro-inflammatory) and IL-13 (anti-inflammatory) were not quantifiable hence were not included for this comparison. Also, we found no difference in IL-2 and TNF-α expressions between groups (Supplementary Fig. 1C and D). Overall, our human intervention study provides a foundation that curcumin intervention can present favorable outcomes against pneumonitis thus warranting further mechanistic elucidation.

Fig. 1.

Plasma inflammatory cytokine profile was improved by curcumin intake in subjects recruited from Chengdu. (A-F) Individual cytokines, namely IFN-γ, IL-1β, IL-8, IL-6, IL-4, and IL-10, were quantified using ELISA kits (see the Methods section for details). (G) Pro- and anti-inflammatory cytokines were combined and then compared between groups. Data are present as mean ± standard error of the mean. *indicates statistically different between groups. A p-value of 0.05 or less was considered statistically significant; *p < 0.05, **p < 0.01.

Curcumin reduces acrolein-induced pneumonitis by suppression of Damage-Associated molecular pattern (DAMP) release through apoptotic death of the alveolar epithelial cell

Acrolein is a ubiquitous and highly oxidizing respiratory pollutant and humans are mainly exposed to the acrolein through cigarette smoke and mobile exhaust; acrolein induces alveolar epithelial cell deaths, GSH depletion, thus provoking inflammatory responses [10], [38], [39]. DAMPs released by alveolar epithelial cell death are a key player in the systemic inflammation response of pneumonitis [40]. Also, since the influx of acrolein drastically depletes GSH, an intracellular antioxidant, the GSH depletion renders the cell susceptible to oxidative stress, resulting in accelerated apoptotic cell death [11]. In contrast, it was previously demonstrated that curcumin transactivates Nrf2 which is a well-known transcription factor of GSH synthesis-related genes [16]. Accordingly, it is reasonable to postulate that reduction in systemic inflammatory cytokines in curcumin-fed human subjects might be, at least in part, via GSH repletion. Hence, to gain mechanistic insight by which curcumin supplementation improves air pollutant-associated pneumonitis in the context of systemic inflammatory responses, we established an animal model of pneumonitis using nasal instillation of acrolein. First, systemic inflammatory cytokine levels were examined using mice plasma samples. As expected, exposure to acrolein dramatically increased inflammatory cytokines (i.e., IL-1β, IL-6, and TNF-α), but curcumin administration significantly lowered all three cytokines (Fig. 2A) which generally agrees with our human intervention study (Fig. 1).

Fig. 2.

Curcumin mitigated acrolein-induced pneumonitis via suppressing apoptosis regardless of immunogenic cell death. (A) Systemic inflammatory cytokines (i.e., IL-1β, IL-6, and TNF-α) were quantified in plasma samples from acrolein-treated mice. Subsequently, lung damages were examined by (B) H&E staining, (C) alveolar diameter, relative mean linear intercept (MLI), (D) destructive index, and (E) inflammation score for alveolar airspace expansion. Also, markers for (F) early lung fibrosis, and (G) leukocytes (Ly6G), neutrophils (MPO), and macrophages (F4/80) in lung epithelial cell surface [labeled using laminin (LAM)] were assessed by immunofluorescence staining. As neutrophil recruitment can be controlled by damage-associated molecular pattern (DAMP), the (H) level of a DAMP molecule (i.e., HMGB1) was examined by immunofluorescence staining. (I) Immunofluorescence of N-GSDMD and p-MLKL were analyzed to determine the immunogenic cell death mechanism caused by acrolein inhalation. (J) Apoptosis, a non-immunogenic cell death, was tested by a TUNEL assay. All data are presented as mean ± standard error of the mean, and *indicates statistically different between groups at p < 0.05.

The systemic inflammatory responses led us to further investigate if the increase in the plasma IL-1β, IL-6, and TNF-α is reflected by tissue damages in the respiratory tract following acrolein exposure. Specifically, abnormal permanent enlargement of peripheral airspaces distal to the terminal bronchioles is the hallmark of chronic respiratory diseases (e.g., emphysema and pneumonitis; [41], [42], [43]). As shown in Fig. 2B, there was a dramatic increase in the destruction of alveolar architecture in mice lungs and thus compromised pulmonary structural integrity with exposure to acrolein compared to the CON group mice. In contrast, compromised lung tissue integrity, induced by the intranasal acrolein exposure, was reversed in the ACR + CUR group. Quantitative assessment of airspace size characteristics revealed that airspace enlargement from the acrolein exposed mice was notably mitigated in the ACR + CUR group mice (Fig. 2C, left panel). Similarly, mean linear intercept (i.e., a measure of morphometric change based on serial measurements of the lung using test lines) was significantly higher in the lung tissues from the acrolein-exposed mice than those of the control group (CON vs ACR group), whereas administration of curcumin ameliorated the pathological pulmonary lung injury signature induced by acrolein exposure (ACR vs ACR + CUR; Fig. 2C, right panel). Consistently, overall tissue destructive index and inflammation score were all improved in the curcumin group compared to the acrolein-treated mice group (ACR vs ACR + CUR; Fig. 2D and E, respectively). When it comes to pulmonary fibrosis, no advanced fibrotic phenotype was observed (data not shown), which might be due to our shorter acrolein exposure time than other reports [44]. However, early markers related to fibrosis, namely protein expressions of α-smooth muscle actin (SMA; green fluorescence), transforming growth factor (TGF)-β, and IL-6, in alveolar epithelial cell-specific marker, laminin-positive cell (LAM; red fluorescence) were significantly increased in the ACR group but were reversed in the ACR + CUR group (Fig. 2F; quantification results are provided in Supplementary Fig. 2A).

Recruitment of leukocytes to inflamed sites is critical for immune defence [45]. Thus, lymphocyte antigen 6 complex locus G6D (Ly6G), myeloperoxidase (MPO), and F4/80 were analyzed as representative markers for leukocytes, neutrophils, and macrophages, respectively. Positive signals (i.e., green fluorescence) from all three markers in the LAM-positive cell (red fluorescence) were dramatically increased by acrolein exposure while normalized by curcumin administration (Fig. 2G; quantification results are provided in Supplementary Fig. 2B). Since neutrophil recruitment is controlled by DAMP [46], expression of high mobility group box 1 (HMGB1), heat shock protein (HSP)60, and S100 DAMP molecules were assessed. We noted that HSP60 and S100 were not responsive, yet HMGB1 positive nucleus signals were dramatically induced by the acrolein and decreased by the curcumin (ACR vs ACR + CUR; Fig. 2H). Additionally, DAMP-releasing factors [i.e., immunogenic cell death (ICD)] were assessed; specifically, N-terminal gasdermin-D (N-GSDMD), p-(mixed lineage kinase domain-like protein (MLKL), and protease-activated receptors (PAR) were monitored as representative markers for pyroptosis, necroptosis, and parthanatos, respectively. We did not note any statistical significance in the markers although there was a trend of increase in p-MLKL in both CUR and ACR + CUR groups, compared to the CON group and ACR group, respectively (Fig. 2I and Supplementary Fig. 2C). PAR was not detected in any groups (data not shown). Despite the dramatic change in HMGB1, we did not validate the release and binding activity of HMGB1 to its receptors (e.g., toll-like receptors or receptors for advanced glycation end-products) since there was no noticeable change in markers for ICD. Instead, we shifted our focus toward non-ICD. We particularly focused on apoptosis which is plausible as recent studies reported a release of HMGB1 in apoptotic processes [47], [48]. Labeling of DNA strand breaks in situ by fluorescent TUNEL demonstrated a higher number of TUNEL-positive cells in the alveolar septa from acrolein-exposed mice compared with control mice (ACR vs CON; Fig. 2J). In the ACR + CUR group, on the other hand, curcumin markedly reduced the TUNEL-positive cells compared to those in the ACR group mice led us to further assess apoptosis mechanisms in lung tissues.

Curcumin suppresses intrinsic and extrinsic apoptotic signaling pathways triggered in the response to acrolein

To ascertain if apoptosis is indeed the main cell death signaling pathway occurring DAMPs in the alveolar epithelial cells, we conducted a histological examination of the lung sections. The alveolar epithelium consists of alveolar epithelial type 1 (AT1) and alveolar epithelial type 2 (AT2) cells. AT1 cells mainly occupy most of the alveolar surface and are essential for the air-blood barrier function of the lungs. AT2 cells crucially serve as progenitors of the alveolar epithelial cells following lung injury because they secrete surfactant protein and can differentiate into the AT1 cells. [49], [50] In the analyses, an antibody specific to podoplanin (PDPN) and surfactant protein C (SpC) was used to label AT1 and AT2 alveolar epithelial cells, respectively. The immunofluorescence analyses demonstrated a massive decrease in populations of SpC-positive AT2 cells in the alveolar septa following acrolein exposure. Such reduction in the AT2 cells was alleviated by the curcumin administration (Fig. 3A, lower left panel). On the other hand, no change was observed in PDPN-positive AT1 cells following acrolein exposure (data not shown). Related, cleaved caspase-3-positive AT2 cells were greatly increased in the ACR group lung while mitigated in the ACR + CUR group (Fig. 3A, lower right panel). In the lung tissues, inhalation of acrolein elicited a remarkably enhanced expression of cleaved caspase-3 and cleaved poly (ADP-ribose) polymerase, which represent the apoptotic index, whereas curcumin clearly ameliorated the apoptotic signals in mice following acrolein exposure (Fig. 3B and Supplementary Fig. 3A). Intrinsic (i.e., mitochondrial) and extrinsic (i.e., death receptor-mediated) apoptotic pathways synergistically contribute to organ failure, and both pathways converge on ‘effector’ caspases such as caspase-3, which executes the process of apoptotic cell death [51], [52]. Since curcumin markedly reduced the cleaved active forms of caspase-3 in the lung tissues (Fig. 3A and B), we further sought impacts of curcumin on the upstream signaling molecules involved in both intrinsic and extrinsic apoptotic pathways. Cleavage of caspase-9, an initial indicator of apoptosome-mediated apoptosis, was induced in lung tissues by acrolein exposure, while curcumin mitigated it to a level comparable to the control group (ACR vs ACR + CUR group; Fig. 3C). Additionally, acrolein exposure led to activation of the mitochondrial apoptotic pathways, as evidenced by upregulated pro-apoptotic members of the B-cell lymphoma 2 (BCL-2) protein family [BCL-2-associated X protein (Bax) and BCL-2-binding component 3 (PUMA)] and downregulation of anti-apoptotic Bcl-2 in lung tissues [53] (Fig. 3D and Supplementary Fig. 3B). On the other hand, enhanced activation of extrinsic apoptotic signaling was demonstrated by an increase in cleaved caspase-8, Fas cell surface death receptor (FAS), and phosphorylation of Fas-associated protein with death domain (FADD) in the acrolein-exposed mice lung tissues [54] (Fig. 3E and Supplementary Fig. 3C). Localization of Bax and cleaved caspase-8, via spatial distribution analysis, was assessed by immunofluorescence of lung tissues; SpC-positive cell-specific expressions of Bax and cleaved caspase-8 confirmed curcumin-mediated inhibition of both intrinsic and extrinsic apoptotic signaling (Fig. 3F and G).

Fig. 3.

Curcumin suppressed acrolein-induced apoptosis of alveolar epithelial type 2 cells via both intrinsic and extrinsic pathways. (A) The level of cleaved caspase-3 (c-Casp3) in alveolar epithelial type 2 cells [labeled using surfactant protein C (SpC)] was measured by immunofluorescence staining, and (B) Full length PARP, c-PARP, Full length Casp3 and c-Casp3 in lung tissue lysates were measured by immunoblotting. Next, key markers for (C and D) intrinsic (i.e., c-Casp9, Bax, PUMA, and Bcl-2) and (E) extrinsic (i.e., c-Casp8, FAS, p-FADD) apoptosis were examined by immunofluorescence or immunoblotting. Further, (F and G) alveolar epithelial type 2 cell-specific intrinsic and extrinsic apoptosis was confirmed by immunofluorescence, followed by spatial distribution analyses. All data are presented as mean ± standard error of the mean, and *indicates statistically different between groups at p < 0.05.

Curcumin inhibits DNA damage response and JNK/c-Jun signaling pathway in a ROS-dependent manner

Oxidative stress is attributed to excessive intracellular ROS formation, which ultimately inflicts apoptotic cell death via progressive oxidative damages to essential cellular components [55]. On the contrary, a recent study demonstrated that curcumin protected lung mesenchymal stem cells against intracellular ROS production, and restored mitochondrial membrane potential in a dose-dependent manner [56]. In addition, the study also exhibited that ROS-induced suppression of BCL-2 was reversed by dose-dependent curcumin treatment. In particular, ROS-mediated apoptotic death appears to be the main cause of alveolar epithelial cell dysfunctions [57]. Acrolein results in excessive ROS formation and thus oxidative stress by GSH depletion, thereby contributing to the apoptotic death of alveolar epithelial cells associated with pathological pulmonary lesions [39]. p53 is a well-known upstream regulator of both pro-apoptotic BCL-2 protein family members (intrinsic apoptosis) and FAS (extrinsic apoptosis) [58]. We noted that acrolein promoted transactivation of p53 in SpC-positive cells of lung tissue sections and its expression in lung tissue, which was lowered by curcumin treatment (Fig. 4A and B, and Supplementary Fig. 4A and B). The activation of p53 can be mediated by a layer of DNA damage responses including activation of ATM Serine/Threonine kinase (ATM) and H2A histone family member X (H2AX) [59]; thus, phosphorylations of ATM (p-ATM) and H2AX (γH2AX) were measured. As expected, acrolein increased both DNA damage response markers while ACR + CUR had fewer p-ATM (Fig. 4C and Supplementary Fig. 4C) and γH2AX positive cells with SpC (Fig. 4D-F and Supplementary Fig. 4D and E).

Fig. 4.

Curcumin inhibited DNA damage response and JNK/c-Jun signaling pathway by preventing GSH depletion. Representative markers for DNA damage response (DDR) pathway in alveolar epithelial type 2 cells [labeled using surfactant protein C (SpC)] were comprehensively examined by immunofluorescence staining. (A) p53 transactivation, an upstream regulator of apoptosis, was assessed as it plays a pivotal role in the DDR pathway. (B) Qualitative analysis of p53 in a whole lung tissue lysate was performed by immunoblotting. (C) Phosphorylation of ATM was measured, which induces p53 in response to DNA damage. Active ATM also (D-F) phosphorylates H2AX (γH2AX) examined by immunofluorescence staining and immunoblot assay. Alveolar epithelial type 2 cell-specific expression of γH2AX was confirmed via spatial distribution analysis. (G-I) JNK/c-Jun axis, another DDR-related signaling, was analyzed by immunofluorescence staining and spatial distribution analyses. Representative oxidative damage markers were examined by immunofluorescence staining; oxidative damages to DNA and lipid were assessed by measuring (J) 8-OHdG level and (K) 4-hydroxynonenal level, respectively. Subsequently, (L) qualitative analysis of GSH-protein complex in a whole lung tissue lysate was performed by immunoblotting. All data are presented as mean ± standard error of the mean, and *indicates statistically different between groups at p < 0.05.

A mechanistic link between c-Jun N-terminal kinase (JNK) and apoptosis should be noted. Regulatory roles of JNK in ROS-mediated apoptosis have been established [60]. ROS-mediated JNK activation (assessed by p-JNK) leads to a translocation of c-Jun to the nucleus, resulting in apoptosis [61]. In the present study, translocation of c-Jun into the nucleus was assessed by immunofluorescence. Low c-Jun-positive signals were detected in the nucleus in the CUR group compared to the CON group; c-Jun was significantly activated in response to acrolein exposure, which was normalized in the ACR + CUR group as low as CON group (Fig. 4G and H; Supplementary Fig. 4F). Consistently, p-JNK was dramatically elevated in the ACR group while it was reduced in the ACR + CUR group (Fig. 4I and Supplementary Fig. 4G).

Subsequently, we examined levels of markers related to oxidative stress in the lung tissues. Inhalation of acrolein increased the expressions of 8-OHdG and 4-hydroxinonenal (4-HNE), known representative indicators of oxidative damage to DNA and lipid, respectively [62], [63], while the markers were reduced in the ACR + CUR group, which was expected (Fig. 4J and K, and Supplementary Fig. 4H and I).

Acrolein-induced GSH depletion increases intracellular oxidative stress via excessive ROS formation [64] thus, we examined total lung GSH protein level. Not surprisingly, acrolein inhalation drastically depleted GSH in the lung tissues. In the ACR + CUR group, the reduced GSH level was dramatically recovered, suggesting that curcumin inhibits the ROS-dependent apoptotic signaling pathways by intervening with acrolein-induced pulmonary GSH depletion (Fig. 4L and Supplementary Fig. 4J and K).

Curcumin enhances de novo GSH synthesis and restoration through Keap1 conjugation

To ensure whether acrolein inhalation directly impacts the pulmonary system, acrolein adducts were examined via immunofluorescence analysis. We confirmed that nasal acrolein instillation resulted in a noticeable increase in the level of acrolein adducts in the alveolar epithelial cells, which was lowered by the curcumin administration (Fig. 5A). Accordingly, we attempted to validate whether curcumin is involved in GSH conjugation. Glutathione S-transferases (GST) is a multigene family of isozymes, known to catalyze the conjugation of GSH to an array of hydrophobic and electrophilic substrates for detoxifications, including acrolein. In our experimental condition, however, we did not observe a significant difference in pulmonary GST protein expression amongst groups (Fig. 5B and Supplementary Fig. 5A and B). Next, the first step of GSH biosynthesis is catalyzed by glutamate-cysteine ligase (GCL), which is composed of GCLC (catalytic), and GCLM (modifier) subunits [65]. GCLC exhibits all of the catalytic activity of the isolated enzyme and feedback inhibition by GSH [66] and GCLM is enzymatically inactive but plays an important regulatory function by lowering the K_m of GCL for glutamate and raising the K_i for GSH [67]. Interestingly, GCLC was increased in the ACR + CUR group compared to the ACR group, whereas GCLM was not statistically different (Fig. 5C and Supplementary Fig. 5C). This data elucidated that curcumin has the potential capacity to increase GSH synthesis through elevating catalytic activity (Fig. 5C). The second step in GSH synthesis is catalyzed by GSH synthetase (GSS). GSS is important in determining overall GSH synthetic capacity under stress conditions [68]. As a result, we noted that curcumin-treated groups (i.e., CUR and ACR + CUR) showed higher protein expression of GSS, compared to the CON and ACR group, respectively (Fig. 5C).

Fig. 5.

Curcumin promotes de novo GSH synthesis via enhancing redox sensing of Nrf2. (A) The level of ACR-protein adducts was assessed in lung tissues by immunofluorescence. Next, (B and C) glutathione (GSH) conjugation (i.e., GST) and synthesis (i.e., GSS, GCLM, and GCLC) enzyme expressions were examined by immunoblotting. As GSTs and GSH synthesis enzymes are governed by the Nrf2-Keap1 pathway, (D-F) Nrf2 and Keap1 levels were measured by immunofluorescence. Further, (G) computational molecular docking prediction was carried out to identify potential interaction(s) between curcumin and Keap1 protein. (H) Proposal model of mechanism: increased redox sensing of Nrf2 by curcumin. All data are presented as mean ± standard error of the mean, and *indicates statistically different between groups at p < 0.05.

These GSH synthesis enzymes (e.g., GCLs, and GSS) are regulated by Nrf2-target genes in response to oxidative stress [69]. Nrf2 protein in the cytoplasm is characterized by the presence of two different pools. One is free-floating Nrf2 (fNrf2), which functions as a redox-sensitive probe and the other is Keap1-bound Nrf2 (kNrf2), that is destined for ubiquitination and proteasomal degradation. Under basal conditions, there is a constitutive synthesis of new Nrf2 protein and persistent ubiquitin-dependent degradation of Nrf2 while maintaining equilibrium. Hence, only a small fNrf2 pool exists sufficient for constitutive activity [70]. Under oxidative stress conditions, on the other hand, the Nrf2 pool is increased due to hindered ubiquitination and is ready to react with a proper antioxidant response. Nrf2 capacity of binding to Keap1, a gatekeeper to control the availability of the redox-sensitive Nrf2, is diminished because of Keap1 self-ubiquitination [71], [72]. In our experimental conditions, curcumin increased the nuclear localization of Nrf2 in response to acrolein treatment (Fig. 5D and E). In particular, we noted a decreased Keap1 protein level in alveolar epithelial cells of curcumin-treated groups (CUR and ACR + CUR) compared to their respective counterparts (CON and ACR, respectively; Fig. 5F). Our results are in agreement with other studies that suggest a direct interaction between curcumin and Keap1 [73], [74]. The fNrf2 can sense the change of redox milieu and transmit redox signals to the cell nucleus via gradient nuclear translocation [75]. Hence, by increasing fNrf2, curcumin enhances the redox sensing of Nrf2, and thus GSH restoration is elevated in the response to acrolein.

To predict specific binding sites of Keap1 protein, in silico molecular docking analysis was conducted. Molecular docking is a well-established method to predict target binding site(s) of ligands [76], and a computer-based structure–activity relationships model is a useful tool to design ligand-based chemicals [77]. The cysteine residue of Keap1 protein is associated with dissociation of Keap1-Nrf2 complex. A.L Eggler et al., demonstrated the importance of 6 cysteine residues modified in dissociation of NRF2-Keap1 complex (Cys151, Cys257, Cys273, Cys288, Cys297, Cys613) [78]. Moreover, J.W Shin and his colleague suggest docking model of Cys151, Cys273, and 288 residues of curcumin with Keap1. Especially, Cys151 is completely confirmed to bind with Keap1 and curcumin through site-directed mutagenesis [16]. However, binding modes of other cysteines residues (Cys257, Cys297, and Cys613) have not been explored yet. Under our prediction conditions, there were 3 scenarios where α,β-unsaturated carbonyl groups of curcumin (the sole electrophile of curcumin; [79]) can directly interact with Keap1 protein; Fig. 5G and Supplementary Fig. 5E depict the 3 cysteine residues of Keap1, residing in either the intervening region domain (Cys257 and Cys297) or C-terminal region domain (Cys613). Specifically, our molecular docking simulations suggested covalent binding potential between unsaturated carbonyl groups of curcumin and reactive target cysteine residues of Keap1 by Michael addition. A hydrogen-bonding interaction has been found in the keto group of curcumin (i.e., carbonyl groups) with Cys257 residue (pink arrow in Fig. 5G-1 and purple arrow in Supplementary Fig. 5E-1) and the phenolic hydroxyl group of curcumin with the carboxyl group of Glu218 residue (yellow arrow in Fig. 5G-1 and purple arrow in Supplementary Fig. 5E-1). Also, a π-π stacking interaction was confirmed in phenolic groups of curcumin and Tyr255 as well (blue arrow in Fig. 5G-1 and green arrow in Supplementary Fig. 5E-1). As a result of a covalent bond between curcumin and Cys297 (pink arrow in Fig. 5G-2 and black bar in Supplementary Fig. 5E-2), the phenolic hydroxyl group of curcumin forms a hydrogen bond with the Phe262 residue backbone carboxylic acid group (yellow arrow in Fig. 5G-2 and purple arrow in Supplementary Fig. 5E-2). Last, curcumin forms a covalent bond with the Cys613 (pink arrow in Fig. 5G-3 and black bar in Supplementary Fig. 5E-3) and forms a hydrogen bond with the carboxamide group of Asn346 (yellow arrow in Fig. 5G-3 and purple arrow in Supplementary Fig. 5E-3), and the hydroxymethyl group of Ser351 (blue arrow in Fig. 5G-3 and purple arrow in Supplementary Fig. 5E-3). Covalent docking affinity of curcumin to Cys257, Cys297, and Cys613 were − 2.60, −1.87, and − 2.01 kcal/mol, respectively. Furthermore, we confirmed the direct binding Keap1 protein and curcumin through immunoprecipitation. The curcumin is reported a having fluorescence properties [23]. Thus, in order to examine the binding Keap1 protein and curcumin, we pulled down the Keap1 protein using antibody and beads, in turn, incubated with curcumin in the Keap1-antibody-bead complex. The fluorescence intensity of curcumin was measured at 520 nm wavelength. Consequently, the intensity was dramatically increased in curcumin incubated group (Supplementary Fig. 4D). The result indicated that curcumin directly binds with Keap1 protein. Dissociation of the Nrf2-Keap1 complex can be induced by changes in cysteine residues of the Keap1 protein thus our results provide strong evidence that, through the aforementioned cysteine residues, curcumin can enhance the redox sensing of Nrf2 thereby enhancing GSH in response to acrolein exposure.

Conclusion

Our study provides detailed insights into the molecular mechanism underlying the protective effect of curcumin against pneumonitis symptoms caused by air pollution. In our randomized controlled trial, curcumin-supplemented human subjects in Chengdu, the third most air-polluted area in China, showed a noticeable reduction of circulating pro-inflammatory cytokines, predictive biomarkers of clinical pneumonitis, with improved anti-inflammatory cytokines. Our follow-up pre-clinical study provided evidence that curcumin mitigates acrolein-induced pneumonitis by reducing oxidative stress and suppressing both intrinsic and extrinsic apoptosis. Using molecular docking analysis, we also revealed that curcumin could directly bind to Keap1 protein which can lead the enhanced redox sensing and thus sensitized GSH synthesis and restoration. Indeed, curcumin treatment increased nuclear Nrf2 in response to acrolein and consequently attenuated the acrolein-induced depletion of GSH, a major antioxidant and redox regulator in cells (Fig. 5H). Conclusively, both animal study and human trial clearly demonstrated the preventive potential of curcumin, as a safe and effective anti-inflammatory agent, against air pollution-induced respiratory diseases including pneumonitis and COPD.

Compliance with ethics requirements.

All studies involving human participants were approved by the Ethical Committee of Foshan University, Foshan, China (IRB Approval number: Fosu2019012). All experiments involving animals were conducted according to the ethical policies and procedures approved by the Animal Care and Use Committee of Korea University (Approval number: KUIACUC-2019–0060). All animal care and experimental procedures in this work were by Guide for the Care and Use of Laboratory Animals (National Institutes of Health).

CRediT authorship contribution statement

Eun Hee Jo: Writing – original draft, Methodology, Data curation, Investigation. Ji Eun Moon: Conceptualization, Writing – review & editing, Methodology, Data curation, Investigation. Moon Han Chang: Conceptualization, Writing – review & editing, Investigation. Ye Jin Lim: Investigation. Jung Hyun Park: Methodology. Suk Hee Lee: Writing – review & editing, Methodology. Young Rae Cho: Software, Methodology, Visualization, Investigation. Art E Cho: Validation, Writing – review & editing. Seung Pil Pack: Writing – review & editing. Hyeon-Wee Kim: Writing – review & editing, Funding acquisition. Liana Crowley: Writing – review & editing. Brandy Le: Writing – review & editing. Aykin-Burns Nukhet: Writing – review & editing. Yinfeng Chen: Software, Writing – review & editing. Yihang Zhong: Formal analysis, Investigation. Jiangchao Zhao: Conceptualization, Supervision, Writing – review & editing. Ying Li: Conceptualization, Resources, Supervision, Writing – review & editing. Hanvit Cha: Supervision, Investigation, Writing – review & editing. Jeong Hoon Pan: Conceptualization, Supervision, Writing – review & editing, Investigation. Jae Kyeom Kim: Conceptualization, Funding acquisition, Supervision, Writing – review & editing. Jin Hyup Lee: Conceptualization, Funding acquisition, Supervision, Writing – review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

The following are the Supplementary data to this article: