Chemical composition and anti-oxidant potential on essential oils of Thymus quinquecostatus Celak. from Loess Plateau in China, regulating Nrf2/Keap1 signaling pathway in zebrafish

Chemical profile and antioxidant potency of essential oils (EOs) of Thymus quinquecostatus Celak. (thyme oils) obtained from Loess Plateau in China had been studied. 130 constituents of thyme oils were determined using gas chromatography-mass spectrometry (GC–MS) and carvacrol ethyl ether was firstly reported as a new natural product, which has been used as a synthetic flavoring substance with no safety concern. The thyme oils showed the anti-oxidant activity using 2,2 diphenyl-1-picrylhydrazyl (DPPH), 2,2′-azino-bis-(3-ethylbenzothiazoline-6-sulfonate) (ABTS), ferric reducing antioxidant power (FRAP) and thiobarbituric acid reactive substances (TBARS) and conferred protection against oxidative stress in zebrafish. In addition, a class of carvacrol analogues was found to develop as potential natural antioxidant products of thyme oils from Loess Plateau by the correlation analysis. YL-thyme oil performed the best antioxidant activity in this research, which could be recommended as preferred sources of thyme oils. Furthermore, YL-thyme oil exhibited a potent antioxidant capacity by reactive oxygen species (ROS) scavenging, enhancing the endogenous antioxidant system, inhibiting lipid peroxidation and activation of Keap1/Nrf2 pathway in zebrafish.

extraction of eos. Dry samples (300 g) of thyme from the four producing areas were severally soaked in deionized water at room temperature about 25 °C to stand overnight, and then were subjected to hydrodistillation in a modified clevenger-type apparatus for 10 h. The obtained EOs were extracted with n-hexane. Subsequently, the extracts were exsiccated by anhydrous sodium sulphate and stored in a stopper vials at 4 °C in a refrigerator until the time of analysis.
GC-MS analysis. EOs were analyzed by an HP 5890 gas chromatograph (GC) using an HP 5973 mass selective detector (Finnigan Trace-MS 2000, USA) with the electron ionization mode (70 eV). The capillary column was a HP-5MS (5% phenyl methyl siloxane, 30 m × 0.25 mm i.d., 0.25 μm film thickness). The temperature of the inlet and detector, injection volume and split ratio were set to 260 °C, 260 °C, 1 µL (TBME solution), 40 : 1, respectively. The oven temperature started at 50 °C, ramped to 100 °C at a rate of 5 °C/min, ramped to 150 °C at a rate of 3 °C/min, and ramped to 250 °C at a rate of 5 °C/min, then held at 250 °C for 30 min. Helium (99.999%) carrier gas was kept with a constant flow of 1.2 mL/min. A standard n-alkane mixture (C 6 -C 24 ) was also analyzed using the above conditions to determine the retention index (RI).

Identification of compounds.
The thyme oils were identified by the retention indices (RI), relative to the series of alkanes (C 6 -C 24 ) at the same chromatographic conditions, referring to the Van Den Dool method 11 . The data were analyzed by the Xcalibur 1.1 software, compared with NIST/EPA/NIH database (1998, version 1.6) and several references 12,13 . The book "Identification of Essential Oils Components by Gas Chromatogrtaphy/ Mass Spectrometry, 4th edition" (Robert P. Adams, PhD) equally played a vital role in identifying the individual compounds.
DPPH assay. The DPPH test was carried out as described before 14 . 0.1 mL thyme oils at various concentrations of (0.02-5.67 mg/mL) were added in 0.1 mL ethanol solution of DPPH (0.001 mg/mL) acquired from Sigma-Aldrich (Saint Louis, MO, USA). The absorbance of each solution was measured at 517 nm, with ethanol being a blank. Ascorbic acid was used as a positive control. Simple regression analysis was used to derive the IC 50 value.
ABTS assay. The antioxidant activity of the samples was determined by using the method described previously 15 . An ABTS stock solution was prepared by dissolving ABTS (Sigma-Aldrich) (7 mM) in water. An ABTS radical cation (ABTS ·+ ) solution was then prepared by mixing the ABTS stock solution with potassium persulfate solution (2 mM) in equal quantities and in the dark at room temperature. The ABTS ·+ solution was www.nature.com/scientificreports/ diluted with PBS (10 mM, pH 7.4) to an absorbance of 0.70 ± 0.02 at 734 nm and stabilized at 30 °C. Next, 0.1 mL thyme oils at different concentrations (0.02-2.32 mg/mL) were mixed with 0.1 mL dilution ABTS •+ solution with ethanol; after 10 min, the absorbance of the mixture was measured at 734 nm in the dark at room temperature. As a positive control, ascorbic acid was chosen to be employed. Simple regression analysis was used to derive the IC 50 value.
FRAP assay. FRAP activity was estimated by the assay described by Benzie and Strain (1996) 16 . 5 mL of freshly prepared FRAP reagent and 20 uL thyme oils (2.32-16.87 mg/mL) were mixed and the resulting solution was vortexed and then incubated at room temperature for 70 min. After the incubation, the absorbance was measured at 594 nm. Ferrous sulfate solutions (FeSO 4 ) (1-33 mmol·L −1 ) were used to obtain the standard curve and trolox (0.975 mg/mL) was used as the positive control. The results were expressed as mmol of ferrous sulfate equivalents per gram per liter of thyme oils. The ferric reducing activity of all samples was expressed as FeSO 4 equivalent (mmol·L −1 ·g −1 DW).

Measurement of lipid peroxidation by thiobarbituric acid reactive substances (TBARS)
assay. The TBARS assay is widely used to measure lipid oxidation and antioxidant activity in food and physiological systems. Modified TBARS were used to measure the lipid peroxide using mouse liver homogenates as lipid-rich media 17  Antioxidant activity in Tg (krt4: NTR-hKikGR) cy17 zebrafish larvae. 24-hpf Tg (krt4: NTR-hKikGR) cy17 zebrafish (15 embryos/well) were added to 24-well plates, and incubated with or without metronidazole (MTZ) (5 mM) (Sigma-Aldrich). Thyme oils were diluted with fish water to different concentrations (5, 10 and 20 μg/mL) to a final volume of 2 mL. After 24 h of incubation, zebrafish larvae were anesthetized using 0.16% tricaine, and then fluorescence was detected by FSX100 Bio Imaging Navigator Equipment. N-acetyl-lcysteine (L-NAC) (0.2 mM) was used as a positive control. L-NAC is a classical antioxidant that can reduce the damage to cells caused by oxidative stress and has been widely used as an antioxidant positive control in vivo 19,20 .
The relative antioxidant capacity of the thyme oils can be determined by counting the number of fluorescence spots using Image-Pro Plus software. A fixed square region (18.00 cm × 6.00 cm) of the images on zebrafish body was selected and calculated the number of fluorescence.
Measurement of intracellular ROS production in AB zebrafish larvae. The dichloro-dihydrofluorescein diacetate (DCFH-DA) method was used to quantify intracellular ROS levels 21 . All treatment solutions were diluted to different concentrations using fish water. 24-hpf AB zebrafish (15 embryos/well) were added to 24-well plates, and incubated with or without thyme oils (5, 10 and 20 μg/mL) to a final volume of 2 mL for 1 h. Then, 2,2-Azobis (2-amidinopropane) dihydrochloride (AAPH) (12 mM) was added to incubate for 24 h. L-NAC (0.2 mM) was used as a positive control. A ROS kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) was used for analyzing generation of ROS in larvae at 48 hpf. The treatment groups were treated with 30 μM DCFH-DA and incubated in the dark at 28 ± 0.5 °C for 40 min. The zebrafish larvae were then washed with fish water three times and were observed using a fluorescence microscope (SZX16, Olympus, Tokyo, Japan). Fluorescence intensity of each zebrafish larva was determined using the Image-Pro Plus software. www.nature.com/scientificreports/ Assessment of antioxidative enzyme and lipid peroxidation activities in AB zebrafish larvae. The test was carried out as described before 18 . Briefly, 24-hpf AB zebrafish were transferred into sixwell plates (100 larvae/3 well/group), and incubated with or without AAPH (12 mM) and YL-thyme oils (5, 10 and 20 μg/mL) to a final volume of 5 mL. L-NAC (0.2 mM) was used as a positive control. All treatment solutions were diluted to final different concentrations using fish water. Larvae in each group were pooled together and homogenized on ice with 500 μL ice-cold physiological saline at 48 hpf. The supernatants were collected for analysis of antioxidative enzyme and lipid peroxidation activities after centrifugation at 3,500 rpm for 15 min at 4 °C. SOD, CAT activity and MDA levels were assessed using commercial kits (Nanjing Jiancheng Bioengineering Institute) in accordance with the manufacturer's protocols. SOD and CAT activity were expressed as unit per milligram protein and MDA levels were showed as nmol per milligram protein. And the protein content was determined using the bicinchoninic acid method 22 .
Quantitative real-time polymerase chain reaction (qRT-PCR) assay in AB zebrafish larvae. 24-hpf AB zebrafish were transferred into six-well plates (30 embryos/well), and incubated with or without AAPH (12 mM) and YL-thyme oils (5, 10 and 20 μg/mL) to a final volume of 5 mL. L-NAC (0.2 mM) was used as a positive control. All treatment solutions were diluted to final different concentrations using fish water. Total zebrafish RNA was extracted from 30 larvae (48 hpf) using the TRIzol reagent (Invitrogen, Waltham, USA) following the manufacturer's instructions. RNA concentrations and quality were evaluated on the basis of OD 260 / OD 280 ratio. Then the synthesis of cDNA was carried out by using the HiScript II Q RT SuperMix (Vazyme, Nanjing, China), and RT-PCR was done by using a RT-PCR system (Biorad, CA, USA) using the SYBR Green mix (Takara, Dalian, China). Thermal cycling was set at 95 °C for 5 s and 60 °C for 30 s with 40 cycles. Transcription of target genes was calculated using the 2 −ΔΔCt method. β-actin was chosen as the house-keeping gene. The primer sequences of the genes that were detected are shown in Table 1.

Statistical analysis.
All the assays were conducted in triplicate and results were expressed as mean values ± standard deviation (SD). Analysis of variance (ANOVA) test was performed to check significant differences. Comparison between the groups was performed using Student-Newman-Keuls (SNK). Differences from controls were considered significant when p was less than 0.05 or 0.01. Removal of the compounds with concentrations less than 0.1% had no effect on the overall analysis. Therefore, the thyme volatile constituents with the content more than 0.1% were evaluated using Principal Component Analysis (PCA) and Hierarchical Clustering Analysis (HCA), which could highlight the similarity and dissimilarity of the EO constituents and expound the active ingredients of thyme oils. Correlation analyses were performed by using a two-tailed Pearson's correlation test. All statistical analyses were carried out by using OriginPro 2017 and GraphPad Prism (version 6.01).

Results and discussions
Chemical compositions of thyme oils. The aerial parts of T. quinquecostatus were hydrodistilled and produced a yellow EOs with characteristic odor, with the yield of 4.67 mL/kg (YL), 2.33 mL/kg (JB), 5.00 mL/ kg (QY) and 2.67 mL/kg (LD). Total ion chromatograms (TICs) given by GC-MS of the thyme oils was shown in Fig. 3 (YL-thyme oil (Fig. 3a), JB-thyme oil (Fig. 3b), QY-thyme oil (Fig. 3c) and LD-thyme oil (Fig. 3d)). And the retention indices (RI) and percentages of thyme oil chemical compositions were listed in Table 2.

No
Gene symbol Forward primer (5′-3′) Reverse primer (5′-3′) AAG GCC GTT TG  CAT GAG GGT TGA AGT GCG GA   2  Cat  ACT ACA AGA CTA ATC AGG GCA TTA AGAA CCC ACA GGA ATC AGA GGG AACT   3  Keap1  CCA ACG GCA TAG AGG TAG TTAT  CCT GTA TGT GGT AGG AGG GTT   4  Nrf2  TTG TCT TTG GTG AAC GGA GGT  CTC GGA GGA GAT GGA AGG AAG   5  Hmox1  ATG CCC TTG TTT CCA GTC  www.nature.com/scientificreports/ The borneol was observed in amounts of 6.50% in YL-thyme oil, similar to that of JB-thyme oil (5.61%). It showed the content of 1.52% in the QY-thyme oil, similar to that of LD-thyme oil (1.38%). Nowadays, it is a famous compound in prescription for cardiovascular and cerebrovascular diseases 25 . Additionally, β-bisabolene is a popular resource for the synthesis of many natural products and it has been approved as a food additive in Europe 26 . The content of β-bisabolene was the highest in QY-thyme oil (2.84%), followed by that of LD-thyme oil (2.36%), JB-thyme oil (1.03%) and YL-thyme oil (1.12%).
The individual thyme oil composition from the four regions. Notably, the carvacrol ethyl ether was firstly discovered as a new natural product and its amount is significant in YL-thyme oil (31.80%) and JB-thyme oil (23.32%). It is the pale yellow clear oily liquid and has spicy odor. Carvacrol ethyl ether has been used as a synthetic flavoring substance with no safety concern at current levels of intake evaluated by the Joint FAO/WHO Expert Committee on Food Additives (JECFA). The interesting component ρ-vinyl guaiacol was indicated with the highest content in QY-thyme oil (23.55%), followed by that of YL-thyme oil (3.97%). Vinyl guaiacol appeared as colorless or canary yellow liquid that mostly found in the volatile from corn alcohol fermentation and it is a high value-added product widely used in cosmetic, pharmaceutical and chemical industries 27 . Additionally, JB-thyme oil displayed the highest content of terpinen-4-ol (10.56%) while 4.96% in YL-thyme oil. Thymol, the isomeric of carvacrol, was observed as a dominant compound (16.32%) in QY-thyme oil. It exhibited significantly antifungal, antibacterial and anti-inflammatory properties, together with anticonvulsant and antiepileptogenic activities as adjuvant to AEDs 28 . Finally, there were two components that had not been identified. One compound (RT = 16.32, RI = 1,277.300) composed 54% of the total volatile ingredients in LD-thyme oil. And another one (RT = 8.6, RI = 1,027.931) accounted for 10.16% of the volatile composition in YL-thyme oil.
Multivariate analysis. The content of compounds greater than 0.1% of the thyme oils had been used for chemometric analysis. Both the chemical composition and their content of the total EOs from the four regions were obviously different. It was interesting that the thyme oils from JB and YL were in great similarity. It might result from the close distance between JB and YL, as well as the similar growing conditions such as climate, strength of illumination, humidity of soil and air etc. Thus, growing condition might be the main influencing factor to the chemo-variation in the thyme oils produced from the four regions.
PCA and HCA analyses of the thyme oil compounds with content greater than 0.1%. PCA is an unsupervised multivariate method and transforms the original set of variables to a new set of uncorrelated variables called principal components (PCs). By plotting the PCA scores, it is possible to visually assess similarities between samples and determine whether samples can be grouped or not 29 . In present study, score plots were created in order to assess clustering tendency of samples, and three significant principal components   (Fig. 4a). As shown in Fig. 4c, YL-thyme oil had relatively high scores on PC1, while the QY-thyme oil with the mostly negative PC1 scores. On PC2, the LD-thyme oils had quite negative scores. The loading vectors responsible for the separations were distributed as Fig. 4b: loading 33, 41, 40 and 44, corresponding to compounds borneol, isobornyl acetate, carvacrol methyl ether and carvacrol ethyl ether, respectively. They were principally responsible for the discrimination of the YL-thyme oil on the PC1. The vectors 28 and 36, represented by linalool and γ-terpineol, were principally involved in the discrimination on the LD-thyme oil. And the vector 42 (thymol) and 46 (ρ-vinyl guaiacol) were strongly featured in separating QY-thyme oil with high negative scores on the PC1.
As with PCA, HCA is also an unsupervised multivariate method, which evaluates the samples by taking account of the chosen distance between samples and group average method was adopted when Euclidean Distance was applied as the measurement 30 . Using D linkage = 22.9 as a criterion for the selection of the number of groups, three clusters were obtained: Cluster I (QY-thyme oil), Cluster II (LD-thyme oil) and Cluster III (JB-thyme oil and YL-thyme oil) (Fig. 4d) and the obtained clusters are in complete agreement with the groups previously defined by PCA.
In vitro antioxidant activities of thyme oils from the four regions. Phytochemicals are known to have a complex nature hence no single assay will accurately reflect all of the radical sources or all antioxidants in a mixed or complex system 31 . In this study, the in vitro antioxidant activity of thyme oils was evaluated using the DPPH, ABTS, FRAP and TBARS methods. DPPH assay. DPPH molecule that contains a stable free radical has been widely used to evaluate the radical scavenging ability of antioxidants. Comparing with the standard antioxidant ascorbic acid (IC 50 = 0.026 mg/mL), the YL-thyme oil possessed the highest antioxidant activity (IC 50 = 0.512 mg/mL), followed by 0.530 mg/mL in  Table 2. Chemical constituents of the thyme oil in the four regions. a: "-" means the compound is not identified; b: "tr" represent the percent of compound is lower than 0.05%; c:"RI" represent the Retention indices on the DB-5 column.
ABTS assay. ABTS method is more reliable due to solubility of the ABTS reagent in both aqueous and organic solvents and rapid reaction with lipophilic as well as hydrophilic antioxidant species as compared to DPPH 32 . The IC 50 values of ascorbic acid, YL-thyme oil, JB-thyme oil, QY-thyme oil and LD-thyme oil were 0.088, 1.848, 2.951, 3.205 and 4.077 mg/mL, respectively. The YL-thyme oil exhibited the best antioxidant capacity with the rate of 98.2% at 2.109 mg/mL. These results revealed that the YL-thyme oil possessed the highest antioxidant activity among the four regions (Fig. 5b). Characterized by being simple, rapid, accurate and highly sensitive . , DPPH and ABTS are often combined to determine the activity of EOs 33 . Moreover, the YL-thyme oil and JB-thyme oil showed similar DPPH and ABTS radical scavenging activity. These results were consistent with the multivariate statistical analyses (PCA and HCA), which indicated that chemical compositions of thyme oils greatly affected the antioxidant activity. FRAP assay. FRAP assay measures the reducing potential of an antioxidant reacting with a ferric tripyridyltriazine [Fe 3+ -TPTZ] complex and produces a colored ferrous tripyridyltriazine [Fe 2+ -TPTZ], so FRAP is a reasonable screen for the ability to maintain redox status in cells or tissues 31 . Calibration curves (Y = 0.0379X + 0.0103), in the range of 1 mmol·L −1 -33 mmol·L −1 , showed good linearity (R 2 ≥ 0.9992). The reducing power capacities results follow the same pattern as in case of radical scavenging assay, in which the YL-thyme oil was the most active one (2.917 mmol·L −1 ·g −1 DW), followed by that from JB (2.549 mmol·L −1 ·g −1 DW), LD (2.413 mmol·L −1 ·g −1 DW) and QY (1.956 mmol·L −1 ·g −1 DW). Comparing these results with the standard antioxidant trolox (7.28 mmol·L −1 ·g −1 DW), it was revealed that the thyme oils had significantly lower reducing potential than the standard (Fig. 5c).

TBARS assay.
Lipid peroxidation is an oxidative alteration of polyunsaturated fatty acids in the cell membranes that generates a number of degradation products. It can be initiated when fatty acids or fatty acyl side chains are attacked by free radicals. This oxidation process could affect the structural and functional damage of biomolecule 34 . For this reason, it is necessary to determine the function of these thyme oils on lipid peroxidation www.nature.com/scientificreports/ phenomenon. In biological systems, MDA is a reactive species that participates in the cross-linking of DNA with proteins and damages liver cells 35 . In the present study, Fe 2+ -H 2 O 2 system was used to induce lipid peroxidation in mouse liver homogenate. The lipid peroxidation inhibition effects of thyme oils increased with the increase of sample concentrations and significant differences (p < 0.05) among the tested thyme oils were found. By comparing the IC 50 values of YL-thyme oils with that of the authentic antioxidant ascorbic acid, it was found that the antioxidant activity of (IC 50 = 0.080 mg/mL) was quite comparable with that of ascorbic acid (IC 50 = 0.058 mg/mL). At a concentration of 0.563 mg/mL, the inhibition effect of YL-thyme oil was 88.90%. The others were ranked in the following order: JB-thyme oil (IC 50 = 0.198 mg/mL), QY-thyme oil (IC 50 = 0.241 mg/ mL) and LD-thyme oil (IC 50 = 0.376 mg/mL) (Fig. 5d). These results indicate that the thyme oils significantly inhibited MDA formation in the liver tissue and could protect cell membranes from lipid oxidation.

The correlations of thyme oil components with antioxidant activities. Furthermore, it is
important to correlate and understand which compounds contribute to the different antioxidant assays, showing specific antioxidant potential for the different radicals depending on their chemical structure. The correlations between thyme oil components and antioxidant activities were reported in Table 3. It is well known that the smaller the IC 50 value is, the better the antioxidant activity is. Therefore, compounds present in thyme oils showed a strong and positive correlation with the free radical scavenging activity (DPPH· and ABTS ·+ ), FRAP as well as TBARS. In contrast to this finding, β-bisabolene was inversely and strongly related to the DPPH (0.96561, p < 0.05). This inverse relationship would indicate that the higher content of the β-bisabolene did not exert protection, but actually promoted oxidation. Interestingly  www.nature.com/scientificreports/ data 36 . However, for FRAP, all of them were not statistically significant. These compounds mentioned above were found in the highest amounts in the YL and JB thyme-oil. This indicated that the higher level of these components is likely to be closely related to the relatively high antioxidant activity of YL and JB-thyme oil. These differences observed between assays are related to the individual molecular structure of each compound which indicated that stereoisomerism, functional groups distribution and any other structural parameters such as the oxidation state of the C ring, the hydroxylation and methylation pattern also are expected to affect the final value. In particular, there were good correlations of carvacrol ethyl ether both with IC 50 value of DPPH and ABTS and carvacrol acetate both with IC 50 value of ABTS and TBARS, which could be recommended as an indicative component for the antioxidant capacity. It is acknowledged that carvacrol is used to defined the quality of the Thymus species and widely used in the food industry as antimicrobial and antioxidant agent. Thus, in the present work, carvacrol methyl ether, carvacrol ethyl ether and carvacrol acetate were all found to be significantly associated with antioxidant activity, implying the carvacrol analogues as potential natural antioxidants of thyme oils from Loess Plateau in China.
In vivo antioxidant activity of thyme oil on the zebrafish. Although in vitro assays are used for rapid screening but their results cannot be directly extrapolated to in vivo conditions. The experimental models are therefore used to simulate in vivo conditions. Hence, we also evaluated the antioxidant capacity of thyme oils using zebrafish model.

Lethal and teratogenic effects of thyme oil.
To determine the toxicity of thyme oils, we investigated the survival rate of zebrafish larvae. The thyme oil had no toxic effect up to 40 μg/mL. Therefore, concentrations of thyme oil less than 40 μg/mL are suitable for the in vivo experiments. When the concentration reached 40 μg/mL, these larvae treated with the thyme oils at 48 hpf from four producing areas showed different degrees of morphological abnormalities, including yolk sac edema, pericardial edema, and curved body shape (Fig. 6a). Pericardial edema and yolk sac edema were the most pronounced morphological alterations. Pericardial edema was first observed in the LD-thyme oil and all zebrafish larvae in this group were dead at the concentration of 40 μg/mL. The curved body shape was first observed in the QY-thyme oil at the concentration of 40 μg/mL. Figure 6b shows the lethal effects of thyme oil at 48 hpf and mortality rates in the treatment group exhibited a dose and time-dependent increase. The values of the 1/3 LC 1 , LC 1 and LC 10 of the four thyme oils were determined to be around 5, 10 and 20 μg/mL. Antioxidant activity in Tg (krt4: NTR-hKikGR) cy17 zebrafish larvae. The in vivo capability of thyme oils on inhibiting oxidative stress was further evaluated using MTZ-induced oxidative insult in Tg (krt4: NTR-hKikGR) cy17 transgenic zebrafish model. A transgenic line of Tg (krt4:NTR-hKikGR) cy17 is established which Table 3. The correlations between thyme oil components and antioxidant activities. ** and *: significant at the 0.01 and 0.05 probability levels, respectively. www.nature.com/scientificreports/ overexpresses the NTR-hKikGR fusion protein under the control of the skin-specific krt4 promoter. It provides as a unique quantitative and fast tool to study the signaling molecules which modulate skin apoptosis in living animals 37 . The fluorescence spots of zebrafish are shown in Fig. 7a. and the anti-oxidation results of the thyme oil on transgenic zebrafish treated with MTZ are shown in Fig. 7b. The results showed that the number of fluorescence spots on the epidermal cells of the zebrafish was significantly decreased after the addition of MTZ.
In the treatment of different concentrations of thyme oils (5, 10 and 20 μg/mL), the number of fluorescence spots on the epidermal cells increased significantly compared with the MTZ-treatment group, indicating that the thyme oils can resist the overproduction of ROS and skin cell apoptosis induced by MTZ. The antioxidant capacity of the thyme oils is better than that of the L-NAC-treatment group, especially the antioxidant capacity of YL-thyme oil is the most prominent, followed by that from JB-thyme oil, QY-thyme oil and LD-thyme oil.
Protective effects of thyme oil on ROS generation in AAPH-induced oxidative stress. The antioxidative effect of thyme oil on ROS production in zebrafish larvae was examined by detection of DCFDA. Fluorescent photographs of larvae showed that the untreated zebrafish had low fluorescence intensity, whereas AAPH-induced group generated a marked increase in the observed fluorescence intensity, indicating that ROS was produced in the presence of AAPH in the zebrafish larvae.  www.nature.com/scientificreports/ The scavenging efficacy of the thyme oil on the ROS production in AAPH-induced zebrafish larvae was measured. At 48 hpf, the larvae in the YL-thyme oil-treatment and JB-thyme oil-treatment groups (Fig. 8b) exhibited markedly lower fluorescence intensities than those in the AAPH-treatment group (Fig. 8a), which indicated that ROS significantly decreased after thyme oil exposure. Conversely, fluorescence intensity in QY-thyme oil-treatment and LD-thyme oil-treatment groups (Fig. 8c) did not decrease significantly compared to the AAPH-treatment group. Hence, these results indicate that thyme oil are effective antioxidants in AAPHinduced oxidative stress model.

Mechanism of YL-thyme oil on the AAPH-induced zebrafish larvae. The outstanding in vitro and
in vivo antioxidant activities presented above gave an indisputable confirmation that YL-thyme oil and JB-thyme oil showed better antioxidant activities due to the close distance between two of them. YL-thyme oil exerted the best antioxidant activity among the four regions and further studies are required to elucidate the biological mechanisms and signaling pathways underlying the protective effects of the thyme oil.
Effects of YL-thyme oil on MDA Levels, SOD and CAT activities in AAPH-induced zebrafish larvae. The antioxidant enzymes, SOD and CAT, work within the cells to remove most of superoxides and peroxides before they react with metal ions to form more reactive free radicals. SOD, the first line of defense against free radicals, converts the superoxide radical to H 2 O 2 and O 2 by reduction. The H 2 O 2 is transformed into water and oxygen by CAT 38 . MDA, an end product of lipid peroxidation, represented the level of lipid peroxidation 39 .
The MDA levels and antioxidant enzymes activities (SOD and CAT) were examined and shown in Fig. 9. AAPH-induction decreased the activities of SOD, and CAT in zebrafish and increased MDA levels when compared to control group, whereas AAPH-induced zebrafish protected by YL-thyme oil showed remarkable damage prevention. After being treated with YL-thyme oil (5 μg/mL), the activities of SOD and CAT were increased by 1.30 folds, and MDA level decreased by 75% as compared with those of AAPH-induced group. The results implied that the protective effect of YL-thyme oil against oxidative stress might be rationalized because of the impairment of antioxidant defenses system. Obviously, AAPH damaged the antioxidant defense system and promoted the lipid peroxidation by reducing the activities of the antioxidation enzymes and upregulating the formation of MDA in zebrafish, while YL-thyme oil could prevent these harms.
Effects of YL-thyme oil on gene expression. In order to investigate the mechanism of YL-thyme oil-treated antioxidant activity, the mRNA expression levels of genes related to the antioxidant activities were measured (Fig. 10). Nrf2 and its endogenous inhibitor, Keap1, function as a ubiquitous, evolutionarily conserved intracellular defense mechanism to counteract oxidative stress and the Keap1/Nrf2 signaling pathway is an Moreover, Nrf2/Hmox1 pathway is widely studied in vertebrates and play a pivotal role in neurodegenerative disorders. Hmox1, as Nrf2-dependent gene, provides cytoprotective effect and play a crucial role in the development of oxidative and age-related disorders 40 . Also, Hmox1 is known to play an important role in cellular protection against oxidative insult in cardiovascular disease, including diabetes, and in the alleviation of vascular diseases and various chemical reagents were known to facilitate antioxidant response via activating Hmox1 expression 41 . In the present study, YL-thyme oil-treated zebrafish larvae exhibited significant downregulation of Keap1 expression and upregulation of Nrf2 expression; furthermore, the expression level of Sod1, Cat and Hmox1 were significantly increased in the treatment of YL-thyme oil groups. AAPH-induction decreased the expression of Sod1, Cat, Nrf2 and Hmox1 in zebrafish by 1.39, 1.41, 2.03 and 2.02 folds, respectively and increased Keap1 levels by 3.47 folds when compared to control group. In the YL-thyme oil-treatment group and L-NAC-treatment group, the expression levels of genes encoding the Sod1 (Fig. 10a) and Nrf2 (Fig. 10d) were significantly increased relative to that in the AAPH-treatment groups. The expression levels of genes encoding Cat (Fig. 10b) and Hmox1 (Fig. 10e) were significantly increased relative to the AAPH-treatment group; however, no obvious changes were detected in the L-NAC-treatment groups relative to the control groups. In contrast, the expression levels of genes encoding Keap1 was downregulated in the YL-thyme oil-treatment group (Fig. 10c).
Taken together, our data demonstrated that thyme oil might exert the antioxidant activity in AAPH-induced oxidative stress model by activating the cellular Nrf2/Keap1 pathway.
conclusions This is the first study on the chemical constituents, antioxidant activities and biological mechanism of EOs of T. quinquecostatus. One new natural compound, carvacrol ethyl ether, was firstly discovered in the thyme oils. It has been used as a synthetic flavoring substance with no safety concern. The thyme oils from the four regions were classified into three groups by chemometric analysis, out of which the borneol, isobornyl acetate, carvacrol methyl ether, ρ-vinyl guaiacol, thymol, carvacrol ethyl ether and linalool contributed the most to the classification, indicating that these compounds may be used for the geographical markers of the thyme oils. The YL-thyme oil was found to produce superior EO yields and possess the strongest anti-oxidant activities in vitro www.nature.com/scientificreports/ and in vivo, suggesting that domestication of this thyme oil is economical for the food industry. It was also proved with correlation analysis that borneol, carvacrol methyl ether, carvacrol ethyl ether, carvacrol acetate, isobornyl acetate and dauca-5,8-diene could be the characteristic constituents responsible for the antioxidant effect of the thyme oils. This indicated that the carvacrol analogues might be developed as potential natural antioxidants of thyme oils from Loess Plateau in China. Furthermore, Keap1/Nrf2 pathway contributed to the anti-oxidant processes of YL-thyme oil. Taken together, our results suggest that thyme oil could be useful as the natural antioxidant and have the potentiality of enhancing the protective effect against oxidative stress. The findings of the present study can provide new insights for enhancing the valuable components and antioxidant capacity of this plant in their EOs. Therefore, this study could lay foundation for the applications of thyme oils in food, pharmaceutical and cosmetic industry, and also provide promising sources for natural antioxidants and www.nature.com/scientificreports/ a new possible breakthrough point for further bioactivity studies of thyme. Admittedly, there are limitations to the current study. CC-MS has been used to quantified the thyme oils components and we didn't examine the data further by GC-FID which might be more precise than GC-MS in quantifying complex mixtures. In addition, we cannot completely determine the major contribution of components for the antioxidant activities of thyme oils. The potential antioxidants, especially the carvacrol analogues from the correlation analysis should be assessed in future studies. Furthermore, further studies can be performed to delineate other candidate genes and signaling pathways, especially those closely correlated with anti-oxidation, in thyme oils-mediated effects.
ethics statement. All experimental procedures were carried out in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals and approved by the Experimental Animal Ethics Committee of the Academic Committee of Beijing University of Chinese Medicine.