Chemical Constituents of the Essential Oil Extracted from Elsholtzia densa and Their Insecticidal Activity against Tribolium castaneum and Lasioderma serricorne

Storage pests pose a great threat to global food security. Here, we found that the essential oil (EO) extracted from E. densa possesses obvious effects against the insects that threaten stored-products. In this work, we investigated the chemical constituents of the essential oil extracted from Elsholtzia densa, and their insecticidal (contact and fumigant) toxicity against Tribolium castaneum and Lasioderma serricorne. A total of 45 compounds were identified by GC-MS, accounting for 98.74% of the total EO. Meanwhile, 11 compounds were isolated from the EO, including limonene, β-caryophyllene, ρ-cymene, trans-phytol, α-terpineol, linalool, acetophenone, 1,8-cineole, ρ-cymen-7-ol, 1-O-cerotoylgly-cerol, and palmitic acid. Furthermore, acetophenone, ρ-cymen-7-ol, and 1-O-cerotoylgly-cerol were isolated for the first time from Elsholtzia spp. The results of the bioassays indicated that the EO had the property of insecticidal toxicity against T. castaneum and L. serricorne. All of the compounds showed different levels of insecticidal toxicity against the two species of insects. Among them, 2-ethyl-1H-imidazole had no insecticidal toxicity against T. castaneum, but possessed fumigant and obvious contact toxicity against L. serricorne. ρ-Cymen-7-ol had beneficial insecticidal toxicity against the two species of insects, and fumigant toxicity against L. serricorne. ρ-Cymen-7-ol (LD50 = 13.30 μg/adult), 1-octen-3-ol (LD50 = 13.52 μg/adult), and 3-octanol (LD50 = 17.45 μg/adult) showed significant contact toxicity against T. castaneum. Acetophenone (LD50 = 7.07 μg/adult) and ρ-cymen-7-ol (LD50 = 8.42 μg/adult) showed strong contact toxicity against L. serricorne. ρ-Cymene (LC50 = 10.91 mg/L air) and ρ-cymen-7-ol (LC50 = 10.47 mg/L air) showed powerful fumigant toxicity to T. castaneum. Limonene (LC50 = 5.56 mg/L air), acetophenone (LC50 = 5.47 mg/L air), and 3-octanol (LC50 = 5.05 mg/L air) showed obvious fumigant toxicity against L. serricorne. In addition, the EO and its chemical compounds possessed different levels of repellent activity. This work provides some evidence of the value of exploring these materials for insecticidal activity, for human health purposes. We suggest that the EO extracted from E. densa may have the potential to be developed as an insecticidal agent against stored product insect pests.


Introduction
As the world's population continues to grow, the demand for food crops is increasing every year. However, a major difficulty in the production and storage of cereal products, grains, and Chinese herbs is insect infestation [1]. Pests are a problem worldwide because of migration, high reproductive rates, and wide distribution. Insect pests play an important role in damaging crops, causing both direct and indirect economic loss [2]. These insect pests are active throughout the year on various crops, and are associated with a loss of more than one billion US dollars per year worldwide [3]. Traditional methods of pest control are

Insects
T. castaneum and L. serricorne were provided by the laboratory of quality research and product development of traditional Chinese medicine, College of Life Science, Northwest Normal University, and were confirmed by Dr. Liang J. Y. (College of Life Science, Northwest Normal University, Lanzhou, China). The two kinds of insects were artificially subcultured and purified for more than 4 to 5 generations in dark incubator at 28-30 • C, with relative humidity maintained at 70-80%. They were reared on a mixture of wheat flour and yeast (10:1, w/w). The mixed-sex adult insects used in all bioassays were about 1-2 weeks old.

GC-MS Analysis
This was carried out according to the method reported previously [23]. The chemical compounds of the EO were analyzed by GC-MS. The column of the GC was a quartz capillary column HP-5MS (30 m × 0.25 mm × 0.25 µm), the temperature was 250 • C, high purity helium was used as the carrier gas, the flow rate was 1.0 mL/min, the injection volume was 1.0 µL, and the split ratio was 100:1. The heating program was set as follows: the initial temperature was 60 • C and was maintained for 2 min, then the temperature increased to 180 • C at a rate of 10 • C/min, was maintained for 1 min, before finally the temperature reached 260 • C at a rate of 20 • C/min, prior to holding it for 15 min. EI-MS was carried out at 70 eV. The EO samples were diluted in acetone to prepare 1% solutions. Further identification was made by comparison of their MS with those stored in Wiley 275 and NIST 11, or with the literature [24].

Isolation and Identification of Pure Compounds
The crude EO (75 mL) of E. densa was chromatographed on a silica gel column (160-200 mesh, Qingdao Marina Chemical Plant, Qingdao, China) (column length 90 cm, diameter 5 cm), then eluted with a gradient of petroleum ether-ethyl acetate (from 100:0 to 0:100). Each 100 mL of eluate was collected as a fraction. With the monitoring of thin layer chromatography (TLC) profiles, similar fractions were combined and at last 10 fractions were obtained. Among them, the fractions (1-5) were pooled and further purified by silica gel column chromatography, and 11 compounds were isolated from them. The isolated compounds were elucidated with nuclear magnetic resonance. 1 H and 13 C NMR were performed on an NMR spectrometer (Agilent Technologies, 400 MHz (Vnmr mercury-400 plus) and 600 MHz (Agilent DD2-600 MHz) for proton) at a temperature of 25 • C in the deuterated chloroform (CDCl 3 ).

Contact and Fumigant Activity
According to the methods of Liu and Ho [25], the drip method and filter paper sheet method were used for the whole process of contact and fumigant activity, respectively. From the raised adults, 10 healthy adults with good activity and consistent growth were selected (regardless of gender). They were placed in an activity test glass bottle (5.5 cm high, 2.5 cm in diameter). The EO and isolated compounds were dissolved in acetone to prepare a serial testing solution, with acetone solvent as the negative control. According to the results of preliminary experiments, five concentrations of the EO and its isolated compounds were determined in formal experiments. Each treatment and control of different concentrations was replicated five times. The death/survival of the test insects were observed and recorded 24 h later, and abnormal activity of the insects was regarded as death.

Repellant Activity
Repellant activity was assessed according to the method reported previously [26]. The EO, its isolated compounds, and commercial compounds were dissolved in acetone to prepare a serial testing solution (78.63, 15.73, 3.15, 0.63, 0.13 nL/cm 2 ). Acetone and DEET were used negative and positive controls, respectively. All of the experiments were carried out on Petri dishes (9 cm in diameter). A filter paper disk (9 cm in diameter) was cut into two halves. One half was uniformly treated with 500 µL of a testing solution, and the other half was treated with 500 µL of acetone. There were 20 insects in each Petri dish center. The number of insects that stayed on the treated (Nt) and control (Nc) halves were recorded after 2 and 4 h. Every treatment was repeated five times. Then, the percentage repellency (PR) of the EO and each compound was calculated using the formula: where Nc is the number of insects present in the negative control half, and Nt is the number of insects present in the treated half. At the same time, the average repellent rate was graded by reference [27], as shown in the

Data Analysis
In the bioassay of fumigant and contact, the effects of the EO and compounds were expressed by LC 50 and LD 50 values. Probit analysis [28] was used for the calculation of these values. LC 50 , LD 50 , and PR values were transferred into arcsin square root values, and subjected to One-Way Analysis of Variance (ANOVA) under Tukey's HSD test at p < 0.05 (SPSS V 19.0, IBM, NY, USA).

Structural Analysis of Isolated Compounds
From the EO of the aerial part of E. densa, 11 compounds (1-11) were isolated and identified. Among them, three compounds (7, 9 and 11) were isolated from the EO of E. densa at first. Terpenoids were isolated as the main components of the EO of E. densa. Terpenoids are generally used as flavors in the food industry, and as fragrances in the cosmetics industry.
Eleven compounds were matched with the corresponding data ( 1 H and 13 C NMR data) in the literature, and their structures are listed in Figure 1.  (9) 1-O-cerotoylgly-cerol (10) Palmitic acid (11)

Contact Activity
The contact activities of E. densa against T. castaneum and L. serricorne is listed in Table 4. The EO of E. densa showed obvious contact activity against T. castaneum and L. serricorne, with LD 50 values of 29.20 and 24.29 µg/adult, respectively. When compared with positive control pyrethrins, the EO demonstrated weak contact activity against T. castaneum and L. serricorne. However, compared with other EOs described in the literature, with which similar bioassays were used, the EO of E. densa possessed stronger contact activity against T. castaneum and L. serricorne. This included in comparison to the EOs of Ajania fruticulosa, Evodia lepta and Citrus wilsonii, which exhibited contact activity against T.castaneum at LD 50 values of 105.67, 166.94 and 48.49 µg/adult, respectively [43][44][45]. Meanwhile, the EO of Zingiber zerumbet showed contact activity against L. serricorne, with an LD 50 value of 48.3 µg/adult [46]. For T. castaneum, only 2-ethylimidazole possessed no contact activity (LD 50 > 300 µg/adult), even at the highest testing concentration (50%, v/v), while the other 16 compounds showed different levels of contact activity. Among them, ρ-cymen-7-ol, 1-octen-3-ol and 3-octanol showed significant contact activity against T. castaneum, with LD 50 values of 13.30, 13.52 and 17.45 µg/adult, respectively. Phytol, disobutyl phthalate, and dioctyl phthalate exhibited weak contact activity, with LD 50 values of more than 100 µg/adult. For L. serricorne, 17 testing compounds exhibited strong contact activity at all the assayed concentrations. Most notably, ρ-cymen-7-ol, acetophenone, geraniol, 1-octen-3-ol, and 3-octanol showed very significant contact activities against L. serricorne, with LD 50 values of 8.42, 7.07, 3.14, 3.39 and 7.75 µg/adult, respectively.
Based on the above experimental results, we found that the contact activity of the EO was significantly weaker than that of the some compounds against T. castaneum and L. serricorne, suggesting that complex mixtures of active ingredients might also be beneficial in terms of pest resistance and behavioral desensitization. The different compounds showed different levels of contact activity against T. castaneum and L. serricorne, which may be due to the different sensitivity of different insects to the substances. As the temperature rises [47], the contact activity may become less effective.

Fumigant Activity
The fumigant activity of E. densa EO and its components against T. castaneum and L. serricorne is described in Table 5. The EO of E. densa showed significant fumigant activity against T. castaneum and L. serricorne, with LC 50 values of 18.45 and 14.49 mg/L air, respectively. The results showed that T. castaneum was more tolerant than L. serricorne. When compared with the famous botanical insecticide, MeBr and Phosphine, the EO demonstrated weak fumigant activity against T. castaneum and L. serriocorne. However, compared with other EOs tested using a similar bioassay in the literature, the EO of E. densa in the present study exhibited stronger fumigant activity against T. castaneum and L. serricorne, e.g., EOs of Amomum maximum and Litsea cubeba exhibited fumigant activity against T. castaneum at LC 50 values of 23.09 and 22.97 mg/L air, respectively [48,49]. The EOs of Alpinia kwangsiensis and Amomum tsaoko showed fumigant activity against L. serricorne at LC 50 values of 9.91 and 8.70 mg/L air, respectively [50,51]. The results demonstrated that T. castaneum is more tolerant than L. serricorne.

Repellent Activity
The results of repellency assays for the EO and individual compounds selected against T. castaneum and L. serricorne are shown in Figure 2. The results showed that at the testing concentrations, the EO and compounds exhibited different levels of repellent activity against T. castaneum and L. serricorne. Total repellent activity against T. castaneum was greater than that for L. serricorne. Acetophenone, ρ-cymen-7-ol, α-terpineol, and the EO showed strong repellent activities against T. castaneum at all testing concentrations, according to most data distributions in Figure 2. It was noteworthy that only linalool possessed weak repellent activity against T. castaneum, while the EO and seven compounds showed significant repellent activity at 78.63 and 15.73 nL/cm 2 at 2 h after exposure, and had PR values of almost 100%. At the testing concentration range of 3.15-0.13 nL/cm 2 , EO, limonene, α-terpineol, acetophenone, and ρ-cymen-7-ol possessed significant repellency against T. castaneum, with PR values of 80-90% at 2 h post-exposure. β-Caryophyllene, ρ-cymene, phytol, and linalool exhibited mild repellent activity with the testing concentration range of 3.15-0.13 nL/cm 2 at 2 h after exposure. 1,8-Cineole displayed significant repellent activity at all testing concentrations, with PR values of 100, 94, 74, 64 and 48%, respectively, at 2 h after exposure, and the activity increased with the concentration of exposure. Although seven compounds showed repellent activity against T. castaneum that was greater than that of EO, phytol, and linalool at the highest testing concentration at 4 h after exposure, these substances had PR values of 98, 88 and 86%. The PR values of β-caryophyllene, acetophenone, and ρ-cymen-7-ol showed significant repellent activity of almost 100% at a testing concentration of 15.73 nL/cm 2 against T. castaneum at 4 h after exposure, while the PR values of β-caryophyllene, acetophenone, and ρ-cymen-7-ol were the same as the positive control DEET at 15.73 nL/cm 2  For L. serricorne, β-caryophyllene, 1,8-cineole and ρ-cymene exhibited significant repellent activity when applied at all testing concentrations, and had PR values ranging from 50-100%, at the same level as DEET. The EO, acetophenone and ρ-cymen-7-ol showed obvious repellency only at the testing concentrations of 78.63 and 15.73 nL/cm 2 . At 4 h after exposure, β-caryophyllene, ρ-cymene, and ρ-cymen-7-ol were better than the EO and other compounds at 78.63 and 15.73 nL/cm 2 , according to PR values. α-Terpineol and phytol exhibited moderate repellency against L. serricorne at 4 h after exposure. Limonene and linalool exhibited weak repellent activity against L. serricorne at all testing concentrations at 4 h after exposure. Most notably, at the lowest testing concentration, ρ-cymen exhibited strong repellent activity (60 and 48%, respectively) against L. serricorne at 4 h after exposure. In general, the variety of the repellent activity was easily affected by the sensitivity of the insects, the testing concentration, and the exposure duration. The repellent effect is complicated to interpret, and thus further work is needed to clarify it.
The results of the repellent assays showed that the EO and nine individual compounds exhibited different levels of repellent activity against T. castaneum at the highest testing concentration when exposed for 2 h. The PR values of the EO, limonene, β-caryophyllene, ρ-cymen, α-terpineol, 1,8-cineole, acetophenone, and ρ-cymen-7-ol were all almost 100%, and no statistically significant differences occurred between them. Although limonene accounted for 22% of the total EO, it exhibited weaker repellent activity than the EO against T. castaneum at all testing concentrations, while the EO was still highly repellent against T. castaneum. In this situation, the repellent activity of the EO did not seem to be greatly affected by limonene, and it is speculated that secondary components of the EOs may play an important role in the repellent activity, and have a synergistic effect that enhances the active effect. Another speculation is that some relatively inactive compounds play an important role in the multicomponent mixture. Some compounds are not significantly active on their own, but can enhance the active effects of other compounds. The repellent activity of the main components in EOs does not completely determine the overall effect of EOs, but synergistic and antagonistic interactions between the various components may have an important effect on the overall activity. Trongtokit et al. [53] reported that interactions between two or more components are caused by different ratios. The appropriate combination of compounds can increase the synergistic effect considerably. Compound proportioning is a complex and interesting issue that deserves further investigation. and ρ-cymen-7-ol were the same as the positive control DEET at 15.73 nL/cm 2 against T. castaneum at 4 h after exposure. The results indicated the EO and compounds possessed approximately equal repellent activity with DEET against T. castaneum at the testing concentrations of 78.63 and 15.73 nL/cm 2 at 2 and 4 h post-exposure. Unfortunately, ρcymene showed weak repellent activity against T. castaneum with the testing concentration range of 3.15-0.13 nL/cm 2 at 4 h after exposure, while other compounds exhibited strong repellent activity; it had PR values of 40-80%. For L. serricorne, β-caryophyllene, 1,8-cineole and ρ-cymene exhibited significant repellent activity when applied at all testing concentrations, and had PR values ranging from 50-100%, at the same level as DEET. The EO, acetophenone and ρ-cymen-7-ol showed obvious repellency only at the testing concentrations of 78.63 and 15.73 nL/cm 2 . At 4 h after exposure, β-caryophyllene, ρ-cymene, and ρ-cymen-7-ol were better than the

Conclusions
In this work, the EO extracted from E. densa and its chemical compounds exhibited effective insecticidal toxicity and repellency against T. castaneum and L. serricorne. Compared with synthetic chemical insecticides, plant-derived materials have many advantages in pest management. They are scalable, safe for human health, and environmentally friendly. E. densa is used in traditional Chinese medicine, and has a long history. The characteristics highlighted above increase the feasibility of developing spice plants from the Elsholtzia family into eco-friendly pesticides, and suggest that the EO of Elsholtzia may have promise as a bioinsecticide or green repellent for pest management in warehouses and grain stores. However, further work should be devoted to investigating the mutual interactions among individual compounds, including their mechanism of action and possible relationships between toxicity and repellency. Meanwhile, further work should also be devoted to the effect of temperature on insecticidal and repellant activity. Whether the EO is actually safe for humans and the environment in the long term requires further experiments to verify. In addition, the target of action against insects should be confirmed.