Repellency of zerumbone identified in Cyperus rotundus rhizome and other constituents to Blattella germanica

The compounds 1,8-cineole and zerumbone (ZER) from the Cyperus rotundus rhizome along with another 11 previously identified rhizome essential oil constituents and α-humulene, which lacks the only carbonyl group present in ZER, as well as binary mixtures of ZER and seven active compounds were tested for repellency to male B. germanica. The results were compared to N,N-diethyl-3-methylbenzamide (deet). In filter-paper choice tests, ZER was the most repellent compound, and α-humulene was ineffective, which indicates that the α,β-unsaturated carbonyl group of ZER is a prerequisite component for repellency. At 81.5 μg cm−2, enhanced repellency was produced by binary mixtures of ZER and 1,8-cineole, (+)-dihydrocarvone or (R)-(+)-limonene (70:30, 50:50 and 30:70 ratios by weight). These mixtures were very effective against male B. germanica within 24 h and were more repellent than a single compound or deet alone. The optimum ZER content was determined to be more than 50%. In Ebeling choice box tests at 652.4 μg cm−2, these compounds and deet resulted in complete repellency to intact male B. germanica, while they exhibited 35–47% repellency to antennectomized male one. Mixtures formulated from the active constituents of the C. rotundus rhizome could be useful as potential repellents for controlling B. germanica.

are difficult to apply to inaccessible places, do not reach deep, insecticide-free harborages and cannot be applied to sensitive environments 9 .
Biorepellents and botanical insecticides derived from plants may provide potential alternatives for household and medically important pest insects because plants are a potential source of bioactive secondary metabolites that are perceived by the general public as relatively safe and pose fewer risks to the environment with minimal impacts on human health [15][16][17][18] . Many efforts have been focused on plants as potential sources of commercial repellents, in part, because certain plant preparations meet the criteria of minimum-risk repellents 19 . Previous studies have shown that the methanol extract and essential oil from the rhizome of coco-grass, Cyperus rotundus L. (Poales: Cyperaceae), possessed good repellency against male B. germanica. Limited information is available concerning the potential uses of the C. rotundus rhizome and its constituents for the control of B. germanica, despite the insecticidal properties of the plant 20 .
In this study, our aim was to assess whether the two monoterpenoids, 1,8-cineole and zerumbone (ZER), which were extracted from C. rotundus rhizomes, and another 11 previously identified constituents without insecticidal activity from the rhizome essential oil 20 had repellent activity against male B. germanica compared to deet, a positive control, in a filter-paper choice bioassay. Deet is registered as an insect repellent in South Korea 21 . The repellency of binary mixtures of ZER and seven active compounds was also evaluated. Finally, the repellency of α-humulene was examined to discern whether the α,β-unsaturated carbonyl group of ZER is essential for repellant activity. The relationship between repellency and boiling points (BPs) of the test compounds is also discussed.

Results
Bioassay-guided fractionation and isolation. The fractions obtained from solvent partitioning of the methanol extract of the C. rotundus rhizome were tested against male B. germanica using a filter-paper choice assay (Table 1). Significant differences in repellency were observed among the fractions and were used to identify the peak activity fractions for the next step in purification. At a concentration of 244.6 μg cm −2 , the hexane-soluble fraction was the most potent repellent, while no repellency was obtained using the chloroform-, ethyl acetateor water-soluble fractions.
The repellency of the eight selected active compounds and deet at 652.4 μg cm −2 against male B. germanica was evaluated using Ebeling choice boxes ( Table 3). The repellency was significantly different among treatments. ZER resulted in complete repellency at the 10 h post-treatment interval, but after that peak, the cockroach repellency of this compound decreased considerably over time. Overall, all these compounds were less potent repellents than deet.
Repellency of binary mixtures. The differences in repellency after the application of the seven binary mixtures were significant ( Table 4). The repellency of the binary mixtures of ZER and 1,8-cineole, (+)-dihydrocarvone or (R)-(+)-limonene at the 70:30, 50:50 and 30:70 ratios was between 94 and 90%, between 91 and 85% and between 82 and 79% at 24 h post-treatment, respectively. The mixtures with the 70:30 ratio were significantly more effective than either the corresponding compound or deet alone. The repellency of the other four binary mixtures at the 70:30, 50:50 and 30:70 ratios was between 83 and 80%, between 79 and 73% and between 73 and 68%, respectively.
Behavioral response. Antennectomy of male B. germanica resulted in a diminished response to the five tested compounds and deet at 652.4 μg cm −2 in the Ebeling choice boxes (Table 5). There was a significant difference (P < 0.0001) in the repellency of ZER to intact and antennectomized male B. germanica. Similar differences in the response of male cockroaches to (+)-dihydrocarvone, β-caryophyllene, (1S)-(−)-verbenone, 1,8-cineole and deet in intact versus antennectomized treatments were also observed.

Discussion
In this study, the repellency of 1,8-cineole and ZER from the C. rotundus rhizome along with 11 previously identified constituents and binary mixtures of ZER and seven active compounds against male B. germanica was assessed. The repellency of these materials was compared to deet, which is a currently available synthetic repellent, to determine whether they would be suitable for future commercial cockroach repellents. Historically, C.  Table 2. Repellency of nine pure organic compounds and a commercial repellent, deet, against male Blattella germanica using a filter-paper choice assay. Means followed by the same lowercase letter in the same column or the same uppercase letter in the same row are not significantly different (P = 0.05, Bonferroni method).
rotundus has been widely used as an analgesic, a sedative, an antiplasmodic agent, an antimalarial and a treatment for stomach disorders 24,25 . Many plant extracts and essential oils produce repellency to different insect species [26][27][28] . Repellent constituents derived from plants against cockroaches include flavonoids 29 , phenylpropanoids 30 and terpenoids [31][32][33] . For example, the monoterpenoids, E,Z-and Z,E-nepetalactone as well as pulegol, isopulegol and pulegone work against B. germanica, and the phenylpropanoids, safrole and isosafrole, work against Periplaneta americana (L.) (Blattodea: Blattidae) and have been reported as repellent compounds [30][31][32] . In the current study, the repellent constituents of the C. rotundus rhizome were determined to be the sesquiterpenoid ZER and the monoterpenoid 1,8-cineole. Of the tested compounds, ZER was the most potent repellent. Potent repellency was also produced by β-caryophyllene, (+)-dihydrocarvone, (R)-(+)-limonene, (S)-(−)-limonene, (−)-(E)-pinocarveol and (1S)-(−)-verbenone. This study is the first report of the repellency of ZER and pinocarveol against B. germanica, although 1,8-cineole has been reported to have repellency against B. germanica nymphs 34 as well as contact and fumigant toxicity and ovicidal and feeding-deterrent activity against Tribolium castanaeum (Herbst) (Coleoptera: Tenebrionidae) 35 and fumigant toxicity against three major stored-grain insects 36 .    Volatile constituents derived from most plant extracts and essential oils consist of alkanes, alcohols, aldehydes and terpenoids (particularly monoterpenoids) 34,37 and some are insect repellents 38 . Because of their high volatility, the essential oils and their constituents are usually effective against arthropods only for a relatively short period, which is typically less than 2 h [38][39][40] . However, available information on B. germanica repellency is limited because most previous studies that investigated repellency used a bioassay with a short time period, which was mainly 5 to 10 min. The compounds 1,8-cineole and α-pinene have been reported to have 92 and 82% repellency, respectively, against B. germanica nymphs after a 2 h post-treatment at a concentration of 5 ppm 33 . However, Appel et al. 41 reported that mint oil extract at the surface concentration of 2 mL of mint oil to 457.5 cm 2 was extremely repellent (approximately 100%) to adult males of B. germanica and P. americana for 14 days.
In the current study with Ebeling choice box tests, (−)-(E)-pinocarveol and ZER were very effective against male B. germanica for 8-10 h, while (+)-dihydrocarvone, β-caryophyllene and (1S)-(−)-verbenone were very effective for 4 h. Nevertheless, these compounds were less potent repellents than deet. The differences in repellency could be attributed to the differences in the quantitative losses because of the different volatility of the compounds, as described by Brown and Hebert 42 . The ability of a chemical vapor to function as a repellent has been reported to be related to its BP, and BPs between 230 and 260 °C at atmospheric pressure were found to be the most desirable for an effective repellent 42 . The BPs of (E)-carveol and (−)-α-copaene coincide with the optimal range but were less effective than either ZER with higher BPs, compounds with lower BPs or compounds with similar BPs. Thus, other factors, such as structural characteristics (e.g., types of functional groups and carbon skeleton), rather than BP appear to be associated with repellency of these compounds against male B. germanica. For example, β-caryophyllene was more effective than caryophyllene oxide. Interestingly, α-humulene, which lacks only the α,β-unsaturated carbonyl group present in ZER, was virtually ineffective. This finding indicates that the α,β-unsaturated carbonyl group of ZER is a prerequisite component for repellency to B. germanica. Peterson et al. 31 studied the repellency of catnip, Nepeta cataria L. (Lamiales: Lamiaceae), as well as its essential oil constituents, against male B. germanica. They reported that the iridoid monoterpenoid, E,Z-nepetalactone, was significantly more active than the Z,E-isomer. In addition, ZER, (+)-dihydrocarvone, β-caryophyllene, (1S)-(−)-verbenone, 1,8-cineole and deet produced complete repellency against intact male B. germanica, while these compounds exhibited 35-47% repellency against antennectomized male B. germanica. These results indicate that antennae are partly responsible for the perception of the tested compounds. Detailed tests are needed to fully understand their modes of repellent action.
It is well known that repellency against various insect species was more pronounced in binary mixtures of phytochemicals compared to the corresponding single compounds. For example, a significantly enhanced repellency was produced through binary mixtures of E,Z-and Z,E-nepetalactone that were consistently more repellent than the isomers when they were tested alone against Anopheles gambiae Giles (Diptera: Culicidae), expect for equivalent or near-equivalent mixtures, which had significantly lower repellency 43 . In the current study, repellency against male B. germanica produced by the binary mixtures of ZER and 1,8-cineole, (+)-dihydrocarvone or (R)-(+)-limonene at three mixture ratios were more pronounced compared to repellency produced by the corresponding compound and deet alone. The improved effectiveness of repellency could be attributed to the lower evaporation rate and/or better persistence of ZER in the combined presence of another compound, as described by Khan et al. 44 and Tuetun et al. 45 . The optimum ZER content was determined to be more than 50% according to our laboratory results. This original finding indicates that binary mixtures could be promising repellant products that are novel and effective against B. germanica. However, these terpenoids are volatile, and the volatility problem may be alleviated by special formulations, such as microencapsulation, that can reduce volatile loss, as described by Peterson et al. 31 .
In conclusion, C. rotundus rhizome-derived products containing active compounds, especially zerumbone and 1,8-cineole, could be useful as repellents in the control of B. germanica populations in sensitive environments in which conventional insecticides would be inappropriate, as long as special formulations (e.g., microencapsulation) that facilitate the slow release of active compounds can be selected or developed. Further research is needed for the practical applications of plant-derived preparations as novel cockroach repellent products to establish their safety profiles in humans. In addition, their effects on non-target organisms and the environment need to be established. Finally, detailed tests are needed to understand how to improve repellency potency and stability for eventual commercial development.

Methods
Instrumental analysis. The 1 H and 13 C NMR spectra were recorded in CD 3 OD on an Avance 400 WB spectrometer (Bruker, Rheinstetten, Germany) at 400 and 100 MHz, respectively, using tetramethylsilane as an internal standard. The chemical shifts are given in δ (ppm). The DEPT spectra were acquired using Bruker software. The mass spectra were obtained on a GCMS-QP 2010 Ultra gas chromatograph-mass spectrometer (Shimadzu, Columbia, MD, USA). Silica gel 60 (0.063-0.2 mm) (Merck, Darmstadt, Germany) was used for column chromatography. Merck pre-coated silica gel plates (Kieselgel 60 F 254 ) were used for analytical thin layer chromatography (TLC). A Spectra System P 2000 high-performance liquid chromatograph (HPLC) (Thermo Scientific, Waltham, MA, USA) was used to isolate the active compounds.
Chemicals. Two constituents, 1,8-cineole and ZER, were identified in this study, and another 11 previously identified constituents from the C. rotundus rhizome essential oil 21 are listed in Table 6 along with their BPs and purities. α-Humulene, which is a zerumbone analog lacking the α,β-unsaturated carbonyl group, was also used in this study for structure-activity relationship (Fig. 1). All compounds were purchased from Sigma-Aldrich (St. Louis, MO, USA). For the relationship between toxicity and BPs of the tested compounds, the BP values of these compounds were obtained from ACD/ChemSketch (ACD/LAB 12.0 for Microsoft Window, Advanced Chemistry Development, Inc., Montreal, Canada) ( Table 1). Deet (97.0% purity) was purchased from Sigma-Aldrich. All of the other chemicals used in this study were reagent-grade quality and are available commercially.
Cockroaches. The stock cultures of B. germanica (susceptible KSS strain) 20 were maintained in the laboratory without exposure to any known insecticide. Cockroaches were reared in glass jars (43-cm diameter × 30 cm) containing Purina calf food pellets (Pyeongtaek, Gyeonggi, South Korea), distilled water and a cardboard shelter at 27 ± 2 °C and 60 ± 5% relative humidity (RH) under a 12:12 h light:dark cycle. Because male B. germanica are more sensitive than females to olfactory stimuli 31,46 , adult males were used for the repellency bioassays. Bioassay-guided fractionation and isolation. Air-dried rhizomes (600 g) of C. rotundus were pulverized, extracted with methanol (3 × 3 L) at room temperature for 2 days and filtered. The combined filtrate was concentrated by rotary evaporation at 40 °C to yield approximately 152 g of a dark greenish tar. The extract (100 g) was sequentially partitioned into hexane-(16.4 g), chloroform-(45.3 g), ethyl acetate-(10.7 g) and water-soluble (27.6 g) portions for the subsequent bioassays. The organic solvent-soluble portions were concentrated under vacuum at 40 °C, and the water-soluble portion was concentrated at 50 °C. To isolate the active constituents, 244.6 μg cm −2 of each C. rotundus rhizome-derived fraction were tested in a filter-paper choice bioassay, as described by Petterson et al. 31 .
The hexane-soluble fraction (15 g) was the most biologically active fraction (Table 1) and was chromatographed on a 5.5 × 70 cm silica gel (600 g) column through elution with a gradient of hexane and ethyl acetate Filter-paper choice assay. A filter-paper choice test described previously by Petterson et al. 31 was used to evaluate the repellency of all compounds against male B. germanica (7-10 days old). Each test was conducted between 20:00 to 23:00 h with overhead florescent lighting at 27 ± 2 °C and 60 ± 5% RH in 5-min test periods, as described previously 31   ) of each test compound in 1 mL methanol, and the other side was treated with 1 mL methanol. Deet served as the positive control and was used in a similar manner. After drying the samples in a fume hood for 1 min, each treated paper was placed on the bottom section of a disposable Petri dish (15-cm diameter × 2 cm). The lid of the dish had a 1 cm hole cut in the center for introducing the cockroach directly into the center of the arena. One insect at a time was introduced. The hole was then blocked by using a small piece of tissue paper to prevent the cockroaches from escaping. Immediately after the introduction of the insect, the number of seconds it spent on the treated or untreated side within 300 s was timed with two stopwatches. If a compound produced ≥40% repellency at a given concentration, further bioassays were conducted. Filter papers and cockroaches were used once then discarded. Each trial was repeated 10 times.
To determine an effective mixture ratio for ZER and another seven active compounds, the repellency of binary mixtures at five tested ratios (100:0, 70:30, 50:50, 30:70 and 0:100 by weight) was tested at a concentration of 81.5 μg cm −2 at the 24 h post-treatment, which was based on the preliminary test results. Deet served as the positive control. All treatments were replicated 10 times.
Ebeling choice box assay. The residual repellency of the eight selected compounds to male B. germanica was tested in Ebeling choice boxes 47 as described previously by Appel et al. 41 , with a slight modification. A choice test was carried out using a series of two connected acrylic boxes (each 5.5 × 11 × 11 cm). The top side of each compartment of the choice box had nine holes (0.5-mm diameter) on the surface to permit airflow. Except for the top side of the dark compartment of the choice box, the inner surface of other five walls was evenly painted with 652.4 μg cm −2 of the test compounds in 2 mL of methanol using a brush, and the amount was based on the preliminary test results. Food and water were placed on the bottom section of the dark compartment of the choice box. Deet served as a positive control and was similarly formulated. Negative controls (i.e., no test material or repellent) consisted of 2 mL methanol only. Treatments were allocated randomly to the choice boxes. After drying the samples in a fume hood for 3 min, 20 adult male B. germanica were released into the untreated compartment of the choice box and were allowed to enter the treated compartment for 24 h. Cockroaches were able to move freely between the dark (treated) and the lighted (untreated) compartments through two holes (1-cm diameter) in the partition separating the sides. Choice boxes had the same environmental conditions as those used for colony maintenance. Banks of white fluorescent lights were 1.6 m above the choice boxes and produced a light intensity in the untreated compartment of 300-350 lux using an INS Digital Lux Meter (Markson Scientific, Phoenix, AZ, USA). The number of B. germanica in each compartment was recorded at every 15 min interval until 24 h post-treatment. Six replicates were used for each treatment in a completely randomized design.
For the tests that used antennectomized male B. germanica, a razor blade was used to remove the antennae at the scape, as described by Petterson et al. 31 . The cockroaches were allowed to recover from the procedure for 24 h before they were exposed to 652.4 μg cm −2 of each compound and deet in 2 mL methanol according to the choice box method above. Data analysis. The repellent index was calculated according to the following formula: % repellency for filter-paper choice assay = [(Tu − Tt)/Tn] × 100, where Tu is the number of seconds a cockroach spent on the untreated side, Tt is the number of seconds a cockroach spent on the treated side, and Tn is the total number of Figure 2. Procedures to isolate the repellent constituents. The Cyperus rotundus rhizome methanol extract was sequentially partitioned into hexane-, chloroform-, ethyl acetate-and water-soluble portions. The hexanesoluble fraction was the most biologically active fraction and high-performance liquid chromatography was performed. Each fraction (244.6 μg cm −2 ) was tested in a filter-paper choice bioassay to isolate the active compounds from the fraction. seconds (300) 31 ; % repellency for the Ebeling choice box assay = 100 − [(Ta/Tb) × 100], where Ta is the number of cockroaches in the treated group and Tb is the number of total cockroaches that were tested 41 . The percentages of repellency were transformed to arcsine square-root values for analysis of variance (ANOVA). The Bonferroni multiple-comparison method was used to test for significant differences among treatments 48 . A Student's t-test was used to test for significant differences between the two treatment methods 48 . A compound with <10% repellency was considered to be ineffective. Means ± standard errors (SEs) of untransformed data are reported.