# Journal of Medical Entomology

### Editor-in-Chief

William K. Reisen, PhD

1.712
1.944
22 out of 94

Published on behalf of

## Dihydronepetalactones Deter Feeding Activity by Mosquitoes, Stable Flies, and Deer Ticks

John E. Feaster, Mark A. Scialdone, Robin G. Todd, Yamaira I. Gonzalez, Joseph P. Foster, David L. Hallahan
832-840 First published online: 1 July 2009

## Abstract

The essential oil of catmint, Nepeta cataria L., contains nepetalactones, that, on hydrogenation, yield the corresponding dihydronepetalactone (DHN) diastereomers. The DHN diastereomer (4R,4aR,7S,7aS)-4,7-dimethylhexahydrocyclopenta[c]pyran-1(3H)-one, DHN 1) was evaluated as mosquito repellent, as was the mixture of diastereomers {mostly (4S,4aR,7S,7aR)-4,7-dimethylhexahydrocyclopenta[c]pyran-1(3H)-one, DHN 2} present after hydrogenation of catmint oil itself. The repellency of these materials to Aedes aegypti L. and Anopheles albimanus Wiedemann mosquitoes was tested in vitro and found to be comparable to that obtained with the well-known insect repellent active ingredient N,N-diethyl-3-methylbenzamide (DEET). DHN 1 and DHN 2 also repelled the stable fly, Stomoxys calcitrans L., in this study. DHN 1, DHN 2, and p-menthane-3,8-diol (PMD), another natural monoterpenoid repellent, gave comparable levels of repellency against An. albimanus and S. calcitrans. Laboratory testing of DHN 1 and DHN 2 using human subjects with An. albimanus mosquitoes was carried out. Both DHN 1 and DHN 2 at 10% (wt:vol) conferred complete protection from bites for significant periods of time (3.5 and 5 h, respectively), with DHN2 conferring protection statistically equivalent to DEET. The DHN 1 and DHN 2 diastereomers were also efficaceous against black-legged tick (Ixodes scapularis Say) nymphs.

• Nepeta cataria
• catnip
• nepetalactone
• iridoid
• repellent

There is a need for natural insect repellents that can substitute for existing synthetic actives (Peterson and Coats 2001, Krajick 2006, Moore et al. 2006). Many plant species produce essential oils that are used as natural sources of insect repellent and fragrant chemicals (Hay and Waterman 1993, Isman 2006). These oils are rich in volatile compounds such as monoterpenoids, with which repellency has generally been associated. Citronella oil, known for its insect repellency, is obtained from the graminaceous plants Cymbopogon winterianus Jowitt and C. nardus L. The active ingredients in this oil have been reported to be the acyclic monoterpenoids citronellal, citronellol, and geraniol (Moore et al. 2006). However, repellency of citronella has often been found to be relatively short-lived (Lindsay et al. 1996, Fradin and Day 2002, Barnard and Xue 2004, Moore et al. 2006). The repellent p-menthane-3,8-diol (PMD) is a monoterpenoid prepared from oil of the lemon-scented gum tree Corymbia citriodora (Hook) Hill and Johnson, which has shown efficacy as an insect repellent against mosquitoes, the biting midge Culicoides impunctatus Goetghebeur, and Ixodes ricinus L. ticks (Trigg 1996a, b; Trigg and Hill 1996; Barnard et al. 2002; Barnard and Xue 2004). However, eucalyptus oil was shown to be attractive to the biting midge Culicoides imicola Kieffer (Braverman et al. 1999).

Certain plants of the genus Nepeta (catmints) produce an essential oil rich in a class of monoterpenoid compounds known as iridoids (Inouye 1991). Iridoid monoterpenoids have long been known for their effects on insect behavior. Certain iridoids are known attractants, for example, acting as aphid sex pheromones (Dawson et al. 1996, Birkett and Pickett 2003). The iridoids of interest here are the methylcyclopentanoid nepetalactones (Fig. 1) (Regnier et al. 1967b, Cavill 1969). Nepetalactone was shown to be repellent to some insect species (Eisner 1964, 1995), although until recently this compound had not been tested against any of the common insect pests of human society. Several recent papers have reported repellency of nepetalactone toward cockroaches, termites, houseflies, and mosquitoes (Peterson et al. 2002; Coats et al. 2003a, b; Peterson and Ems-Wilson 2003; Schultz et al. 2006). Catmint oil exhibited activity indicative of spatial repellency against Aedes aegypti L. (Bernier et al. 2005). It also exhibited efficacy in topical repellency against Ae. aegypti, Anopheles albimanus Wiedemann, and Anopheles quadrimaculatus Say, although DEET was shown to be more effective (Bernier et al. 2005). In another study, catmint oil and nepetalactones were compared with DEET and found to be significantly less effective in deterring the biting of Ae. aegypti mosquitoes (Chauhan et al. 2005).

Fig. 1

Chemical structures of nepetalactone and dihydronepetalactone stereoisomers referred to in the text.

Dihydronepetalactone (DHN), the reduced form of nepetalactone (Fig. 1), has been reported as a minor component of the nepetalactone-rich oils of certain Nepeta spp. including N. cataria L. (Sakan et al. 1965, Regnier et al. 1967a). Catalytic hydrogenation of catmint oil containing a mixture of nepetalactone diastereomers enriches the oil with a mixture of DHN diastereomers (Regnier et al. 1967a). Isodihydronepetalactone (4R,4aR,7S,7aR)-4,7-dimethylhexahydrocyclopenta[c]pyran-1(3H)-one; DHN 3 in Fig. 1) is present in the defensive secretions of certain ants (Cavill and Clark 1967, Cavill et al. 1982). Jefson et al. (1983) described the repellent effect of a dihydronepetalactone diastereomer toward the ant species Monomorium destructor. In this case, a 7-(R)-configured DHN was tested. In this paper, we report on the repellency of DHN diastereomers as pure compounds and as mixtures that are formed from the hydrogenation of catmint oil. This hydrogenation produces a fragrant mixture enriched with DHN diastereomers that (because they are derived from the [7S]-nepetalactones of the oil) have an absolute stereochemical configuration of (S) at carbon 7, the cyclopentyl ring carbon that bears the methyl group. These materials show repellent activity toward mosquitoes, stable flies, and ticks in laboratory tests.

## Materials and Methods

### Test Materials.

Catmint oil (Berjé, Bloomfield, NJ) was partially purified by fractional distillation. Analysis of the fractionated oil was carried out by combined gas chromatography mass spectroscopy (GC-MS) and nuclear magnetic resonance spectroscopy (NMR). For GC-MS, a Hewlett-Packard (now Agilent; Santa Clara, CA) 5890 gas chromatograph was coupled to a Hewlett-Packard 5972A benchtop mass spectrometer using a heated transfer line. Chromatographic separation was effected using a HP5-MS (25 m by 0.2 mm) column (Hewlett-Packard), and a GC oven programmed from 120 (2-min hold) to 210°C (5-min hold) at 15°C/min. The carrier gas was He (1 ml/min). Qualitative GC-MS analysis showed the presence of three principal peaks with mass spectra characteristic of methylcyclopentanoids (Clark et al. 1997). 13C NMR analysis confirmed that these principal constituents were nepetalactone stereoisomers, and quantitation based on the carbonyl peak around 170 ppm indicated of the presence of 84.5 mol% Z,E-nepetalactone, 14.3 mol% E,Z-nepetalactone, and 1.2 mol% E,E-nepetalactone.

To convert the nepetalactones to dihydronepetalactones, hydrogenation of this material was carried out in ethanol at 30 psig and room temperature for 48 h using Pd/SrCO3 (41,461–1; Aldrich, Milwaukee, WI). The reaction mixture was filtered over Celite (Sigma-Aldrich) to remove catalyst, and the solvent was removed under vacuum, yielding a clear, pale yellow oil. Quantitative analysis of dihydronepetalactones was conducted by GC-MS, using 1,3-dibromobenzene as an internal standard and a column calibrated with authentic standards. This indicated that the principal component (66.58 wt%) represented a dihydronepetalactone isomer, with m/z 168 and a characteristic fragment ion of m/z 113 in the mass spectrum. Based on the analysis of the 1H NMR spectrum of the product, this principal diastereomer was DHN 2, (4S,4aR,7S,7aR)-4,7-dimethylhexahydrocyclopenta[c]pyran-1(3H)-one, as shown in Fig. 1. The minor DHN diastereomers could be similarly identified as DHN 1 [(4R,4aR,7S,7aS)-4,7-dimethylhexahydrocyclopenta[c]pyran-1(3H)-one; 7.64 wt%] and DHN 3 [(4R,4aR,7S,7aR)-4,7-dimethylhexahydrocyclopenta[c]pyran-1(3H)-one, isodihydronepetalactone; 2.75 wt%]. A small amount (3.4 wt%) of nepetalic acid was present in the hydrogenated material, together with traces of unreacted nepetalactones (0.12 wt%) and caryophyllenes (0.81 wt%). The remaining constituents (constituting ≈20 wt%) were not identified.

For preparation of E,Z-nepetalactone, a number of plants were grown from seeds of the catmint Nepeta racemosa Lam. (Chiltern Seeds, Cumbria, United Kingdom). The GC analysis of extracts of leaves from individual plants allowed identification of plants producing mainly E,Z-nepetalactone in their oils (Clark et al. 1997). These plants were selected and grown to maturity. Leaf material from these plants was harvested, freeze-dried, and extracted into ethyl acetate (Sigma-Aldrich), and the extracts were concentrated. Purification of E,Z-nepetalactone from the concentrated extract was accomplished by silica gel chromatography in hexane/ethyl acetate (9:1; Sigma-Aldrich), followed by preparative thin-layer chromatography on silica using the same solvent mixture. After removal of the solvent and redissolving in hexane, the E,Z-nepetalactone was crystallized over dry ice. Analysis by GC-MS and NMR (1H and 13C) confirmed the identity of the crystalline material as E,Z-nepetalactone.

Hydrogenation of the E,Z-nepetalactone was carried out in ethanol (Sigma-Aldrich) using ESCAT#142 catalyst (Engelhard; now BASF Catalysts, Iselin, NJ) at 50°C for 4 h. Examination of the products by GC-MS and NMR (1H and 13C) confirmed that the E,Z-nepetalactone had been quantitatively converted to the corresponding dihydronepetalactones, with one in significant excess. Results from NMR analysis of the major diastereomer are as follows: 1H NMR (500 MHz, CDCl3): d 0.97 (d, 3H, J = 6.28 Hz), 0.98 (d, 3H, J = 6.94 Hz) d 1.24 (m, 2H), 1.74 (m, 1H), 1.77 (m, 2H), 1.99 (m, 2H), 2.12 (dd, 1H, J = 6.86 and 13.2 Hz), 2.51 (m, 1H), 3.78 (tr, 1H, J = 11.1 Hz), 4.33 (dd, 1H, J = 5.73 and 11.32 Hz); 13C (500 MHz, CDCl3): d 15.43, 18.09, 27.95, 30.81, 31.58, 35.70, 42.51, 51.40, 76.18, 172.03. Based on the observed 1H NMR spectrum of the DHN product obtained, it was concluded that the diastereomer was DHN 1, (4R,4aR,7S,7aS)-4,7-dimethylhexahydrocyclopenta[c]pyran-1(3H)-one.

### Arthropods.

Two species of mosquitoes, Aedes aegypti L. (the yellow-fever mosquito, colony a mixture of Fort Detrick and Rockefeller strains, obtained ≈1960) and Anopheles albimanus Wiedemann (a malaria vector found in the tropics, colony obtained in 1992 from Johns Hopkins and from Walter Reed Army Institute of Research), and the muscid fly Stomoxys calcitrans L. (stable fly, colony obtained from USDA Kerrville, TX, and the University of Nebraska in 1991) were obtained from ICR’s (formerly Insect Control & Research) laboratory colonies. The mosquitoes were reared in an insectary at ≈27°C and ≈80% RH with a 12-h photoperiod. The adults were provided access to rabbit blood once a week for 20–30 min. They were allowed to develop and lay their eggs. These eggs were hatched, and the resulting first-instar larvae were placed in plastic trays (24 by 16 by 4 in deep) filled with water to ≈7.5 cm. The larvae were fed daily with yeast and later with liver powder until they pupated. The pupae were separated by sex, and female pupae were transferred to bowls (6.5 in diameter by 5 in deep) and allowed to emerge as adults. S. calcitrans adults were fed only on cotton balls soaked with citrated bovine blood (flies 0-5 d old were provided this food source for 9–15 min/d, 6 d/wk, flies 6 d and older were provided this source at all times). When eggs were laid, they were transferred to a medium (Lab Diet Fly Larvae Media; PMI Nutritional, Henderson, CO) in 1-gal glass jars in which they hatched and completed their larval development to pupae (densities were not recorded). The pupae were retrieved and allowed to emerge as adults (of both sexes) in screen cages (12 by 12 by 12 in). Flies aged 3-8 d of both sexes were used in the test; they had not received any blood meals over the previous 24 h.

Nymphal black-legged ticks (Ixodes scapularis Say) were obtained from a colony maintained by Dr. Tom Mather, University of Rhode Island. These ticks were known to be free of Borrelia burgdorferi (the spirochaete that causes Lyme disease), but the presence or absence of other pathogens carried by this tick species was unknown.

### In Vitro Repellency Assays.

Repellency tests were carried out in an enclosed chamber containing five wells, each covered by a Baudruche membrane (Joseph Long, Belleville, NJ) treated with either repellent or solvent alone as a negative control (in all cases, isopropyl alcohol [IPA]; Sigma-Aldrich) (Rutledge et al. 1976, Rutledge and Gupta 2004). The wells were filled with citrated bovine blood, ATP (Sigma-Aldrich) was added (72 mg ATP disodium salt per 26 ml of blood), and the wells were maintained at 37 ± 1°C using a circulating water bath. A volume of 25 μl IPA, containing candidate repellents at various concentrations (see tables for concentrations tested), was applied to each membrane. The wells were inserted beneath the cage. Insects (250 Ae. aegypti, 100 An. albimanus, or 100 S. calcitrans) were released into the chamber, and a sliding cover was removed from above the wells, exposing them to the insects. Probes (for mosquitoes) or landings (for stable flies) on each of the wells were tallied for 20 min at 2-min intervals. No attempt was made to prevent insects from landing or probing more than once as this was impractical. This procedure was replicated for a total of five batches of insects, five sets of treated membranes, and five aliquots of blood. The positions of the treated membranes on the wells were changed for each replicate so as to avoid bias; all five variables were tested at each position (the only exception was one test in which there were only three replicates, Table 1). The times over which these test were run ranged from 0900 to 1400 hours EST, with the photoperiod beginning at 0600 hours.

View this table:

Percentage reduction in landings or probes was calculated for each treatment at each observation time using the following equation: $$mathtex$$\% \ reduction = 100 - [(T/C)\times 100$$mathtex$$

where T was the mean number of mosquitoes probing (or stable flies landing) on a treated well for that replicate at time tx, and C was the mean number of insects landing or probing on the IPA control well at time tx. These resulting percentages were then arcsine transformed (this normalized the data in some cases), and analyses of variance (ANOVAs) were conducted (CoStat; CoHort, Monterey, CA) using the calculated repellency. Two approaches were considered for the analysis. The first calculated repellency at each 2-min period across all five replicates (n = 50). The second calculated repellency from the total number of landings for all 10 2-min periods per replicate (n = 5). The latter approach is used in this paper. Multiple comparisons of means were conducted using the Duncan’s test (when numbers of replicates per treatment were equal) and Student-Newman-Keuls test (when these numbers were not equal). Differences were assessed at the 0.05% level of significance. Repellencies are reported in the text and tables before arcsine transformation.

### In Vivo Repellency Assays.

All test subjects provided informed consent, and test protocols were approved by the internal DuPont Human Studies Review Board and an external Institutional Review Board (Essex IRB, Lebanon, NJ). For repellency tests with mosquitoes, the test cages (2 by 2 by 2 ft) had two sleeved entry ports on each of two opposite sides and a hand rest in the center. The sides and top were screened, and the floor was equipped with a mirror to facilitate observations. Two hundred adult female An. albimanus (3-8 d old), which had never received a bloodmeal and which had been deprived of their normal diet of 10% sucrose 24 h before the experiments, were released into the test cage. Each subject was prequalified as attractive through having an arbitrary 10 mosquito landings on their untreated forearms within 30 s of arm insertion into the cage. This also prequalified the mosquitoes. Dilutions in IPA of DHN, DHN 2, and DEET were tested over 2 d: at 10% (wt:vol) in the first test and 5% (wt:vol) in the second test. A volume of 1.0 ml of each compound (either 5 or 10% [wt:vol] in IPA) was applied to 250 cm2 (i.e., 0.4 mg AI/cm2) areas on the forearms of six subjects, with each subject having one compound applied to each forearm. After allowing the applied solutions to dry for 30 min, the subject’s forearms were placed into the test cage for 5-min periods (long enough for any mosquito ready to bite to do so) at 30-min intervals, and the number of mosquitoes probing or biting during each exposure period was recorded. Complete protection time (CPT) was recorded for each repellent on each subject. The CPT was the time elapsed between treatment and when the first confirmed bite occurred. The first confirmed bite was defined as a bite that was followed by a second bite either within the same or the next exposure period (Barnard et al. 2006).

For testing for tick repellency, a volume of 25 μl of test compound (30% [wt:vol] in IPA) was applied within 4-cm-diameter circles drawn on the left forearms of six human subjects (i.e., 0.6 mg [AI]/cm2). Each subject had two repellents applied individually within two circles on this forearm; a single 4-cm-diameter circle drawn on the other arm was left untreated to act as a control for tick attractiveness. Laboratory-reared unfed nymphs of I. scapularis were brought within 1 mm of the untreated circles on cotton swabs. If normal questing behavior was observed and the tick crawled onto the untreated area, it was deemed qualified and presented to a treated area. A qualified tick that quested at, or crawled onto, the treated area within 1 min was recorded as having not been repelled. A qualified tick that did not quest or ceased questing within 1 min and/or retreated from the treated area was recorded as repelled. Additionally, a qualified tick that crawled onto the treated area, but fell off within an additional 1 min, was recorded as repelled. Each subject had five qualified ticks offered to each treated circle at hourly intervals. Exposures continued until three of any group of five offered ticks were not repelled. Repellent failure time was the time when the first of at least three of a set of five ticks was not repelled. This was used as a measure for comparing treatments not as an acceptable threshold for repellency. Each treatment was tested on four subjects. Ticks were only used once after which they were killed by immersion in IPA.

## Results

Initial screening tests for repellency of DHN were conducted using an in vitro blood feeding test system (Rutledge et al. 1976, Rutledge and Gupta 2004). A mixture of dihydronepetalactone diastereomers was prepared by hydrogenation of catmint oil enriched in Z,E-nepetalactone. This material, consisting predominantly of DHN 2 and referred to thus for clarity, was evaluated for its repellent effect against female Ae. aegypti mosquitoes. Table 1 shows the effect of repellent concentration on the time taken before the mosquitoes first probed each membrane. Mosquitoes began probing negative control wells at an average of 4.7 min, but 12 min elapsed before they probed the 1% (wt:vol) DEET-treated membrane and ≈19 min before the 5% (wt:vol) DHN 2-treated membrane. Membranes treated with DHN 2 at 1 and 2.5% (wt:vol) allowed first probes at an average of 8 and 9.3 min, respectively. Probing of membranes treated with dihydronepetalactones at each observation time by Ae. aegypti is shown graphically in Fig. 2. There were only 0.03 probes per observation period on the membranes treated with DHN 2 at 5% and only 0.47 probes per observation period on the 1% DEET-treated membranes. More probes occurred on membranes treated with the two lower concentrations of DHN 2 (1 and 2.5% wt:vol); 1.40 and 0.63, respectively (Table 1). These data show that at all concentrations tested, DHN2 significantly reduced probes with respect to the untreated control.

Fig. 2

Distribution of probing over time, during tests of DHN 2 (as the predominant DHN in hydrogenated catmint oil) and DEET against Ae. aegypti mosquitoes. The negative control is IPA.

The dihydronepetalactone diastereomer produced by hydrogenation of purified E,Z-nepetalactone, DHN 1, was also tested in vitro for repellency against Ae. aegypti mosquitoes, and the results are shown in Table 2. DHN 1 at 1% (wt:vol) was found to discourage mosquito "first probing" for an average of 16 min. This result is marginally better than, but statistically indistinguishable from, DEET at the same concentration in this test, for which the mean time to first probe was 14.8 min. Lower concentrations of DHN 1 (0.5 and 0.2% wt:vol) inhibited first probing for an average of 9.6 and 8.4 min, respectively. The incidence of probing by female Ae. aegypti over time is shown in Fig. 3. Throughout the 20 min of exposures, fewer probes occurred on membranes treated with the diastereomer DHN 1 at 1.0% (wt:vol) or 1% DEET than on the control membranes. Intermediate numbers of probes occurred on the membranes treated with 0.5 or 0.2% DHN 1. These data show that, at all three concentrations tested, treatment with DHN 1 significantly reduced probing on the membranes as compared with the IPA treatment. There was also a relationship between DHN concentration and the time taken for the first mosquito to begin probing a treated membrane. Overall, the data indicate that DHN 1, derived from hydrogenation of E,Z-nepetalactone, is a significant deterrent to probing by Ae. aegypti in this test.

View this table:
Fig. 3

Distribution of probing over time, during tests of DHN 1 and DEET against Ae. aegypti mosquitoes. The negative control is IPA.

DHN 1 and DHN 2 were similarly tested for repellency of adult female An. albimanus mosquitoes. In these experiments, an additional positive control compound was included, namely p-menthane-3,8-diol (PMD). The incidence of probing by An. albimanus mosquitoes on membranes treated with dihydronepetalactone stereoisomers was analyzed at each observation period, as shown in Fig. 4. Probing started on exposure of the insects to the test wells and gradually increased in intensity over time. The numbers of probes of membranes treated with DHNs were fewer than on the negative control membranes at each observation time. The DHN diastereomers showed repellency to An. albimanus throughout the experiment.

Fig. 4

Distribution of probing over time, during tests of various repellents against An. albimanus mosquitoes. Pure DHN 1 and hydrogenated catmint oil (containing predominantly DHN 2) was tested. Negative controls used IPA; PMD and DEET solutions were tested for comparison. All repellents were 1% (wt:vol) in IPA.

The results are presented in Table 3. The data indicate that at all four treatments were modestly effective in preventing probing by An. albimanus, with no statistically significant differences between them. Treatment with diastereomers DHN 1 or 2 significantly reduced probes compared with the control treatment.

View this table:

Tests for repellency against the stable fly Stomoxys calcitrans were conducted. An accurate time to "first landing" could not be determined, because some landings occurred before the first exposure period of 2 min in three or more of the five replicates for each test variable. The numbers of landings by stable flies at each observation period is shown graphically in Fig. 5. Landings started on exposure of the flies to the test wells, reached a maximum at 4 min, and gradually decreased. Table 4 shows that the total number of landings per replicate on membranes treated with DHNs were significantly fewer than on the negative control membranes and similar to those observed with DEET. PMD seemed to be less effective in repelling landings than either the DHNs or DEET, although there was no statistically significant difference in the numbers of landings. The data show that treatment with DHN diastereomers significantly reduced landing on the membranes compared with the negative control treatment and that DHNs were statistically equivalent to DEET and PMD in repellency.

Fig. 5

Distribution of landing density over time, during tests of various repellents against stable flies (S. calcitrans). Pure DHN1 and hydrogenated catmint oil (containing predominantly DHN2) were tested. Negative controls used IPA; PMD and DEET solutions were tested for comparison. All repellents were 1% (wt:vol) in IPA.

View this table:

Tests were conducted in vivo (i.e., on human subjects) for repellency against black-legged tick (I. scapularis) nymphs. The data (Table 5) indicate that DEET provided a mean CPT from this tick species of 124 min, whereas DHN 1 was repellent for 109 min and DHN 2 for 85 min. Thus, it is clear that both DHNs are repellent toward the deer tick, with DHN 1 perhaps marginally less so than the well-known repellent DEET. It must be noted that there was a wide a range in the CPT for each treatment (see Table 5 for SEMs), which limits the utility of these results.

View this table:

Tests for repellency of DHN 1 and DHN 2 were also carried out against An. albimanus using adult human subjects. This species was chosen instead of Ae. aegypti because of its tolerance of DEET (Barnard 1998, Klun et al. 2004). Testing was conducted on 2 d with six human subjects: first with 10% dilutions of DHN 1, DHN 2, and DEET and then with 5% of these three materials. The results (Table 6) indicate that both DHNs conferred complete protection from bites for hours. For example, at 10% (wt:vol), DHN1 gave 3.5 h and DHN2 gave 5.0 h of protection, comparable to that afforded by DEET (6.4 h) at the same concentration. In this test format, repellency of DHN 2 was statistically indistinguishable from that of DEET at the same concentrations. In the in vitro assay, DHN 2 appeared superior to DHN 1 with this species. Although DHN 1 appeared to give shorter protection when applied to human subjects (Table 6), the performance of DHN 1 was statistically indistinguishable from that of DHN 2 in this test.

View this table:

## Discussion

In the study presented here, the efficacy of dihydronepetalactone, either in pure form (DHN 1 derived from hydrogenation of purified E,Z-nepetalactone) or as the principal constituent in hydrogenated Z,E-nepetalactone enriched catmint oil (mixture of DHN diastereomers, predominantly DHN 2), as a repellent to mosquitoes, stable flies, and ticks was examined. Analysis of mosquito and stable fly landings on treated membranes in an in vitro repellency assay (Rutledge et al. 1976, Rutledge and Gupta 2004) indicated that, in fact, these chemicals possessed promise as topical repellents against blood-seeking insect pests. In these tests, deterrence of insect landings was comparable to that of the commonly used insect repellent active ingredient, DEET, with Ae. aegypti, and seemed to be superior to that of another natural product repellent, PMD. The in vitro tests with Ae. aegypti indicated that DHN 1 was more effective than DHN 2 in preventing probes. The in vitro data for An. albimanus showed equivalent reduction of probes by DHN 1 and 2, DEET, and PMD. An. albimanus is known to be insensitive to DEET and other repellents relative to other mosquito species (Barnard 1998, Klun et al. 2004), and this is borne out by the data reported in this paper, where percentage reduction in probes was less than for Ae. aegypti. DHN 1 and 2 were also deterrent to stable flies. Although these experiments indicated efficacy of DHNs as repellents and allowed some comparison with efficacy of other known repellents, the in vitro test system used has several known limitations. These limitations include the close proximity of the treated membranes within the test apparatus and the absence of a human host. This method is, however, useful for rapid, economical screening without potential hazards to humans. The data are thus indicative of feeding deterrency, and this may be an indicator of repellency in more realistic circumstances. To confirm this indication, further testing using human subjects was conducted.

In in vivo studies with An. albimanus mosquitoes, human subjects were protected from bites for several hours. Protection in these tests, as well as in tests using the nymphal life stage of the deer or black-legged tick nymphs was comparable in duration to DEET. The DHN 1 compound appeared to more effective than DHN 2 with ticks. However, the differences in efficacy between these two DHN diastereomers were minor and were not statistically significant. In contrast, DHN 2 was the more effective diastereomer with An. albimanus mosquitoes. Thus, the data presented here are encouraging that dihydronepetalactone diastereomers and mixtures thereof, e.g., hydrogenated catmint oil, might constitute a useful natural product insect repellent as an alternative to synthetic chemicals, with a broad spectrum of efficacy and comparable properties to existing (natural and synthetic) repellents.

## Acknowledgements

The many staff of ICR, especially T. Foard, N. Spero, and B. Gaynor (who administered the repellency testing described here), F. Zgidou (who reared the mosquitoes), and G. Stevens (who reared the stable flies), are gratefully acknowledged. Similarly, the excellent assistance of N. Keiper-Hrynko, A. Liauw, and B. D’Achille (DuPont CR&D) is gratefully acknowledged. Also within DuPont CR&D, D. Kline (Corporate Process Development) conducted the fractional distillation of catmint oil; D. Dragotta conducted the GC analyses of catmint leaf extracts; L. Manzer and L. Kao conducted small-scale hydrogenation reactions; and E. McCord and E. Lozada (Corporate Center for Analytical Science) conducted the NMR analyses.

## Footnotes

• Hydrogenated catmint oil has been assigned the INCI name "Hydrogenated Nepeta cataria oil" and the tradename "Refined Oil of Nepeta cataria" by the Personal Care Products Council.