As part of an arbovirus vector control strategy, chemical control continues to be a mainstay in mitigating the burden of disease. The current arsenal of chemicals used for this purpose, however, are becoming challenged rapidly because of issues of insecticide resistance and environmental pressure. Newer, environmentally friendly actives are of interest to supplement aging chemistries; therefore efforts to screen compounds for insecticidal activity are warranted. This study evaluated the efficacy of the high-throughput screening system (HITSS) for measuring the behavior-modifying actions of Brazilian Cerrado plant extracts, oils, and other compounds against Aedes aegypti. Different concentrations were evaluated, with 8 of 34 samples tested demonstrating either contact irritancy, spatial repellency, or attractiveness. We concluded several natural products screened in this study showed promise for use against mosquito vectors like Ae. aegypti, and that the compact modular HITSS assay constitutes a robust tool for measuring the behavioral responses of mosquitoes in the search for novel insecticides derived from natural products.
Despite organized vector control efforts, mosquito-borne arboviral diseases continue to pose a serious global public health issue, with the number of cases constantly increasing (Pan American Health Organization [PAHO] and World Health Organization [WHO] 2021). In 2019, more than 3 million cases of dengue were reported in the Americas, constituting the highest recorded incidence of dengue in the history of this region (PAHO and WHO 2020). Until July of 2021, more than 673,000 cases were reported for dengue (PAHO and WHO 2021). To add to the difficulty of arboviral disease control, there are currently no specific treatments or widely acceptable vaccines available for preventing dengue, Zika, and chikungunya (Da Silveira et al. 2019, Sukhralia et al. 2019). Aedes aegypti (L.), the primary mosquito vector for a number of arboviruses, has been shown to be highly domesticated with an extensive global distribution, which renders this mosquito a major threat to humans (Brady and Hay 2020).
From a public health perspective, chemical-based vector control targeted against Ae. aegypti has been a fundamental strategy for disease control and outbreak mitigation (Garcia et al. 2018). The vector control community has been exploring novel active ingredients that exhibit different modes of action. Historically, screening for new chemical actives has mainly focused on arthropod toxicity outcomes, with none to minimal consideration to assessing other important chemical properties such as contact irritancy and spatial repellency (Liu et al. 2021). Both of these actions hold the potential of creating a vector-free space and breaking human–vector contact (Achee et al. 2012). A benefit of investing in products targeting behavioral changes in Ae. aegypti, other than acute toxicity, is the potential mitigation of insecticide resistance. In addition, the use of lower doses implies less toxicity to nontarget organisms (Achee et al. 2012). Natural products have been investigated as repellents with low human toxicity, low cost, and biodegradability. Such products provide complex innovative structures with potent biological activity (Suwansirisilp et al. 2013, Estrada et al. 2019, Peach et al. 2019).
Screening chemical libraries for modes of action beyond mortality requires assays that employ unique methodologies for measuring such endpoints (Leal et al. 2017, Bakker et al. 2020, Richoux et al. 2020). The HITSS (high-throughput screening system) is a behavioral assay that was developed in order to provide rapid screening of a large number of chemical compounds for contact irritant and spatial repellent actions against mosquitoes (Grieco et al. 2005). The assay is currently incorporated into the World Health Organization (WHO) guidelines for efficacy testing of spatial repellents (WHO 2013). The objective of the current study was to investigate the behavioral responses of Ae. aegypti to select natural products including extracts, oils, and compounds as part of a Brazilian screening program for new vector control prototypes, ArboControl, using the HITSS behavioral system.
Laboratory assays were conducted using female Ae. aegypti (Liverpool strain) of 4–7 days old. The colony was established in the University of Notre Dame (USA), where it was maintained at 27°C, 80% relative humidity (RH), in a 12 h light and 12 h dark photoperiod. A blood meal was given, using a glass mosquito feeder (Part 89-940, 50 mm; NDS Technologies, Inc., 891 East Oak Road, Vineland, NJ) covered by an artificial membrane made of collagen sausage casing (Waltons Inc., Wichita, KS). The blood meal (Interstate Blood Bank, Inc., Memphis, TN) was given 1 time per week for maintaining egg production. Newly hatched larvae were sorted into groups of 50 individuals and placed in 500-ml pots, where they were fed on a diet of ground fish chow (Hikari Cichlid) and allowed to develop to the pupal stage. Pupae were sorted by size to separate females from males and the latter placed inside 2-gal cartons until emergence. Adults were provided with a cotton pad soaked in a 10% sugar solution until time of assaying. Groups of 10 or 20 females, depending on the assay type, were transferred to smaller cartons 24 h prior to testing with no sugar solution available.
The HITSS, used for evaluating behavioral mosquito responses, is composed of a metal treatment cylinder, a clear cylinder, end caps with viewing windows and mosquito introduction portals, linking sections, an inner metal spool, and treatment nets, as previously described by Grieco et al. (2007). For contact irritancy assays, the device was assembled to allow mosquito tarsal contact with the chemical-treated material, whereas for the spatial repellency assay, the device was configured such that mosquitoes were introduced into the central, chemical-free chamber, exposed to airborne chemical in a gradient. Full details about tests methodologies can be accessed at https://biology.nd.edu/labs/achee-grieco-lab/training-manuals/.
Test compounds were part of a Brazilian screening program (ArboControl) for innovative natural products to control Ae. aegypti. In total, 34 Arbo-coded samples were tested: 17 isolated compounds, 14 plant extracts, 1 essential oil, 1 oil, and 1 fixed oil. Samples were diluted in acetone or ethanol to different concentrations with 1.5 ml applied to a 275-cm2 netting strip (nylon organdy, mesh size 150 × 150 μm, Filenes Fabrics, Inc., South Bend, IN), using a micropipette inside a chemical hood. Additional untreated control nets were prepared with ethanol or acetone diluent only, matched in tests to the dilution solvent used for active compound. All nets were allowed to dry for 15 min prior to insertion in the HITSS. The same treated net was used for both contact irritancy and spatial repellency assays, with the exception of the oils, for which an individual treated netting strip was prepared for each assay.
Decisions on test concentrations depended on the availability of each sample, as many were in limited volume for testing. Plant extracts and the fixed oil were tested at 10% (w/v). Essential oil and the additional oil were tested at 10% and 1% (v/v). Compounds were tested at 1% (v/v), 5% (v/v), and 25 nmol/cm2.
The assays of a given sample were conducted on the same day within 1–7 h after netting strip treatment. The average temperature during all tests was 24°C with 35% RH. Acetone was used to clean all metal parts in the assay devices that had direct contact with the netting strips. The remaining parts were cleaned using a detergent solution (Liqui-Nox, Aloconox, Inc., White Plains, NY), rinsed thoroughly with water, and air dried overnight.
For contact irritancy, each sample was assayed with 6 test replicates of 10 ± 2 females with a parallel control. Black opaque felt was placed over the end-cap viewing windows and wrapped around the clear cylinder to prevent any light-induced mosquito behavioral effect. The females were inserted into the metal cylinder and allowed to rest for 30 sec in direct contact with the treated netting strips. The butterfly valves from the linking section were subsequently turned to the open position for 10 min. During this time, the mosquitoes could move freely inside the device. The valves were then closed and the numbers of alive and knocked down (KD) mosquitoes (immobile or unable to stand) in the metal and clear cylinders (number escaping) were recorded separately. Mosquitoes were removed from the assay device prior to commencing subsequent assays.
For spatial repellency assay, a total of 20 ± 2 mosquitoes were placed into the central chamber and allowed to rest for 30 sec with black felt wrapped around the clear cylinder. The valves to the metal chambers containing the treated (or control) material were then simultaneously opened and mosquitoes allowed to move throughout the device freely. After 10 min, the valves were closed and the number of mosquitoes in each cylinder individually recorded. The KD was also noted in each chamber. After removing the test cohort, metal cylinders were allowed to ventilate for 3 min to remove residual volatile chemicals from the assembly before staring a subsequent assay. A total of 9 replicates were conducted for each sample tested.
All data were analyzed using the SAS University Edition software (SAS Institute, Inc. Cary, NC). Contact irritant activity was measured as the percentage of mosquitoes escaping from the metal cylinder containing the treated netting strip into the clear cylinder. The percent escape was corrected as compared to the escape and KD in the control assay. A Wilcoxon 2-sample test (PROC NPAR1WAY) was used to assess the difference between the number of mosquitoes escaping from treatment and control chambers. A spatial activity index (SAI) was used to measure the spatial repellent activity of each sample. The SAI was calculated according to the formula: SAI = (Nc − Nt)/(Nc + Nt), in which the Nc is the number of mosquitoes in the metal control chamber from the spatial repellency device; Nt corresponds to the number of mosquitoes in the metal treatment cylinder. The SAI varies from −1 to 1: negative values indicate attractive activity, whereas positive values indicate repellency. A nonparametric signed-rank test (PROC UNIVARIATE) to determine if the SAI was statistically different from zero. An additional analysis was performed for ranking samples activity by wSAI (SAI ∗ PRESP), where PRESP corresponds to the percent of mosquitoes responding (choosing control or treatment cylinders) in the test.
Of the 34 samples tested, 6 presented statistically significant contact irritancy, 3 oils (essential oil, oil, and fixed oil) and 3 compounds (Table 1). None of the plant extracts and remaining compounds evaluated elicited contact irritancy, toxicity, or mosquito KD.
The essential oil ArboS67 (10%) caused 80.28% escape in the treatment cylinder, with almost 16% KD inside the metal treatment cylinder. The fixed oil ArboBR2295 (10%) resulted in 74.44% escape with no KD. The oil ArboINT086 (1%) presented 66% escape and a very low KD (3.33%) in the metal cylinder. The 3 compounds that caused contact irritancy were further evaluated at different concentrations. ArboS156 (1%) caused 87% escape with 67.2% KD. Compound ArboS109 (5%) caused 79.78% escape, with 23.7% KD. ArboBR1924 (25 nmol/cm2, equivalent to 0.08%) caused 22.78% escape and no toxicity.
Three samples presented spatial repellency activity, 2 essential oils and 1 compound, whereas 2 plant extracts were attractive. The essential oil ArboS67 promoted significant spatial repellency (SAI = 0.61 ± 0.16) (Table 2). The SAI of the oil ArboINT086 was not statistically significant (P > S = 0.078). However, when a weighted SAI was calculated, as wSAI = SAI ∗ PRESP (PRESP is the mean percent responding), the sample presented a significant repellency rate (wSAI = 10.59 ± 3.45, P > S = 0.02) (Table 2). The 2 aforementioned essential oils had no, or very few, KD mosquitoes in the spatial repellency assays. The compound ArboS156 promoted spatial repellency (SAI = 0.70 ± 0.10) (Table 2), with 7.7% KD. Conversely, 2 plant extracts significantly modified mosquito behavior, not as repellents but rather as attractants: Arbo0289 (SAI = −0.61 ± 0.10) and Arbo0294 (−0.64 ± 0.14). No toxicity was noted for these extracts.
Our data demonstrate the capacity of the HITSS to evaluate different behavioral responses of Ae. aegypti including contact irritancy, spatial repellency, and attraction to ranging concentrations of natural products supporting them as potential sources of innovative insecticides. The oils tested presented a degree of repellent activity. However, only the essential oil and the oil tested elicited contact irritancy and spatial repellency. These findings support previous results within the natural products field (Suwansirisilp et al. 2013, Peach et al. 2019).
Of the 34 natural products screened in this HITSS study, the oil ArboINT086 proved the most interesting. This sample was nontoxic to mosquitoes and presented significant results in both assays at a low concentration (1% or 15 mg) when compared to previous essential oil repellency studies (Peach et al. 2019). ArboS156 was the only compound that exhibited both contact irritant and spatial repellent actions but was highly toxic to females. Interestingly, 2 plant extracts showed an attractant response, which could be exploited as oviposition attractants of gravid females in a different vector-control strategy (Suman 2019). The compound ArboS156 had the highest KD, and, therefore, could be evaluated at lower concentrations in future tests.
The current study continues to support the HITSS system as a compact, versatile, and easy-to-implement assay option for assessing behavioral responses of mosquitoes to chemicals. The methodology enabled testing of samples on a milligram scale, which is very important when dealing with natural products which are often only available in small quantities. For evaluating insect irritancy, the WHO recommends a cone assay placed to an insecticide-treated surface (WHO 2006). This experiment allows evaluating compounds applied to different surfaces commonly used as building materials, which is an advantage compared to the HITSS contact irritancy assay. On the other hand, flexibility of the HITSS chambers to be reconfigured for either contact irritant or spatial repellent test modality added to the utility. Our data demonstrated that HITSS is a robust assay with wide utility in vector behavioral studies.
The present study was a cooperation between University of Notre Dame and Universidade de Brasília, supported by CAPES, through CAPES PrInt (88887.364317/2019-00). Marianne Krebs is also thanked for all technical support. We acknowledge the Brazilian Ministry of Health for the funding and fellowships under the ArboControl project grants: TED 74/2016 and 42/2017, and the Genetic Heritage Management Council (CGen)/IBAMA 06/2012-Process 02000.002272/2006-73 for granting authorization to access the plant samples.
Universidade de Brasília, Laboratório de Farmacognosia, Campus Universitário Darcy Ribeiro, 70910-900, Brasília, DF, Brazil.
Department of Biological Sciences, Eck Institute for Global Health, University of Notre Dame, Notre Dame, IN 46556.