Evaluation of insecticide resistance and underlying resistance mechanisms in selected whitefly populations in Sri Lanka

: Whiteflies Bemisia tabaci and Trialeurodes vaporariorum cause high economic damage to commercial crops in Sri Lanka. Insecticide resistance was tested in B. tabaci collected from Kandy, Anuradhapura, Badulla, Matale and Hambantota districts and T. vaporariorum collected from Puttlam District against six insecticides using leaf dip bioassays. LC 50 values and percentage mortalities for the recommended dosages were determined by log-probit mortality curves. Metabolic resistance caused by increased activities of esterases, glutathione S-transferases and cytochrome P 450 monooxygenases, and the insensitive target site resistance caused by acetylcholinesterases were evaluated by biochemical assays. Overexpression of CYP6CM1 monooxygenase protein in B. tabaci was tested using a lateral flow assay kit. The highest and the lowest resistance factors (RFs) were obtained for etofenprox (1311) and profenofos (4), respectively. Resistance percentages of whitefly populations to the recommended dosages were 37−83% for thiamethoxam, 50−83% for acetamiprid, 71% for imidacloprid, 2–36% for profenofos, 15–53% for carbosulfan and 68–89% for etofenprox. Anuradhapura, Badulla and Matale populations had elevated esterases, all populations had elevated GST activities and Badulla population had elevated monooxygenase activities. Populations were homozygous resistant or heterozygous for altered acetylcholinesterase mechanism. High esterase activities were parallel to high neonicotinoid resistance. CYP6CM1 overexpression was linked to thiamethoxam resistance than to acetamiprid resistance. Both the enhanced esterases and altered target sites were associated with high carbosulfan resistance. This is the first report on the insecticide resistance mechanisms of whiteflies in Sri Lanka. The outcome of the present study can be effectively utilized for a proper whitefly control programme in Sri Lanka, avoiding indiscriminate insecticide usage.


INTRODUCTION
Whiteflies in vegetables cause extensive economic damage. The silverleaf whitefly, Bemisia tabaci (Gennadius) (Insecta: Hemiptera: Aleyrodidae) is a injurious pest in agriculture (Bedford et al., 1994) and listed as one of the 100 worst invasive species in the world (Global Invasive Species Database, http:// www.issg.org/database). The greenhouse whitefly, Trialeurodes vaporariorum (Westwood) (Insecta: Hemiptera: Aleyrodidae), although not included in the invasive species list, is considered as another injurious whitefly present globally (Brødsgaard & Albajes, 1999;Karatolos, 2011). In Sri Lanka, both B. tabaci and T. vaporariorum cause high economic damage to commercial crops. The number of host plant species of B. tabaci has been estimated to be more than 600 plant species (Basu, 1995) whereas T. vaporariorum is also known as a highly polyphagous species having a wide host plant range (Brødsgaard & Albajes, 1999). However, the host range of these pests could be even higher at present with its wider global distribution. Out of all the whiteflies, B. tabaci has become prominent because of its ability to transmit a large number of begomoviruses causing complete destruction of crops (Brown & Bird, 1992). Bemisia tabaci is now considered as a cryptic species complex, rather than different biotypes, due to December 2021 Journal of the National Science Foundation of Sri Lanka 49 (4) high genetic varaitions found among populations (De Barro et al., 2011). Both the physiological traits and genetic variations are crucial in determining its species status, and Surendran et al. (2019) had identified that the predominant species type of B. tabaci in Sri Lanka is Asia-I.
Naturally, the control of whiteflies is difficult due to their high reproductive rate, wide host range and habit of lodging underneath leaves. On top of these natural features, the acquired resistance to insecticides by them is the most challenging at present. Insecticide resistance of whiteflies is a global issue. Bemisia tabaci is found to be resistant to more than 40 active ingredients used worldwide (Arthropod Pesticide Resistance Database, www.pesticideresistance.org). In brief, insecticide resistance is a heritable trait acquired by a population to withstand the toxicity of an insecticide repeatedly at its recommended dosage (Insecticide resistance action committee, https://irac-online.org/about/resistance). High resistance of T. vaporariorum to many insecticide groups including pyrethroids, organophosphates and neonicotinoids has been documented elsewhere (Cahill et al., 1995;Gorman et al., 2002;Nauen & Denholm, 2005;Karatolos, 2011). The two major mechanisms of insecticide resistance in insects are; increased metabolism or detoxification by enzymes (mainly through esterases, mono-oxygenases and glutathione-S-transferases), and target site insensitivity by mutations (i.e. altered sodium channels for pyrethroids, altered acetylcholinesterase for organophosphates and carbamates, and altered nicotinic acetylcholine receptors for neonicotinoids) (Karunaratne, 1998;WHO, 1998).
Previously, we reported high resistance to insecticides in three whitefly populations based on laboratory bioassays raising concerns over the existing dosages recommended for chemical control (Marasinghe et al., 2017). The objective of the present study was to investigate the present status of whitefly resistance to insecticides in more detail by analysing the whitefly populations collected from six districts of the country against three neonicotinoids, one organophosphate, one carbamate and one pyrethroid. The activities of metabolic enzymes and altered acetylcholinesterase target site were also evaluated to understand the mechanisms underlying the insecticide resistance of this injurious pest category.

Whiteflies
Bemisia tabaci populations were collected from several intensive agricultural localities from five districts belonging to four agroecological zones of Sri Lanka, i.e. Gannoruwa in Kandy, Huruluwewa in Anuradhapura, Bandarawela in Badulla, Matale in Matale, Suriyawewa in Hambantota and Kalpitiya in Puttalam. All collections were made during the 2016-2018 period and the details are given in Table 1. Bemisia tabaci collected from Puttalam were not enough for experiments. Trialeurodes vaporariorum dominated the whitefly collections from Puttalam district, and those samples were used for experiments for comparison.
Whiteflies were transported to the entomology laboratory of the Horticulture Research and Development Institute, Gannoruwa in cool boxes within few hours of collection. They were reared separately on young brinjal, tomato or cucumber plants in fine-mesh aluminium cages of 80 × 40 × 60 cm 3 under room temperature up to 5-6 generations.
Species identification of both species were reconfirmed using adult/nymphal phenotypic features and the egg-laying pattern, i.e. B. tabaci adults are yellowish in colour and comparatively smaller with wings held vertical and parallel to body axis at rest. The fourth instar nymph (pupa) is almost circular in shape without  long threads. Eggs are laid in a scattered manner on the underside of host leaf. Trialeurodes vaporariorum adults hold their wings in a flattened and overlapped position at rest. The fourth instar nymph is oval/ elliptical in shape with long erect threads. Eggs are laid in a circular pattern on the underside of the host leaf. Slide-mounted pupae were also examined for characteristic morphological features to differentiate the two species (Hodges & Evans, 2005). A set of specimens from each population was stored at -20 0 C for biochemical assays.

Biochemical assays
Adult whiteflies in batches of 20 individuals totalling up to a minimum of 200 individuals from each population were subjected to esterase, glutathione S-transferase (GST), monooxygenase, protein and acetylcholinesterase assays using a kinetic microplate reader (Bio-Tek, USA). Mass homogenates of the batches of 20 whiteflies were prepared separately in 150 µL of distilled water. Crude homogenate was directly used for acetylcholine esterase assay. The supernatant after centrifugation for 2 min at 10,000 g was used for esterase, GST, monooxygenase and protein assays. All the biochemical assays were carried out in microtiter plates. All the tests were replicated and controls were run with distilled water instead of the crude homogenate or supernatant. The assaying methods described in the Field and Laboratory Manual of the World Health Organization (WHO, 1998) were adopted with slight modifications as previously described for agricultureal insect pests (Damayanthi & Karunaratne, 2005;Karunaratne & Weerakoon, 2007).

Esterase assay
Ten microlitres of homogenate was mixed with 200 µL of 1 mM p-nitrophenyl acetate (pNPA) working solution in 50 mM sodium phosphate buffer (pH 7.4). At 405 nm, the absorbance of the reaction was measured kinetically for 2 min and converted to moles using an extinction coefficient of 6.53 mM −1 (corrected for a path length of 0.6 cm).

Glutathione S-transferase (GST) assay
Ten microlitres of homogenate was mixed with 200 µL of substrate solution prepared using 10.5 mM reduced glutathione (95 parts GSH in 100 mM phosphate buffer) and 63 mM 1-chloro-2,4-dinitrobenzene (CDNB) (5 parts of CDNB in methanol). The reaction rate was measured at 340 nm for 5 min. The absorbance was converted to moles using an extinction coefficient of 5.76 mM −1 (corrected for a path length of 0.6 cm).

December 2021
Journal of the National Science Foundation of Sri Lanka 49 (4) of sodium acetate buffer at pH 5.0) + 25 µL of 3 % H 2 O 2 solution in a microtitre plate well. The mixture was left for 2 h for incubation under room temperature and the absorbance was read as an end point assay at 630 nm. The amount of monooxygenase was expressed as equivalent units of cytochrome P 450 (Brogdon et al., 1997).

Protein assay
Protein assays were conducted to obtain protein concentrations of the homogenates to determine specific activities of the enzymes. BIO-RAD protein determination kit, with bovine serum albumin as the standard protein was used. Ten microlitres of the homogenate was mixed with 300 µL of BIO-RAD working solution (prepared according to the manufacturer's instructions) and after a 5 min incubation period at room temperature, the reaction was read at 570 nm as an end point assay.

Acetylcholinesterase (AChE) assay
Two × 25 µL of crude homogenate were placed in two consecutive wells of a microtiter plate, each containing 145 µL of 1 % Triton buffer in 100 mM phosphate buffer at pH 7.8 with 10 µL of 10 mM dithiobis 2-nitrobenzoic acid (0.0396 g of DTNB in 100 mM phosphate buffer at pH 7.0). An aliquote of 25 µL of 10 mM acetylthiocholine iodide (ASCHI) (0.0578 g ASCHI in 20 mL distilled water) was added to one of the wells while 25 µL of 10 mM ASCHI with propoxur (10 mL of 10 mM ASCHI + 20 µL of 0.1 M propoxur in acetone) was added to the other. Acetylcholinestearse (AChE) activity was measured at 405 nm for 5 min. Results were expressed as the percentage remaining activity in the inhibited fraction compared with the control (uninhibited) activity.

Detection of the CYP6CM1 protein in B. tabaci
A field portable ELISA based lateral flow assay designed to detect monooxygenase CYP6CM1 protein was used following the manufacture's protocol (Bayer CropScience). Five to ten whiteflies from each population were homogenised in conjugated polyclonal antibodies in a homogenization vial. After a 2-3 min incubation, streptavidin-gold solution was added to the vial and gently inverted 3-4 times. The mixture was then poured into the sample cavity of the test strip and observed for 5 min. Respective distinct lines at C and T points indicated the expression levels of the CYP6CM1 enzyme by the control and the sample, respectively.

Data analysis
Bioassay data were used only when the control mortality was less than 20 %. Mortalities were adjusted with control mortalities using Abbott's formula (Abbott, 1987). Adjusted mortalities were Probit transformed using Sigma Plot (version 10) software and plotted against the log values of the insecticide concentrations. The LC 50 and the percent mortality for the Department of Agriculture recommended dosages (DOA, 2015) were estimated by regression analysis. Chi square values (χ 2 ) were calculated by evaluating the goodness-of-fit of the linear regression to log-probit transformed data.
The software provides the slope and the 95 % confidence limits for each mortality line. Resistance factors (RF) were calculated using LC 50 values reported for susceptible populations of the same species elsewhere and used to evaluate resistance levels (Cahill et al., 1995;Karatolos, 2011;Vassiliou, 2011).

RESULTS AND DISCUSSION
A total of 48,989 whiteflies from B. tabaci and T. vaporariorum were subjected to insecticide bioassays. Probit analysis separately done for the bioassay results obtained for the six populations with six insecticides are shown in Table 2   N -number tested; CI -confidence intervals; χ 2 -chi square value evaluating the goodness-of-fit of the linear regression to log-probit transformed data; RF -resistance factor; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; 1 susceptible data used from Vassiliou et al. (2011), 2 from Karatolos (2011), 3 from Cahill et al. (1995); NT-not tested; NC-not calculated due to unavailability of susceptible data.

December 2021
Journal of the National Science Foundation of Sri Lanka 49 (4) Mortality of whitefly populations to the insecticide dosages recommended by the Department of Agriculture (DOA) of Sri Lanka (DOA, 2015) as determined by using the mortality curves are presented in Table 3. Percentage mortalities of populations varied from 11−98 %. Whiteflies from Kandy, Matale and Hambanthota gave more than 90 % mortalities for the DOA recommended profenofos dosage, and for other three districts 3−9 fold dose increment was required to achieve a 90 % mortality. All the other tested insecticides required 3− >80fold increase of the recommended dosage to achieve 90 % mortality in these populations. Trialeurodes vaporariorum gave a similar mortality as for B. tabaci to DOA recommended dosage except for etofenprox, which showed a lower efficacy for T. vaporariorum (Table 3). Imidacloprid could be tested only against T. vaporariorum and 29 % mortality obtained for the DOA recommended dosage indicates the high imidacloprid resistance in whiteflies. In general, high resistance to all the tested insecticides is evident in whitefly populations except for profenofos in a few populations (Table 3).  However, the esterase specific activities cannot be used to compare with the present results since the substrate used for the assay was different to that of the present study. Using the same methodology of the present study, a GST specific activity of 0.2 μmol/mg/min has been obtained for their susceptible population, which is about 5-8 times lower than the GST activities detected in our whitefly populations (Vassiliou et al., 2011). No other enzyme activity data from susceptible strains are available for whiteflies or for any other insect pest of agriculture for comparison. However, for anopheline mosquitoes discriminating values are available in the literature for the same assays to determine the enhanced activities; above 0.25 μmol/mg/min for enhanced esterase activity, above 0.40 μmol/mg/min for enhanced GST activity, and 0.35 equivalent units of cytochrome P 450 for enhanced monooxygenase amounts (Perera et al., 2008). According to this classification, Anuradhapura, Badulla and Matale populations have elevated esterases, all six whitefly populations have elevated GST activities and Badulla population has elevated monooxygenase activities (Figure 1). In aphids and mosquitoes, elevation of esterases is a powerful mechanism against organophosphates and carbamates, which have molecular structures rich with esterase bonds (Karunaratne et al., 2018). Thiamethoxam resistant B. tabaci has been synergized and reversed using the esterase inhibitor S,S,S-tributyl phosphorotrithioate (DEF) showing its importance in thiamethoxam resistance (Kandil et al., 2008). The present study also reveals high esterase activity levels parallel to high neonicotinoid resistance, indicating the contribution of esterases to neonicotinoid resistance. Although GSTs are primarily important in declorination of organochlorines, elevated levels of GST activity have been reported in organophosphate, organochlorine and pyrethroid resistant mosquitoes (Che-Mendoza et al., 2008). High GST levels observed in the studied whitefly populations may have a wider general role giving resistance to all insecticide groups.
Results of the ELISA based lateral flow assay for CYP6CM1 monooxygenase protein also showed a clear overexpression of CYP6CM1 monooxygenase protein in Badulla B. tabaci population (Figure 2). The CYP6CM1 expression levels of lateral flow assay nicely matched with the biochemical assay data indicating that CYP6CM1, a typical microsomal P 450 member of CYP6 family, is a major candidate of enhanced monooxygenases of the tested whitefly populations. Imidacloprid resistance in both the B and Q biotypes of B. tabaci has been tightly related to the up-regulation of the cytochrome P450dependent monooxygenase gene CYP6CM1 (Karunker et al., 2008). Overexpression of CYP6CM1 gene has also been detected in thiamethoxam resistant B. tabaci too (Yang et al., 2013). Imidacloprid resistance in B. tabaci has not been tested in our study and thus, the association with imdicaloprid resistance could not be traced. Results of the ELISA based lateral flow assay for CYP6CM1 monooxygenase protein also showed a clear overexpression of CYP6CM1 monooxygenase protein in Badulla B. tabaci population (Figure 2). The CYP6CM1 expression levels of lateral flow assay nicely matched with the biochemical assay data indicating that CYP6CM1, a typical microsomal P 450 member of CYP6 family, is a major candidate of enhanced monooxygenases of the tested whitefly populations. Imidacloprid resistance in both the B and Q biotypes of B. tabaci has been tightly related to the up-regulation of the cytochrome P450-dependent monooxygenase gene CYP6CM1 (Karunker et al.,   According to our bioassay results from Badulla and Matale populations, it can be concluded that CYP6CM1 is more important in thiamethoxam resistance than in acetamiprid resistance. Percentage remaining activities of propoxur inhibited target site AChE are shown in Figure 3. The same AChE assay is used by WHO to categorize mosquito populations as susceptible homozygous (SS) (< 30 % remaining activity), heterozygous (RS) (30−70% remaining activity) and resistant homozygous (RR) (> 70% remaining activity) for altered or insensitive AChE target site mechanism (WHO, 2016). According to these guidelines, Badulla B. tabaci population was homozygous resistant with > 70% remaining activity while all other populations including T. vaporariorum population were heterozygous for altered AChE mechanism with a remaining activity of 30−70% (Figure 3).  High organophosphate and carbamate resistance is expected from the altered AChE target esterases activities (Karunaratne et al., 2018). Acetylcholinesterase insensitivity has been pre OP and/or carbamate resistant whiteflies (Dittrich & Ernst, 1990;Byrne & Devonshire, 1991).
profenofos and carbosulfan resistance detected in Badulla and Matale districts for B. tabaci p due to these two mechanisms.
Whitefly control has become a serious problem in agriculture due to reduced efficacy insecticides. Severity of the problem has intensified with the economical importance of begomovirus vector of many crops. A large number of insecticides belonging to different che been recommended for whitefly control in Sri Lanka (MOAAL, 1997;DOA, 2009 (Karunaratne et al., 2018). Acetylcholinesterase insensitivity has been previously reported in OP and/or carbamate resistant whiteflies (Dittrich & Ernst, 1990;Byrne & Devonshire, 1991). Therefore, the high profenofos and carbosulfan resistance detected in Badulla and Matale districts for B. tabaci populations must be due to these two mechanisms.
Whitefly control has become a serious problem in agriculture due to reduced efficacy of recommended insecticides. Severity of the problem has intensified with the economical importance of B. tabaci being a begomovirus vector of many crops. A large number of insecticides belonging to different chemical groups have been recommended for whitefly control in Sri Lanka (MOAAL, 1997;DOA, 2009). Organophosphates and carbamates have a long history of use for more than 4 decades. Pyrethroids were introduced in the 1990s and neonicotinoids in early years of this century for whitefly control (DOA, 2009). The popularity of neonicotinoids among farmers increased not only due to their high efficacy at the introductory phase, but also due to the wide crop range applicability and the long residual activity. The present study evaluated the efficacy of representative samples of these insecticide groups against B. tabaci and T. vaporariorum using laboratory bioassays with the leaf dip method, which resembles the foliar application method practiced in the field. The percentage mortalities estimated using probit regression lines revealed high resistance of whiteflies to all the tested insecticides, Journal of the National Science Foundation of Sri Lanka 49 (4) December 2021 except for profenofos. Since repeated use of less effective insecticides would further increase resistance generation with environment pollution, taking different approaches to control whiteflies at this stage is necessary. The underlying mechanisms of insecticide resistance in Sri Lankan whitefly populations were evaluated for the first time using biochemical assays. It was evident that both the metabolic resistance and insensitive target site resistance mechanisms have been developed to varying degrees in Sri Lankan whitefly populations. In T. vaporariorum also the resistance levels and resistance mechanisms are found to be comparable to those of B. tabaci.
Etofenprox is a non-ester pyrethroid having a mode of action similar to other pyrethroids, i.e. by interacting with voltage gated sodium channels (VGSC) of insect neurons. Cross resistance studies have shown no effect of enhanced metabolic enzyme activities on etofenprox toxicity in mosquitoes possibly due to the relatively simple structure of the etofenprox molecule compared to most of the other pyrethroids (Hemingway, 1995). Also, the monooxygenase synergist piperonyl butoxide (PBO) and GST synergist diethyl maleate (DEM) were able to synergize permethrin resistance, but not etofenprox resistance in Anopheles mosquitoes (Gross & Bloomquist, 2018). With the high resistance observed in all whitefly populations against etofenprox, and with a least chance for metabolic resistance mechanisms to be operative, there is a high possibility that knockdown type (kdr) target site mutations are involved in this resistance. Future studies are required to investigate the presence of kdr-type VGSC gene mutations in Sri Lankan whiteflies.

CONCLUSIONS
High percentages of the tested whitefly populations have developed resistance to the DOA recommended dosages of thiamethoxam (37−83%), acetamiprid (50−83%), imidacloprid (71%), carbosulfan (15−53%) and etofenprox (68−89%) insecticides. Least resistance was shown to the organophosphate profenofos (2−36%). Both the increased detoxification by esterases, glutathione S-transferases and monooxygenases, and AChE target site insensitivity contribute in varying degrees to the observed resistance. Involvement of kdr type mutations in pyrethroid resistance should be further investigated. This is the first report confirming the resistance status of Sri Lankan whiteflies with underlying resistance mechanisms and the outcome can be utilized effectively for developing a proper whitefly control programme in Sri Lanka avoiding indiscriminate use of insecticides.