Fipronil affects cockroach behavior and olfactory memory Free

The exponential growth of the human population in the post-war period required a major expansion in crop production and resulted in a massive increase in the use of pesticides applied to crops before or after harvest to protect the commodities from deterioration (Tudi et al., 2021). The use of pesticides has increased ∼50-fold since 1950 and ∼2.5 million tons (corresponding to US$15 billion) are currently used worldwide every year (Gro Intelligence, 2018; Zaller, 2020).

Many sprayed insecticides and herbicides reach non-target species, in addition to contaminating air, water and soil (Mali et al., 2020). Pesticides are also responsible for reducing the general biodiversity and pollinator insect populations (Wells, 2007), degrading habitats by reducing plant pollination and reducing insect resources for birds (Palmer et al., 2007), and threatening endangered species (Gill and Garg, 2014). In addition, pesticide use can damage neighboring agricultural activity by forcing pests towards nearby crops that are free of pesticides, thereby causing harm and potential losses in crop yield (Dyck et al., 2021). Insecticides cause both lethal and sublethal effects in invertebrates and some vertebrates (Rani et al., 2021). Hence, there is a need for a better understanding of the potential risks posed by sublethal doses of insecticides on non-target insect species.

Fipronil (Fpl) is a broad-spectrum insecticide, belonging to the phenylpyrazole family, that acts primarily by blocking gamma-aminobutyric acid (GABA) receptors in insects (Gupta and Anadón, 2018). Fpl is a systemic pesticide that distributes amply throughout plant tissues and is toxic to any insect (and, potentially, other organisms) that feed upon the plant (Simon-Delso et al., 2015). The high to moderate solubility of Fpl and its persistence and leaching potential in soil means that this compound can easily contaminate aqueous environments and affect non-target invertebrates such as honey bees and other important pollinators that drink contaminated water (Bonmatin et al., 2015). Fpl has been found in honey bee hives in southern Queensland (Australia), where it has caused a massive loss of colonies involving the death of >600,000 honey bees (Robinson and Sanders, 2021). This finding agrees with evidence that exposure to sublethal doses of Fpl disrupts learning, memory and orientation in gregarious insects such as honey bees (Pisa et al., 2015). Navigation and foraging behavior are also impaired as Fpl reduces the proportion of active bees in the hive and causes behavioral changes, such as a deficiency in visual learning, that reduce the efficacy of foraging flights (Pisa et al., 2015).

Whereas numerous studies have investigated the neurophysiological effects of Fpl in insects (Durham et al., 2001; Narahashi et al., 2010; Kostromytska et al., 2011; Gols et al., 2020), the influence of sublethal doses of this pesticide, such as commonly encountered in the environment, on insects is still poorly understood. Several studies have examined the distribution and degradation of Fpl in soil and other environments (Gunasekara et al., 2007). Degradation typically takes 111–350 days and in soil, concentrations ranging from 0.000636 to 0.0248 mg g⁻¹ have usually been found (Demcheck and Skrobialowski, 2003). In view of studies on bioaccumulation and residual doses of Fpl in the environment (Tingle et al., 2003; Bhatti et al., 2019; Holder et al., 2018), in this work we investigated the ability of sublethal concentrations of Fpl to disrupt olfactory memory and other behavioral parameters in Nauphoeta cinerea cockroaches. To our knowledge, this is the first study to examine the effect of Fpl on these parameters in cockroaches.

All reagents and solutions were of high purity and were purchased from Sigma-Aldrich Chemical Co. (St Louis, MO, USA). Fipronil 800 WG^® (Fpl) was purchased from BASF Agri-Production SAS (Borborema, São Paulo, Brazil). Fpl was dissolved in 0.9% NaCl (saline solution; stock solution) and diluted (working solutions) in saline solution at pre-set concentrations. All solutions used in the tests were prepared daily, before use, to ensure the stability of the compound to be tested.

All experiments were carried out on adult Nauphoeta cinerea (Olivier 1789) cockroaches of both sexes (3–4 months after the adult molt), with an average weight of 500±30 mg measured during the early morning, that were reared in polyethylene plastic boxes (20 cm×50 cm) in sawdust substrate in a controlled environment at ±25°C, and a 12 h light:dark cycle (on at 07:00 h and off at 19:00 h), with water and food ad libitum (basic food composition: bone and meat meal, viscera meal, animal lipids, ground corn, wheat bran, sodium chloride, flavoring, antioxidant, folic acid, iron, copper sulfate, zinc sulfate, iodine, selenium, choline chloride, potassium, magnesium, niacin, pantothenic acid, vitamin E, vitamin B6, vitamin A, vitamin D3, vitamin K3, vitamin B1, vitamin B2 and vitamin B12). The animals were kept in the above conditions, in groups, until the beginning of each experiment. In the case of the animals subjected to memory assays, the selected insects were kept in the same conditions, but partially deprived of water and food for 15 days. The purpose of this deprivation period was to increase the animals' appeal to the solutions used in the olfactory memory tests. Thus, food and water were offered only once a day, in minimal amounts that were completely consumed, to ensure the necessary resting nutrition for the well-being of the animals.

The effect of Fpl on the biological activity of N. cinerea cockroaches was assessed by injecting the insects with sublethal doses of the insecticide (0.1, 0.01 or 0.001 µg g⁻¹ body mass) directly into the hemocoel, via the third abdominal segment, using a Hamilton syringe in a final volume of 10 µl, unless otherwise specified. Insects were injected with Fpl or with saline solution alone in control experiments.

The grooming behavior of the cockroaches was monitored based on the activity of the legs and antennae, which was recorded for 30 min in a controlled environment at 24–25°C, essentially as described by Stürmer et al. (2014). The tests were carried out individually and for each treatment, including the control saline group, n=30 animals were used in multiple trials and never reused again in other protocols.

The locomotory activity was assessed by recording the trajectory of the cockroaches as described by Leal et al. (2018). To accomplish this, four insects each time were randomly selected and placed individually in a white polystyrene box (25 cm in length×15 cm in width×7 cm in height). Their exploratory behavior was recorded over 10 min with a Logitech^® HD Webcam connected to a desktop computer (Dell, São Paulo, Brazil) that recorded and retrieved the videos for posterior analysis. Animal tracking was observed using the software idTracker (Stoelting, Denver, CO, USA) and the data were analyzed using the Insect Locomotion Tracking Program ILTP^®, a freely available software developed by our research group that can be downloaded at http://sites.unipampa.edu.br/gomndi. With a script specially developed for this purpose, the following were analyzed: distance traveled (cm); number of immobile episodes (n) and time spent immobile (%). The control and treated groups were injected with saline solution or Fpl 10 min before the beginning of data acquisition. The tests were carried out individually in groups of 4 animals, each one in an individual box and for each treatment including the control saline group, n=30 animals were used and never reused again in other protocols.

The in vivo metathoracic coxal-adductor nerve–muscle preparation was mounted as described by Carrazoni et al. (2017). For this, the cockroaches were immobilized by chilling (5–7 min) and then fixed ventral-side up on a stage covered with soft rubber. After fixation, one of the third metathoracic legs was tied to an isometric force transducer (AVS Instruments, São Carlos, SP, Brazil) and indirectly electrically stimulated via the nerve fiber (0.5 Hz/5 ms). Muscle twitches were recorded for 120 min using a data acquisition and analysis system (AQCAD and ANCAD, respectively; AVS Instruments). The tests were carried out individually and for each treatment including the control saline group, n=6 animals were used and never reused again in other protocols.

The effect of Fpl on N. cinerea heart rate was evaluated as described by Leal et al. (2020). Briefly, adult cockroaches were immobilized by chilling (5–7 min) and placed ventral-side up on a dissection plate. The lateral margins of the abdomen were cut along each side, and the ventral abdominal body wall was pulled out to show the viscera. After moving the viscera carefully aside, the heart was exposed still contracting, while attached to the dorsal body wall. The heart preparations were washed by bathing in 200μl of saline solution at room temperature (21–24°C). After 5 min of heartbeat stabilization, the treatments were delivered by exchanging the bathing solution. The average heart rate (beats min⁻¹) in the first 5 min was taken as a reference. Heart rate (frequency) was monitored for 35 min, under a 1000× Digital Microscope (Shenzhen Huishixin Technology Co., Shenzhen, China), connected to a desktop computer (Dell) that recorded and retrieved the videos for posterior analysis. The responses to saline alone or Fpl were monitored from the 5th to the 30th minute, and the preparation was washed several times from the 30th to the 35th minute to remove the saline or Fpl and to verify the possible reversal of the toxic effects. The tests were carried out individually and for each treatment including the control saline group, n=3 animals were used and never reused again in other protocols.

The olfactory memory assays were done as described by Matsumoto and Mizunami (2000) with a few modifications. Before the assays, the cockroaches were deprived of food and water for 15 days. The tests were subsequently done in an apparatus (dimensions: length: 29 cm; width: 14.4 cm; height: 6.8 cm) developed specifically for olfactory tests (Fig. 1). Initially, the cockroaches were placed in a compartment at one end (labeled A) and allowed to adapt to the surroundings for 5 min. Subsequently, a gate (B) was lifted, giving access to the arena (C) where a rotating cylinder (D) simultaneously offered two odor sources placed inside a well (E).

The odor sources used were vanilla essence and citronella diluted 1:3 in milli-Q water. Inside each well was placed a piece of cotton (1 cm×1 cm) soaked with 500 µl of the odor sources alternately in the cylinders. Each cylinder was capped with a circle of absorbent paper. For the memory tests, 100 µl of 40% sucrose solution was placed on the absorbent paper of cylinders with citronella odor source and 100 µl of 40% NaCl solution on the paper of cylinders with vanilla odor source. This way, the animals associated the repulsive odor with the reward (40% sucrose solution). The assay began as soon as the cockroach entered the arena and involved recording the time spent probing each odor source during a 10 min period, with the position of the wells being inverted every minute. The memory test was basically divided into two phases that consisted of preference tests and olfactory memory tests. The preference tests served as positive controls to confirm the hypothesis that cockroaches are naturally attracted to vanilla odor and repelled by citronella odor. The tests were carried out individually and for each treatment including the control saline group, n=30 animals were used and never reused again in other protocols.

The results are expressed as the mean±s.e.m. As indicated in the figures, statistical comparisons between two experimental groups were done using Student's t-test; one-way ANOVA followed by Dunnett's test was used to compare treatments only with the control group; for comparisons when the y-axis had more than one variable, two-way ANOVA followed by the Bonferroni post hoc test was used, with all groups being compared with the saline control. P≤0.05 indicated significance. All statistical analyses were done using Prism v.7.0 software (GraphPad, San Diego, CA, USA).

The exposure of N. cinerea to sublethal doses of Fpl (0.1, 0.01 and 0.001 μg g⁻¹ body mass) significantly increased the leg grooming time, but did not markedly affect the antennal grooming time (Fig. 2).

Fpl caused a dose-dependent locomotory deficit (Fig. 3D), with an overall reduction of 46% in the total distance traveled by the cockroaches after injection of the highest dose (0.1 µg g⁻¹) (Fig. 3A). All three doses markedly increased the number of immobile episodes (defined as more than 3 s without movement; ∼200% increase in this parameter at the highest dose) (Fig. 3B), in addition to causing lethargy in the cockroaches (∼80% increase in time spent immobile, i.e. time without movement during the total experiment, at the highest dose compared with the saline control) (Fig. 3C).

In cockroach metathoracic coxal-adductor neuromuscular preparations, only the highest dose of the pesticide (0.1 µg g⁻¹) significantly affected the twitch tension and caused a maximal decrease of ∼45% in the contractile response in 50 min of experiment (Fig. 4A). Incubation with Fpl did not affect the baseline tension of the preparations, i.e. there was no muscle contracture independent of electrical stimulation (Fig. 4B).

In cockroach semi-isolated heart preparations, Fpl (0.001–0.1 µg 200 µl⁻¹) caused irreversible, dose-dependent bradycardia at all doses. In control conditions, in the absence of Fpl, the heart rate (frequency) was considered to be 100%. Thus, the maximal mean heart rates were 70.3±10.6%, 75.8±23.4% and 70.3±28.8% after 30 min for doses of 0.001, 0.01 and 0.1 µg 200 µl⁻¹, respectively (Fig. 5). Extensive washing of the preparations did not lead to a recovery of heart rate, indicating the irreversible nature of the cardiac effect.

The preference test confirmed that N. cinerea cockroaches naturally prefer the odor of vanilla rather than citronella (Fig. 6A). However, when citronella was offered in a 40% sucrose solution, the olfactory memory test showed that the odor preference became similar between citronella and vanilla (Fig. 6B). Treatment with sublethal doses of Fpl (0.001, 0.01 and 0.1 µg g⁻¹) markedly decreased the olfactory memory for all compounds offered at all doses tested (Fig. 6C).

The findings described here indicate that sublethal doses of Fpl adversely affect several physiological systems of N. cinerea cockroaches and cause important behavioral changes. Exposure to Fpl increased the frequency of leg grooming but not that of antennal grooming. Grooming in insects is a fundamental activity associated with body cleanliness, courtship, social signaling, movement activity and bodily excitement (Spruijt et al., 1992; Zhukovskaya et al., 2013), and is modulated primarily by octopaminergic and dopaminergic neuronal pathways (Weisel-Eichler et al., 1999; Libersat and Pflueger, 2004; Leal et al., 2018, 2020). Changes in grooming behavior are related to fundamental factors associated with the survival and persistence of insects in the environment and can influence a species’ fertility and longevity (França et al., 2017). The alterations in grooming behavior observed here suggest a direct effect of Fpl on the central nervous system that involves the modulation of octopaminergic and dopaminergic neurotransmission, possibly in a similar manner to other toxins that we have studied (Stürmer et al., 2014; Barreto et al., 2020).

The influence of Fpl on exploratory activity was assessed by examining changes in the locomotor behavior of cockroaches. Specifically, Fpl reduced the distance walked by cockroaches and increased the number of immobile episodes and time spent immobile. Fpl also affected leg grooming behavior. These findings strongly suggest that sublethal doses of Fpl adversely affect the activity of the central nervous system, leading to locomotory and exploratory deficits. Alterations in the locomotion and exploratory behavior of insects can increase their susceptibility to predators (Marliére et al., 2015) as locomotion is the greatest defensive strategy used to avoid and escape predators in the environment (Adedara et al., 2016). In bees, for example, locomotion, including flight, is responsible for maintaining colony homeostasis through behavioral mechanisms and pheromones (Bortolotti and Costa, 2014). Flight activity is also responsible for recognition of the environment, including biotic factors (pheromones, diseases, stress, etc.) and abiotic factors (rain, food, temperature, etc.) that are important for colony survival, and its failure is associated with swarming and hive abandonment. Thus, behavior and locomotion are essential for insect maintenance and survival (Mizutani et al., 2021).

Although leg grooming behavior is mostly associated with octopaminergic neurotransmission, motivation to walk is dependent on dopamine, an important excitatory neurotransmitter (Stürmer et al., 2014; Borges et al., 2020). Fpl is a well-known agonist of the main inhibitory neurotransmitter, GABA (Byrne, 2019), and sublethal doses of this pesticide can persist in the insect neuromuscular junction (Chapman, 2012). For this reason, we cannot exclude the possibility that the mechanism behind the Fpl-mediated decrease in exploratory activity may include an inhibitory effect on peripheral neuronal activity.

Indeed, experiments with neuromuscular preparations showed that even very small sublethal doses of Fpl caused a reduction in the contractile activity of N. cinerea leg muscle. The effects observed here in cockroaches resemble those seen with sublethal doses of Fpl in the behavior of gregarious insects such as bees. In the case of bees, the remaining activity of sublethal Fpl doses might affect motor activity, which is important for foraging and orientation during the ‘dance’ used to indicate the location of resources to the colony (El Hassani et al., 2005). Thus, a Fpl-induced decrease in the contractile response would disturb the exploratory activity of those pollinators, causing disorientation and interference with the maintenance of honey bee colonies (Bovi et al., 2018).

The bradycardia caused by sublethal doses of Fpl suggests the involvement of octopamine, the main neurotransmitter associated with the modulation of heart rate in insects (Hillyer, 2018). High concentrations of octopamine cause tachycardia, whereas low concentrations are associated with bradycardia (Papaefthimiou and Theophilidis, 2011). In contrast, the role of other neurotransmitters such as acetylcholine (ACh) in insect heart rate is still controversial. Some studies have shown that ACh increases heart rate, while others suggest the opposite effect (Claros-Guzmán et al., 2020). Resolution of the relative contributions of these neurotransmitters to cardiac function will require the use of pharmacological interventions to modulate the octopaminergic and cholinergic pathways, as well as an assessment of acetylcholinesterase activity.

Another important finding of this work was that sublethal doses of Fpl interfered with olfactory memory learning and formation in N. cinerea cockroaches. Olfaction is an important sense used by insects, such as cockroaches and bees, and is closely related to behavior and physiological homeostasis. In cockroaches, which are essentially nocturnal animals, communication occurs mainly through strategies based on smell (involving pheromones and odorant bacteria) and touch (Gullan and Cranston, 2014). The insect olfactory system is one of the most developed among animals and acts by converting a chemical into an electrical signal. Physiologically, the electrical impulses generated through excitation of the system by odorous molecules are conducted from the antennal nerve to the antennal lobe (glomeruli). The information captured through this pathway is directed to a region of the animal's brain known as the ‘mushroom body’ that is associated with learning and memory processes (Modi et al., 2020). Adult cockroaches can associate aversive odors, such as mint, with a reward, such as sucrose solution, and preferential odors, such as vanilla, with a punishment, such as hypertonic saline solution (Watanabe et al., 2003).

Studies have shown that the reward-associated processing is mainly related to octopaminergic neuromodulation, whereas aversion behavior is related to dopaminergic neuromodulation (Gauthier and Grünewald, 2012). Glutamatergic receptors are also involved in learning processes, memory formation and retrieval (Gauthier and Grünewald, 2012; Leboulle, 2012). In the mushroom body, third-order neurons (Kenyon cells, KCs), which play critical roles in olfactory learning (Menzel et al., 2006; Liu et al., 2012), respond only to specific odors with a few spikes and result in sparse odor coding (Gupta and Stopfer, 2014; Lin et al., 2014). One mechanism that contributes to the formation of spatially and temporally sparse odor representations in populations of KCs is a widespread and broadly tuned GABAergic inhibition that feeds signals from KCs back to KCs (Szyszka et al., 2005; Papadopoulou et al., 2011).

The precise mechanism by which sublethal Fpl disturbs olfactory memory learning and formation is unclear but may be associated with GABAergic interneurons and the blockage of chloride channels in GABAergic and glutamatergic neurons. Thus, if sublethal doses of Fpl cause small disturbances in GABAergic interneurons, this could create hyperexcitation, thereby preventing feedback signals to KCs, leading to a destabilization of physiological processes involved in olfaction and associated memory formation. Future experiments involving site-specific intracellular recordings in mushroom bodies and other central structures of the N. cinerea central nervous system may help to elucidate the mechanism involved in the Fpl-induced disturbance of olfactory memory.

The results of this work demonstrate that sublethal doses of Fpl cause striking alterations in the physiology and behavior of N. cinerea cockroaches that affect the cardiovascular system together with olfactory, exploratory and locomotory behavior. These alterations involve mostly central nervous system pathways, probably by affecting octopaminergic and GABAergic neurotransmission. Overall, these findings show that the effects of sublethal doses of Fpl on insect behavior should not be neglected and they reinforce the need to examine the toxicological effects of sublethal doses of insecticides prior to their approval for general and agricultural use.

We thank Dr João Batista Teixeira da Rocha of the Federal University of Santa Maria (UFSM), Santa Maria, RS, Brazil, for providing some of the reagents used in this work. We also thank the Centro Integrado de Pesquisas em Biotecnologia-CIPBIOTEC for offering all the necessary infrastructure in terms of equipment and technical staff for the proper conduct of this work.