Toxicological and morphological analysis of Africanized Apis mellifera selected for tolerance to the neonicotinoid thiamethoxam

Honeybees are the insect most used for pollination purposes due to its efficient characteristics for this function, which reflects in positive aspects for both nature and man. The expansion of agriculture and the development of agrochemicals to combat pests has had negative impacts on honeybee health, causing its disappearance around the world. This research aimed to evaluate the effects of honeybee exposure to the neonicotinoid insecticide thiamethoxam on its survival rate, as well as on morphological and histological changes in the midgut of adult workers from the F4 generation of Apis mellifera queens tolerant to thiamethoxam and non-tolerant worker honeybees. After the bioassays, the midgut was removed for morphological evaluation. The results showed that F4 bees were more tolerance to thiamethoxam and exhibited less significant morphological changes when compared to the non-tolerant control group.


Introduction
In 1956, African honeybees (Apis mellifera scutellata) arrived in Brazil from Africa. In 1957, 26 of those colonies escaped and mated with the European subspecies Apis mellifera ligustica from Italy, Apis mellifera mellifera from Germany and Apis mellifera carnica from Austria, which already inhabited Brazil, emerging Africanized honeybees or polyhibrid honeybees (Kerr, 1967).
The Africanized population was able to adapt to the tropical climate, developing characteristics for its survival, such as temperature control, nest organization and tolerance to some diseases and pathogens (Bourgeois et al., 2016).
Bees are of great importance to the ecosystem, pollinating several species of plants, including crops, increasing production. A. mellifera is one of the best-studied species in the world, due to its significant services to the environment and its disappearance, due to factors such as deforestation, climate change and indiscriminate use of agrochemicals, is raising concern (Stein et al., 2017;Sanchez-Bayo et al., 2016;Oliveira et al., 2018).
The excessive use of agrochemicals in doses above the permitted favors the significant loss of pollinating insects, which can compromise the development of plant species and the survival of all living organisms that depend on plants for a living (Sanchez-Bayo et al., 2016;Oliveira et al., 2018;Chambers et al., 2019).
Increasing honey production and disease tolerance requires a great amount of work from those who manage the apiary, in addition to the interaction of the genetic with the environment. Knowing the responsiveness of individuals to the environment favors genetic enhancement (Zakour & Bienefeld, 2015;Bourgeois et al., 2016;Hassett et al., 2018).
Queen exchange is a process to develop and implement breeding programs, bringing many advantages, such as faster production, allowing colony monitoring during the breeding program. For breeding programs, some characteristics are primal, such as laying capacity, diseases tolerance and longevity (Zakour & Bienefeld, 2015;Bourgeois et al., 2016;Hassett et al., 2018).
Several factors directly affect honeybee lives, for instance, climate, since low temperatures associated with the damage promoted by agrochemicals result in serious problems, such as the nest abandonment. Low temperatures bring a restriction on the supply of food in the field, and for its survival, bees seek to adapt to the climate and its thermoregulatory Research, Society andDevelopment, v. 10, n. 2, e14310212109, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i2.12109 3 mechanisms are activated, which expends a large amount of energy (Suryanarayanan & Kleinman, 2013;Seitz et al., 2015).
After winter, bees forage on plants that were treated with agrochemicals, since crop production occurs particularly during summer and spring, which harms bee development, which can lead to death (Chambers et al., 2019).
The midgut is the gateway to food and the first access route for pathogens and substances that causes damage, so it is an important region for morphological research. As all substance absorption occurs in the midgut, morphological changes caused by the insecticide in this region of the digestive system can compromise other physiological and behavioral processes (Williamson et al., 2014;Tavares et al., 2015;Catae et al., 2014).
Honeybees exposed to agrochemicals may not be able to return to their colony because they do not maintain their sense of guidance and location, becoming a global problem due to the large-scale use of these products (Williamson et al., 2014;Oliveira et al., 2018).
Studying and understanding the action of these agrochemicals on bees it is essential to minimize the impact caused by exposure to insecticides, reducing the lethal and sublethal action effects on pollinators (Carreck & Ratnieks, 2014;Williamson et al., 2014;Oliveira et al., 2018).
Mortality tests and critical electrolyte concentration (CEC) analysis showed that bees were tolerant of the agrochemical (Pizzaia et al., 2021). Currently, in the apiary, the fourth generation of bees descendant from queens tolerant to thiamethoxam (F4) are being kept. Therefore, the study aimed to evaluate the toxicity and morphological changes in the midgut of F4 adult A. mellifera workers tolerant to the neonicotinoid thiamethoxan.

Methodology
This quantitative research is characterized by the use of quantification, both in the collection as in the treatment of information, using statistical techniques (Richardson, 1999).
Two different types of adult A. mellifera workers were collected, first the ones from the F4 generation of queens tolerant to thiamethoxam and second honeybees without tolerance induction. Bees were kept in plastic bottles measuring 9x7x26cm, with perforations on the lids to allow air to enter. Then bees were taken to the Laboratory of Genetics Animal from the Department of Biotechnology, Genetics and Cell Biology of UEM.

Bioassays
For the bioassays, the commercial insecticide Actara 250WG (Syngenta) was diluted in water, as recommended for application on Citrus crops, producing a solution with 250 grams of active ingredient per kilogram. The final concentration used to feed the bees were 5x10 -5 grams of active ingredient per milliliter (g a.i./mL).
Tolerant workers were feed only with candy (bioassays 1) and feed with candy containing the agrochemical (bioassays 2). Non-tolerant workers were feed only with candy (bioassays 3) and feed with candy containing 2 μl of the agrochemical (bioassays 4). The bioassays were performed with three replicates each and 20 bees per repetition for periods of 24, 48 and 72 h. The bees were kept in glass jars for the bioassays (diameter 14.71 cm x 18.5 cm tall). The glass bottles contained filter paper, a water-soaked cotton swab, and another candy container. Bees were stored in a BOD incubator, maintained in 33 ºC ± 2 °C and UR 80 ± 10%. After each period dead honeybees were counted.

Light microscopy
After 24, 48 and 72 h of the bioassays, surviving bees where anesthetized under cold temperatures and the midgut was dissected in solution (0.1 M NaCl, 0.1 M Na2HPO4, and 0.1 M KH2PO4).
The midguts were fixed in aqueous Bouin's solution for 12 h, dehydrated in an alcohol series of increasing concentrations (70%, 80%, 90%, and 100%), diaphanized in xylene, paraffin-embedded and sectioned into 6 µm slices using a microtome Leica RM 2250. Then, sections were spread on glass slides, rehydrated, and stained with hematoxylin and eosin (H/E) (Junqueira & Junqueira, 1983). The analyses were performed under an Olympus light microscope and sections were photographed using a digital camera.

Data analysis
Statistical analysis was performed using the Statistical Analysis System software (SAS, 2012). Initially, a descriptive analysis was conducted for survival bees between treatments and groups of bees. To test the effect of the treatment, the model with interaction in the main effects using variance analysis (ANOVA) -subdivided plots -was used. The Tukey test was applied to verify the multiple differences between the levels of the variables.
Treatments were divided into two groups, F4 generation and control, with three periods of time (24, 48 and 72 h) and two types of food (with and without the agrochemical each). The results for bee survival were analyzed in a factorial scheme 2 x 3 x 2.
Thus, the variables survival, interactions between levels and significant interaction were analyzed through combinations between factors and the F test was used to identify significant differences. The hypothesis of normality of the error and the homogeneity of variances was evaluated by the Shapiro-Wilk test and the Bartlett test, respectively. In all tests, the level of significance was 5% significance (p-value<0.05).

Toxicity in workers tolerant to thiamethoxan
After oral contamination of A. mellifera with thiamethoxam, workers survival was affected, the average and median of survival levels were higher in the candy treatments, especially in the control group ( Figure 1). However, the control group exposed to the agrochemical at the concentration of 5×10 -5 g a.i./mL showed the lowest mean and median survival levels, but also showed more variability (Figure 1a). Regarding the mean and median of survival levels between the experimental periods, the highest levels were observed after 48 and after 72 h higher variability was showed.

Source: Authors.
For this analysis, the effect model in subdivided plots was defined by: After oral contamination of tolerant A. mellifera (bees that received treatment in the field) and the workers who are not part of this group (bees that did not receive treatment in the field), at the concentration mentioned above, for the periods of 24, 48 and 72 h, there was a significant survival level on both groups (Table 1). Regarding the percentage of explainability of the adjusted model, it presented a value of R²= 0.9536, proving to be adequate for the analysis. Concerning the assumptions of errors, there was no violation of the hypothesis of normality and homogeneity of variances (p-value>0.05).
With the results observed in ANOVA, it was verified which variables related to the analyses presented a significant difference and between which factors an interaction occurred. A significant difference was observed for the variables Groups of Bees and Agrochemical and there was a significant interaction between these variables indicating that the response of the Groups of Bees factor is dependent on the Agrochemical factor (Table 1).
The results allowed to verify the significant effects of the interaction between the factors (Table 1). It was observed the effect of the Groups of Bees factor within the levels of the Agrochemical/Candy factor, F4 tolerant A. mellifera workers fed only with candy showed a survival rate of 97.22% and the control workers presented 97.77% of survival rate on in vitro tests.
Regarding the mean of tolerant workers survival fed with candy and the agrochemical, it was observed a significant survival rate of 91.66% when compared to the control group that received the same treatment (52.66%). Comparing the survival means of bees that were fed only with candy and the bees that received food with the agrochemical, it was found that there is proximity in the results and that F4 bees were more tolerant to the agrochemical exposure. However, field tests will be required to corroborate these observations. It is noteworthy that the bioassays were performed with the concentration of active ingredient of thiamethoxam indicated for cochineal pest on Citrus, therefore it is a concentration of neonicotinoid that workers usually encounter during foraging.
During bioassays, some worker behaviors were observed when fed with candy containing thiamethoxam. F4 workers and control consumed most of the candy contaminated with the agrochemical, showing that the at the concentration of 5x10 -5 g a.i./mL the workers did not notice the presence of the insecticide (Falco et al., 2010;Kessler et al., 2015). Kessler et al. (2015), performed a bioassay with two sources of food, pollen and Biogluc, a type of food for A. mellifera and Bombus terrestris. The authors found that these two species did not avoid relevant concentrations of nectar contaminated with three of the most commonly used neonicotinoids (imidacloprid, thiamethoxam and clothianidin). Also, bees of both species prefer to eat more sucrose solutions mixed with imidacloprid or thiamethoxam than with sucrose alone. The results of this study also indicated that bees did not taste neonicotinoids and are not repelled by them.
Bees preferred solutions containing imidacloprid or thiamethoxam, although the consumption of these agrochemicals made them ingest less food in general. Furthermore, stimulation with these neonicotinoids did not cause spasmodic responses of gustatory neurons in the oral parts of bees, nor inhibited the responses of sucrose-sensitive neurons. Kessler et al. (2015) concluded that bees did not control their exposure to neonicotinoids in food, this fact implies that the treatment of flowering crops with imidacloprid or thiamethoxam poses a considerable danger to foraging bees.
Control worker bees fed with thiamethoxam-contaminated candy showed difficulties in locomotion (walking and flight), tremors and spasms, while thiamethoxam-tolerant F4 workers manifested these changes at a lower frequency. Research, Society andDevelopment, v. 10, n. 2, e14310212109, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i2.12109 7 Contamination of A. mellifera with sublethal concentrations of thiamethoxam can reduce the life expectancy of these insects . These authors experimentally verified that in a group of bees exposed to 1/10 of the thiamethoxam LC50 there was a 41.2% life expectancy reduction.
The results suggest that the tolerance to neonicotinoids can occur relatively quickly in A. mellifera, which can be harmful to bee products because forages can carry contaminated materials to the colony, which will be sold to the general public. Furthermore, we are not yet aware of the long-term consequences of this contamination for bees.

Midgut morphological alterations
The digestive tube of bees is divided into three parts, foregut, midgut and the hindgut. The A. mellifera midgut has a tubular shape and starts in the heart valve reaching the pylorus. It is surrounded by two layers of slender muscle fibers: the circular ones forming the folds of the midgut and the longitudinal ones with numerous branches, in the open weft (Snodgrass, 1956;Landim, 2009). The intestine is the organ responsible for the digestion and absorption of nutrients, being an essential region for toxicity analysis.
After 24, 48 and 72 h of feeding the bees only with candy, it was observed through scanning electron microscopy that the muscle fibers circular muscle fibers arranged internally and as longitudinal outermost of the outer part of the midgut showed organized musculature and without disruption (Figures 2 A-C).
When bees from the control group were submitted to contamination by ingestion with thiamethoxam for 24 h the muscle fibers, mainly the longitudinal of the midgut became even tick (Figure 2 D). After 48 h it was observed muscle fibers disorganized and deformities in the musculature (Figure 2 E). After 72 h, the muscle fibers of the midgut became even tinner when compared to control (Figure 2 F).
It was also observed the external musculature of the F4 worker midgut exposed only to the candy food, which presented thicker muscle fibers (Figure 3 A-C), when compared to the field control (Figure 2 A-C), probably an adaptation to contamination with neonicotinoid, because these bees are being kept in hives with contact with the agrochemical since the first generation, a process initiated five years ago (2015). After exposure for 48 and 72 h to the agrochemical, there was a reduction in the thickness of midgut muscle fibers (Figures 3 E-F), resembling the musculature of the non-tolerant group (Figure 2).
All alterations observed were in response to the effect of the agrochemical thiamethoxam in which the dosage applied (5x10 -5 g a.i./mL) was extremely harmful to the insect when administered by ingestion. Since the intestinal epithelium rests on this musculature, the reduction of muscle fibers observed in longer periods of exposure in tolerant bees, as well as, in bees not tolerant to thiamethoxam, can lead to the morphological compromise of this organ, as well as cause damage to its functioning and, consequently, the survival of bees.
According to Landim (2009), the midgut epithelium is made of digestive cells supported on the basal lamina. It shows homogeneous cytoplasm, a nucleus in the middle-apical region facing the lumen, covered by microvilli (Figures 4 A-C).
In the basal region of the epithelium, it was observed the presence of regenerative cells organized in regeneration nests (Figure 4 A-B) these cells are related to the replacement of the main cells (Teixeira et al., 2013).  In the lumen, the peritrophic membrane acts as a mechanical barrier (Figure 4 B-C). It is made with chitin and proteins called peritrophins and coats the wall of the intestinal epithelium, involving the food that passes through the light of the ventricle, moving the cells away from friction with the solid food. Thus, the peritrophic membrane consists of two compartments, an internal side called endoperitotrophic space, which is in contact with food and the outer membrane that is called ectoperitotrophic space and is in contact with epithelial cells (Moussian, 2013;Teixeira et al., 2013). Besides, the peritrophic membrane was disorganized and regenerative cells were not observed (Figure 4 E-F). The absence of these regenerative cells may indicate the attempt to be differentiated into new digestive cells to replace those that were damaged, regenerating the epithelium, in order to keep the bees alive for longer.
In F4 honeybees fed only with candy, there was organized prismatic cells with nucleus evidencing the organization of the epithelium. In its basal region, it was observed regenerative cells indicating cell activity for regeneration, with the presence of peritrophic membrane in the lumen (Figure 5 A-B). The musculature remained organized (Figure 4   The destructive effects of thiamethoxam were highlighted with the presence of vacuoles, muscle fragile, increased secretion, membrane rupture, which is linked to the release of digestive cells . The effects of the agrochemical on honeybee are adverse and may affect several aspects of their life. Some responses occur through physiological and morphological changes altering musculature, epithelium and peritrophic membrane, which results in harmful behavior effects, impairing usual tasks, such as food collection, compromising colony survival for interfering with other physiological processes Roat et al., 2013), including changes in oxygen supply, synapses, neuronal degeneration, which may explain deficiencies in learning and memory processes of bees (Catae et al., 2018). Similar effects were identified by Diniz et al. (2020) after oral contamination of Scaptotrigona bipunctata bees with a combined insecticide containing the neonicotinoid and pyrethroid chemical groups. In this study, several morphological changes were observed in the midgut, such as the degradation of the peritrophic membrane, epithelium disorganization and the musculature loosening.
F4 honeybees tolerant to thiamethoxan showed significance tolerance to the agrochemical, with a higher survival rate when compared to the non-tolerant control group. Also, they responded better to contamination, when compared to honeybees that were not previously exposed to the agrochemical in the field.

Conclusion
Adult F4 A. mellifera workers showed tolerance to the neonicotinoid thiamethoxan, due to the estimated interaction with the effect of the agrochemical and survival of honeybees that contaminated by ingestion. Through scanning electron microscopy and light microscopy, morphological changes were detected in the midgut of contaminated honeybees. However, F4 tolerant bees showed less significant changes and a faster recovery in the internal morphology of the midgut after 72 h of thiamethoxan contamination.