Saliva has diverse functions in feeding behavior of animals. However, the impact of salivary digestion of food on insect gustatory information processing is poorly documented. Glucose-aversion (GA) in the German cockroach, Blattella germanica, is a highly adaptive heritable behavioral resistance trait that protects the cockroach from ingesting glucose-containing-insecticide-baits. In this study, we confirmed that GA cockroaches rejected glucose, but they accepted oligosaccharides. However, whereas wild-type cockroaches that accepted glucose also satiated on oligosaccharides, GA cockroaches ceased ingesting the oligosaccharides within seconds, resulting in significantly lower consumption. We hypothesized that saliva might hydrolyze oligosaccharides, releasing glucose and terminating feeding. By mixing artificially collected cockroach saliva with various oligosaccharides, we demonstrated oligosaccharide-aversion in GA cockroaches. Acarbose, an alpha-glucosidase inhibitor, prevented the accumulation of glucose and rescued the phagostimulatory response and ingestion of oligosaccharides. Our results indicate that pre-oral and oral hydrolysis of oligosaccharides by salivary alpha-glucosidases released glucose, which was then processed by the gustatory system of GA cockroaches as a deterrent and caused the rejection of food. We suggest that the genetic mechanism of glucose-aversion support an extended aversion phenotype that includes glucose-containing oligosaccharides. Salivary digestion protects the cockroach from ingesting toxic chemicals and thus could support the rapid evolution of behavioral and physiological resistance in cockroach populations.

2. Materials and Methods

2-1. Two cockroach strains
The wild-type strain (Orlando Normal) was collected in Florida in 1947. The glucose-averse strain (T164) was collected in Florida in 1989 and was artificially selected with glucose-containing toxic bait to fix the glucose-averse (GA) trait in this population. The two strains were maintained on rodent diet (Purina 5001, PMI Nutrition International, St. Louis, MO) and distilled water at 27°C and ~40% RH on a 12:12 h L:D cycle. We tested sexually mature virgin females (5-day old) in bioassays because at this stage females forage extensively in support of oocyte maturation.

2-2. Acceptance-rejection assays and consumption assays
We developed two independent assays to assess feeding.

Acceptance-rejection assay. This was a qualitative yes-no assay that evaluated the instantaneous responses of cockroaches to tastants. It involved a small amount of test solution placed on the paraglossae—the most proximal mouthpart where sugar responsive gustatory sensilla have been observed —and a rapid assessment of either its acceptance or rejection. In this assay, the test solution stimulated gustatory sensilla on the paraglossae, with only brief interaction with saliva. Either one-day starved or non-starved insects were restrained in a plastic pipette tip with only the head exposed. Before starting the assay, insects were satiated with water, except when they were tested with deterrents. The paraglossae, were carefully touched under a stereo-microscope with a 0.3 µl drop of test solution dyed with 1 mM blue food dye (Erioglaucine disodium salt). Acceptance of the taste substance was defined as ingestion within 1 sec, and rejection was defined as lack of ingestion within 1 sec. During ingestion of the stimulus solution the blue dye could be seen through the clypeus and frons, the translucent front-middle area of the head capsule. The cockroaches ingested <0.01 µl of the test solution before the solution was withdrawn to avoid satiation.

Consumption assay. This assay quantitatively measured the amount of test solution ingested by each insect. Non-starved females were individually restrained in a pipette tip as in the feeding assay. A test solution containing 0.5 mM blue food dye in a microcapillary (10 µl, Kimble Glass) was brought in contact with the paraglossae and the volume consumed was recorded. We observed each female until she stopped drinking, and we considered this a single bout of feeding. Therefore, in this assay the test solution stimulated gustatory sensilla on the paraglossae, and more of the test solution had longer contact with saliva.



2-3. Sugar acceptance and consumption
To assess the feeding responses of WT and GA females in the acceptance-rejection assays, monosaccharides (d-glucose and d-fructose), disaccharides (d-maltose, d-trehalose, d-sucrose) and a trisaccharide (d-maltotriose) were tested. All chemicals were prepared at 0.01, 1, 10, 100 and 1000 mM with 1 mM blue food dye. Two groups of females were prepared: Group A was prepared using non-starved females to assess their acceptance responses of phagostimulant sugars. Each insect was satiated with water before the test, to ensure that their acceptance of phagostimulants was not due to thirst; these females would reject the negative control (water with blue dye) and various deterrents. For each sugar 21 WT and 31 GA females were tested. Group B was prepared to assess their rejection responses of various tastants. Females were starved for one day without food and water to produce hungry and thirsty females that were motivated to accept water and phagostimulants, but would reject deterrents at high concentration. Data were obtained from 20 GA females for glucose. The effective concentration (EC50) of each sugar was obtained from dose-response curves. For Group A, dose-response curves and EC50 values were obtained for WT and GA females in response to six and five sugars, respectively. Glucose was tested as a deterrent in GA females in Group B. In the consumption assay, the amount consumed (µl) of six types of sugar solutions with 1 mM blue food dye was measured at 0, 10, 100 and 1000 mM. Each female received a single concentration of a single test solution. For assays using plain water consumption, 26 WT and 13 GA females were used. For sugar consumption, 21–25 females were tested at each concentration of each test solution.

2-4. Effects of saliva on sugar degradation and feeding
To evaluate the effects of saliva mixed with sugar solutions on the feeding response in the acceptance-rejection assay, saliva from 5 day-old WT and GA females was collected. Individual females were briefly anesthetized with carbon dioxide under a microscope and gently squeezed at the side of the thorax to avoid contamination with regurgitates from the gut. Saliva droplets were collected at the mouthparts into a microcapillary (10 µl, Kimble Glass). Collected fresh saliva was immediately used in experiments. Test solutions were prepared by mixing 3 ml of 200 mM maltose, maltotriose, trehalose, sucrose or 600 mM glucose or fructose with 3 ml of either HPLC-grade water or saliva of WT or GA females. The four tri- and disaccharides were presented to females at a final concentration of 100 mM, and the two monosaccharides (glucose and fructose) were presented at 300 mM in a total volume of 6 ml. These concentrations were chosen based on the EC70 in the acceptance-rejection assay. Saliva and sugars were incubated for 5 min at 25oC. We tested 18–22 females in each strain. To evaluate the effect of saliva alone on the feeding response, either non-starved females or females that were starved for one day without food and water were tested with water only and the mixture of saliva and water at a 1:1 ratio (n = 20). All test solutions contained blue food dye.

2-5. Involvement of salivary glucosidases in sugar degradation
To evaluate whether salivary enzymes are involved in the hydrolysis of disaccharides and trisaccharides, the contribution of salivary alpha-glucosidases was tested in the feeding acceptance-rejection assay using the glucosidase inhibitor acarbose (CAS 56180-94-0, Sigma Aldrich). We first confirmed that the range of 0–125 mM acarbose in HPLC-grade water did not disrupt feeding acceptance or rejection of cockroaches. Test solutions were prepared as follows: 2 µl of either HPLC-grade water or saliva of GA females was mixed with 1 µl of either 250 µM of acarbose or HPLC-grade water, then the mixture was added to 1 µl of 400 mM of six different sugar solutions. The total volume was 4 µl with the final concentration of sugar being 100 mM. As before, saliva and sugars were incubated for 5 min at 25°C and all test solutions contained blue food dye. Only 5 day-old GA females (n = 20–25) were tested in each assay.



2-6. Salivary protein and alpha-glucosidase activity
Salivary protein was measured by the Bradford method (Coomassie protein assay kit, Thermo Scientific, 23200, Rockford, IL, USA) to compare differences between sexes and strains. We conducted 6–8 replications with both sexes of the two strains. To estimate alpha-glucosidase activity of saliva, a colorimetric assay using p-nitrophenol (alpha-glucosidase activity colorimetric assay kit, Biovision, K690-100, Milpitas, CA, USA) was carried out. Either 1 µl of HPLC-grade water or 250 mM acarbose was added to 1 µl of saliva and incubated for 5 min at 25 oC. We conducted 5 replications with both sexes of the two strains.