Topical treatment of adult house flies, Musca domestica L. (Diptera: Muscidae), with Beauveria bassiana in combination with three entomopathogenic bacteria

2.1 Entomopathogens

The L90 strain of B. bassiana was obtained from Dr. Drion Boucias at the University of Florida Institute of Food and Agricultural Sciences (UF/IFAS), Entomology and Nematology Department. This strain was originally isolated from wild caught house flies in upstate New York and is known to be virulent to adult M. domestica (Geden et al. 1995; Johnson et al. 2019; White et al. 2021). The Db11 strain of S. marcescens was obtained from the Caenorhabditis Genetics Center (CGC) of the University of Minnesota (Saint Paul, Minnesota). This strain is a streptomycin-resistant mutant of strain Db10 (Iguchi et al. 2014) and was chosen because while S. marcescens is ubiquitous in the environment this strain was shown to be entomopathogenic (Flyg et al. 1980). The NC19 strain of P. temperata was provided by Dr. Byron Adams at Brigham Young University (Provo, Utah). This strain was selected because previous work demonstrated its insecticidal properties, and its genome has been sequenced and is predicted to contain many genes for insecticidal toxins (Hurst et al. 2015). This species has a symbiotic relationship with Heterorhabditis entomopathogenic nematodes, in which the bacteria colonize the nematode gut and then once the nematode infects the insect the bacteria are regurgitated where they replicate, eventually causing mortality of the insect (Clarke 2014). The Pf-5 strain of P. protegens was purchased from the American Type Culture Collection (ATCC; Manassas, Virginia). This strain was formerly classified as Pseudomonas fluorescens Migula but was reclassified as P. protegens as it was found to cluster away from the fluorescent pseudomonads (Takeuchi et al. 2014). This strain was selected due to its entomopathogenic qualities, as demonstrated in Hermetia illucens L. (Diptera: Stratiomyidae) (Shah et al. 2023).

2.2 House fly rearing and handling

The Orlando Normal strain was used for these studies. This insecticide-susceptible strain had been reared at the Center for Medical, Agricultural and Veterinary Entomology, United States Department of Agriculture, Agricultural Research Service (Gainesville, FL) since 1958. Flies were maintained at 28 °C in wire mesh cages (37.5 × 37.5 × 37.5 cm), and fed a diet consisting of dried milk, sugar, and dried egg yolk in an 8:8:1 ratio by volume. Female flies used in the bioassays were less than 3 days old, presumed to have mated, and were aspirated from rearing cages, lightly sedated with CO₂ to sort and count by sex, and topically treated with pathogen suspensions. Treated flies (20 flies per treatment) were placed into screen-covered 500 ml plastic containers (Instawares, Kennesaw, Georgia) containing 50 mg of diet and a 30 ml plastic container with water and a lid with a dental wick (to prevent flies from drowning), and held at 28 °C.

2.3 Bacterial and fungal cultivation

Bacterial strains were cultured on Luria-Bertani (LB) agar plates (LB broth, Fisher BioReagents, Pittsburgh, Pennsylvania), incubated at 28 °C for 48 h then stored at 4 °C. For preparation of bacteria inocula for bioassays, cultures were set up a day before and left overnight. Cultures consisted of 3 mL LB broth in a 14 mL round bottom tube with the respective bacterial colony picked with a sterile loop from the refrigerated agar plates, then the tubes were incubated in a controlled environment shaker (New Brunswick Scientific, Edison, New Jersey) at 28 °C and 250 rpm. The following morning, the culture tubes were transferred into a 50 ml glass flask at a 1:20 dilution with fresh LB broth and then grown while shaking at 28 °C and 250 rpm until reaching an optical density (OD) of 0.5 as measured using a spectrophotometer (absorbance set to 600 nm; Biochrom LKB Ultrospec II; Cambridge, UK). According to Johnson et al. (2019), 0.5 OD₆₀₀ equates to 1.5 × 10⁸ cfu/ml in P. protegens, 3.0 × 10⁸ cfu/ml in S. marcescens and 2.1 × 10⁸ cfu/ml in P. temperata.

B. bassiana was cultured on Sabouraud dextrose agar with yeast extract (SDAY; 2 % glucose, 1 % peptone, 0.5 % yeast extract; pH 7.0) at 24 °C for 7 days to produce a dense lawn of fungal conidia on the surface of the plate. Plates were dried in a sterile biosafety cabinet for an additional 7 days. After drying, conidia were scraped from each plate with a sterile small metal spatula and stored at 4 °C in a sterile glass vial for up to 4 weeks. Conidial concentrations (conidia/mg) were determined for each batch by suspending 10 mg of dried conidia in 0.1 % Tween^® 20 (Sigmal-Aldrich, Saint Louis, Missouri) and distilled water and using an automated cell counter (Cellometer^® Vision HSL; Nexcelom Bioscience LLC, Lawrence, Massachusetts) to determine the conidial concentration.

2.4 Simultaneous applications of B. bassiana and bacteria

Pathogen suspensions were prepared in a 0.5 % solution of the surfactant, CapSil^® (Aquatrols, Paulsboro, New Jersey) with 1X phosphate buffered saline as the diluent. The surfactant and concentration were chosen due to a previous study that found it to spread well on the fly thorax, cause low to no mortality of house flies, and promote growth of the bacteria species to be tested in the current study (Johnson et al. 2019). A concentration of 1 × 10⁶ of B. bassiana conidia/µl was used as this concentration was known to cause high but less than 100 % mortality (Johnson et al. 2019). Although higher concentrations of B. bassiana would cause 100 % mortality this was not desirable as this would not allow detection of increased mortality in the presence of bacteria. The same concentration (1 × 10⁶ cfu/μl) of bacteria was used in all treatments. For each trial, 0.5 % CapSil was prepared that contained either no microorganisms (control), B. bassiana alone (10⁶ conidia/μl), the bacteria alone (10⁶ cfu/μl), or the two pathogens combined so that 1 μl of the combination contained 10⁶ of both B. bassiana conidia and bacteria cfus. Female flies were immobilized with cold and treated by pipetting 1 μl of the suspensions onto the dorsal thorax. Treated flies were transferred to small observation containers with food and water, and dead flies were counted daily for 7 days. The experiment was replicated on three occasions with different generations of flies and pathogen suspensions using 20 flies per treatment per replication (60 flies per treatment, n = 3).

2.5 Sequential applications of B. bassiana and bacteria

The same concentrations described in the previous section were used in sequential topical applications. In these tests, flies were first treated with one pathogen and then with a second one 48 h after the first application. That is, flies were either treated with B. bassiana first and the bacteria two days later or with the bacteria first and the B. bassiana two days later. The goal was to allow the first pathogen to establish before the second pathogen was applied. B. bassiana conidia usually germinate in 14–24 h. For each trial, 0.5 % CapSil that contained either no microorganisms (control), B. bassiana alone (10⁶ conidia/μl), or the bacteria alone (10⁶ cfu/μl) also were prepared. Flies were sedated and treated with 1 μl of pathogen suspension per fly. Fly mortality was monitored for 7 days after the first application. Trials were replicated three times using 20 flies for each bacteria and B. bassiana combination (60 flies per treatment, n = 3).

2.6 Modified disc diffusion assay to test for antagonistic effects of the three bacteria species on B. bassiana growth

A modified disc diffusion assay (Bauer et al. 1966) was used to determine the compatibility of each bacterial strain with B. bassiana. Suspensions of B. bassiana were prepared by diluting conidia into fresh SDY broth at 1 × 10⁶ conidia/ml. An aliquot of 100 μl of this dilution was spread evenly onto an SDAY plate to create a fungal lawn. The plate was allowed to dry for 30 min in a biosafety cabinet, then one blank 6 mm filter disc (Becton, Dickinson and Company, Washington, DC) was placed in the center of each SDAY plate. Bacteria in LB media (10 µl from the culture tube containing approximately 10⁸ cfu) were pipetted onto the center of the blank disc. Three separate plates for each bacterial strain were evaluated for the three bacteria-B. bassiana combinations. Amphotericin B (10 µl; Fungizone^®, 250 μg/ml; Sigma Alrich, St. Louis, Missouri, USA) was used as a positive control, and a blank disc was the negative control. The zone of inhibition was measured after 72 h by first determining the diameter of the zone around the disc that had no fungal growth. The zone of inhibition is expressed as the shortest linear distance (diameter) from the edge of the central disc to the fungal growth. If there was no inhibition of fungal growth, then the measurement was recorded as 0 mm.

2.7 Statistical analysis

B. bassiana usually kills house flies in 5 days so we chose to examine the mortality data from the pathogen combination bioassays on days 3 (before B. bassiana mortality), 5 (during B. bassiana mortality), and 7 (after B. bassiana mortality) using a linear mixed model fitted with the effects of treatment, time and the interaction using repeated measures ANOVA through the Mixed Procedure as implemented in SAS (Proc Mixed), version 9.4 (SAS Institute, Cary, North Carolina). Residual terms were modelled by considering an autoregressive order 1 error structure and the degrees of freedom were adjusted using the Kenward-Roger method. Adjusted treatment means at each time point (3, 5 or 7 days) were compared using Tukey’s honest significant difference tests at α = 0.05. Data from the central disc test experiment appeared normal on a plot of quantiles (Q-Q plot) and a one-way analysis of variance (ANOVA) was carried out in R 0.99.491 (RStudio, Inc. Boston, Massachusetts); differences between means were evaluated by Tukey’s honest significant difference multiple comparison tests.

3.1 Simultaneous applications of B. bassiana and bacteria

Combinations of B. bassiana and P. temperata did not result in substantial mortality, either alone or in combination, until day 5, and there was no indication that the two-pathogen combinations caused higher mortality than B. bassiana alone throughout the test (Figure 1A). Significant differences were seen with treatment over time (F _{21, 64} = 6.84, P < 0.001). There was no significant treatment effect on day 3 (P > 0.05). On day 5, mortality in the B. bassiana treatments with and without bacteria (40–53 %) did not differ significantly from each other but both treatments caused significantly higher mortality than the control (P < 0.05). On day 7 mortality in the B. bassiana treatments with and without bacteria (65–85 %) did not differ significantly from each other but both treatments caused significantly higher mortality than the control or P. temperata alone (P < 0.001).

Figure 1:

Mortality (%, mean ±SE) of adult female house flies 3, 5, and 7 days after 1 μl topical application of 0.5 % CapSil containing Photorhabdus temperata (A, P.t), Serratia marcescens (B, S.m), and Pseudomonas protegens (C, P.p) (all 10⁶ cfu) and Beauveria bassiana (10⁶ conidia, B.b) alone or in combination at the same time. Bars with the same letter within a time point (3, 5, or 7 days) do not differ at p = 0.05 (Tukeys HSD). Control = CapSil 0.5 % alone.

Results with S. marcescens combinations (Figure 1B) were similar to those with P. temperata, with low mortality until day 5 and no evidence of higher mortality in the two-pathogen combination compared to the treatment with B. bassiana alone. A significant interaction was observed between treatment and time (F _{21, 64} = 4.31, P < 0.001). There was no significant treatment effect on day 3 (P > 0.05). On day 5 mortality in the B. bassiana treatments with and without bacteria (38–54 %) did not differ significantly from each other and the only treatment that was significantly different from the control was B. bassiana alone (P < 0.05). Mortality in the B. bassiana treatments with and without bacteria (65–85 %) did not differ significantly from each other on day 7 but both treatments caused significantly higher mortality than the control (P < 0.001).

In contrast, P. protegens combined with B. bassiana caused 50 % mortality after 3 days and continued in an upward trend until day 7 (Figure 1C). There was a significant difference with treatment over time (F _{21, 64} = 10.52, P<0.0001). On day 3, the two treatments with P. protegens caused significantly higher mortality (53–60 %) than the control or B. bassiana alone (P < 0.01). Mortality in the three pathogen treatments was similar but significantly higher than the control on days 5 (P < 0.001) and 7 (P < 0.001).

3.2 Sequential applications of pathogens

Mortality due to P. temperata or B. bassiana applied alone were similar to the previous tests, with maximum mortality on day 7 of 21 and 85 %, respectively (Figure 2A). Treating flies with B. bassiana followed two days later with P. temperata resulted in no significant treatment effect on day 3. On day 5, mortality in the treatments that included initial B. bassiana applications with and without bacteria were similar to each other and significantly higher than the other three treatments (P < 0.01). On day 7, mortality in the three treatments that included B. bassiana were similar to each other and significantly higher than the controls and P. temperata alone (P < 0.01).

Figure 2:

Sequential applications of bacteria and Beauveria bassiana at three time points: 3, 5, and 7 days. Topical application of 1 μl of 0.5 % CapSil containing Photorhabdus temperata (A, P.t), Serratia marcescens (B, S.m), and Pseudomonas protegens (C, P.p) (all 10⁶ cfu) and Beauveria bassiana (B.b; 10⁶ conidia), alone or in sequence with 48 h between treatments. Bars represent mean percentage mortality ± standard errors, and different letters within a time point denote significant differences (Tukeys HSD P ≤ 0.05) among treatments. Sequential treatments are indicated with an arrow → to indicate the pathogen applied first and second. Control = CapSil 0.5 % alone.

Similar results were observed in sequential treatments of S. marcescens and B. bassiana, with no evidence of accelerated mortality in species combinations compared with B. bassiana alone (Figure 2B). There was a significant difference with treatment over time (F _{28, 80} = 2.7, P < 0.0001). No significant treatment effects were observed on days 3 (P > 0.05) or 5 (P > 0.05). On day 7 mortality in the three treatments that included B. bassiana were similar to each other and significantly higher than the controls and S. marcescens alone (P < 0.01).

With P. protegens, there were no significant treatment effects on day 3 (Figure 2C) (P > 0.05). On day 5, mortality was highest when P. protegens was followed by B. bassiana, but this was only significantly different from the control (P < 0.001). Mortality on day 7 was highest in the three treatments that included B. bassiana (P < 0.05), which were all higher than in the controls.

3.3 Modified disc diffusion assay to test for antagonistic effects of the three bacteria species on B. bassiana growth

S. marcescens and P. protegens were slightly motile in this assay even on 1.5 % agar, and the bacteria grew beyond the diameter of the disc creating a halo effect (Figure 3). Comparing each bacterial strain to the blank control (zone of inhibition = 0 mm), P. temperata and P. protegens both inhibited the growth of B. bassiana (zone of inhibition = 3.25 ± 0.5 mm; P < 0.001), whereas S. marcescens did not inhibit growth (zone of inhibition = 0 mm; P > 0.05). Similarly, when bacteria strains were compared with the positive control, amphotericin B (zone of inhibition = 3.6 ± 0.6 mm), P. temperata (P = 0.808) and P. protegens (P = 0.808) showed no significant difference in zones of inhibition.

Figure 3:

Sabouraud dextrose agar with yeast extract plate with Pseudomonas protegens on the disc (black line) showing zone of inhibition (red line) against Beauveria bassiana lawn, as well as a bacterial halo that grew beyond the disc.