The infectivity of the entomopathogenic fungus Beauveria bassiana to insecticide-resistant and susceptible Anopheles arabiensis mosquitoes at two different temperatures

Mosquitoes were collected inside traditional houses in Karonga (9° 48'51.04" S, 33° 52'.97" E), northern Malawi, using a mouth aspirator. They were placed into small polystyrene cups covered with gauze netting and later identified morphologically using the taxonomic keys of Gillies and Coetzee [36]. Only female mosquitoes identified as members of the Anopheles gambiae complex were retained. A cohort of these mosquitoes was maintained on a 10% sucrose solution soaked in cotton pads for a few hours after which they were subjected to insecticide susceptibility tests. Another cohort of female mosquitoes was blood-fed for oviposition. These mosquitoes were also maintained on a sucrose solution before and during transportation to the laboratory for colony rearing and fungal susceptibility tests.

The Polymerase Chain Reaction (PCR) method [37] was used to test the species integrity of selected laboratory-reared An. arabiensis colonies as well as to identify wild-caught material morphologically identified as An. gambiae complex. Selected laboratory colonies included SENN and SENN-DDT originating from Sudan, and MBN and MBN-DDT originating from Kwazulu/Natal, South Africa. SENN-DDT and MBN-DDT were selected for resistance to DDT from their respective parent colonies.

Newly emerged adult mosquitoes from each An. arabiensis colony were exposed to discriminating dosages of representative insecticides from all insecticide classes recommended by WHO for malaria vector control (Table 1). The standard insecticide susceptibility assays and test kits were used [38]. Samples of 25 non blood-fed mosquitoes per cylinder were exposed to insecticide-treated papers for one hour. After exposure, mosquitoes were transferred to clean holding tubes and provided with cotton pads soaked in a 10% sucrose solution. Knock-down was recorded after 1 h exposure and final mortality was recorded 24 h post-exposure. Each test was duplicated for fully insecticide susceptible colonies and triplicated for insecticide resistant colonies. Tests to determine the insecticide susceptibility status of field-collected mosquitoes were limited to DDT (organochlorine) and dieldrin (cyclodiene) because of sample size limitations. Insecticide susceptibility status was determined according to WHO criteria, whereby colonies were considered resistant when more than 20% of individuals survived the diagnostic dose 24 h post-exposure. A final mortality of 98-100% indicates full susceptibility whilst mortality between 80-97% suggests the possibility of resistance that needs further confirmation [38].

Fungal infection experiments were performed under two different temperature regimes: 21 ± 1°C and 25 ± 2°C, both under 70% ± 15% relative humidity. Infection was carried out using 'fungal suspensors' prepared according to the method described by Scholte et al[15]; this approach is not meant to mimic potential operational delivery of spores but provides a reliable infection method for controlled laboratory-based assays. For each trial three suspensors measuring 6.5 cm in height and 2.5 cm in diameter were lined on the inside with a cylinder of filter paper which also served to hold a plastic vial half filled with a 10% sucrose solution for moistening the filter paper. The sugar solution served as a mosquito attractant to each suspensor. Approximately 100 mg of B. bassiana (isolate IMI 391510) conidial powder was weighed and dusted onto treatment suspensors using a small paintbrush. Treated suspensors and an equal number of untreated, control suspensors were placed individually into small cages (2.5 litres). Cohorts of 25-35 sugar-fed 2-3 day old adult female mosquitoes from each colony were exposed to either treated or untreated suspensors for 24 h. A minimum of three trials was conducted per colony, each consisting of three treatment replicates and three controls. After the 24 h exposure mosquitoes were transferred into clean holding cages and fed on a 10% sucrose solution during the monitoring period.

Mortality in treatment and control samples was recorded daily up until the death of all fungus-infected mosquitoes. All dead mosquitoes were removed from their respective cages using dissecting forceps. Each cadaver was dipped in 70% ethanol in order to eliminate saprophytic fungi from their cuticles. They were then placed in Petri dishes lined with moistened filter paper and sealed with parafilm to maintain high humidity for fungal sporulation. Petri dishes were incubated at 25 ± 2°C for a period of 3-5 days after which cadavers were screened for external fungal growth under a compound microscope. Cadavers with visible fungal growth on their body surface were considered to have died as a result of fungal infection. Final fungal infection proportion was recorded per sample.

Mosquito survival was recorded daily. Cox' regression analysis [39] using SPSS 15.0 software was used to compare survival between treatments and controls. Comparisons of mortality rates were used to assess differences in the effect of fungus infection on survival between insecticide susceptible and resistant colonies, laboratory-reared and wild caught An. arabiensis as well as between the two temperature regimes. For each factor, i.e. insecticide susceptibility status, colony and temperature, two-way interactions (of the factor with fungus treatment) were included in the Cox Regression model to test if the effect of fungal infection on mosquito survival was significantly influenced by the factor.

A sample of 386 An. gambiae complex mosquitoes was collected from the field in Malawi. They were all identified as An. arabiensis by PCR. All samples drawn from the laboratory-reared colonies were confirmed to be as An. arabiensis.

The insecticide susceptibility status of the laboratory An. arabiensis colonies and the wild-caught sample is shown in Figure 1. There was no evidence of insecticide resistance in MBN and field-collected mosquitoes, although the latter were only exposed to 4% DDT and 4% dieldrin because of limited numbers available for exposure assays. There was clear evidence of resistance to 0.75% permethrin in the baseline colony SENN. The SENN-DDT and MBN-DDT selected colonies showed measurable resistance to DDT as expected, as well as resistance to one or more insecticides from each of the other classes (pyrethroids, carbamates and organophosphates).

Daily survival of the field-collected mosquitoes after fungus exposure is shown in Figure 2. Exposure to dry B. bassiana spores resulted in significant reductions in longevity of the wild An. arabiensis mosquitoes (P < 0.05, HR > 1) relative to their respective controls. Comparisons of survival following fungal exposure between wild-caught and baseline laboratory-reared An. arabiensis (MBN and SENN, Figure 3) did not reveal any significant difference in susceptibility to fungal infection (P < 0.05, HR > 1).

Survival curves of the insecticide resistant and susceptible laboratory colonies are shown in Figure 3. Exposure to B. bassiana spores resulted in significant reductions in longevity in all mosquito colonies (P < 0.05, HR > 1) relative to their respective controls, regardless of their insecticide susceptibility levels and temperature regimes. Cox-regression analysis revealed no significant differences in fungus-infected mortality rates between the insecticide-resistant and baseline colonies. Interaction analyses indicated no significant influence of insecticide susceptibility status on fungus-induced mortality rates at both tested temperatures (HR = 0.769; P = 0.13 and HR = 1.456; P = 0.61 for MBN vs MBN-DDT at 21°C and 25°C respectively and HR = 1.03; P = 0.86 and HR = 0.576; P = 0.27 for SENN vs SENN-DDT respectively at 21°C and 25°C). Even where fungus mortality rates appeared slower in insecticide resistant lines, as in MBN-DDT (Figure 3), this is not because of increased fungal resistance but because the respective untreated control lines also survived better. Therefore, fungal susceptibility is not affected by resistance to insecticides. Fungus-induced mortality rates were relatively rapid at 25°C, with 100% mortality taking 10-12 days post-fungus exposure in the baseline colonies (MBN and SENN) and field-collected mosquitoes, and 18-21 days in the DDT-selected colonies (MBN-DDT and SENN-DDT). At 21°C, equivalent mortalities took 20-28 days and 27-30 days, respectively (Figure 3). The differences in mortality rate were mirrored in the sporulation data, with > 90% sporulation of cadavers in the 25°C regime compared with just over 70% at 21°C. However, not all infected mosquito cadavers show fungal sporulation after death because some mosquitoes may be killed by fungal toxins or secondary infections, rather than primary fungal growth and invasion of organs that tends to facilitate sporulation of the cadaver. As observed from the hazard ratio values in Table 2, the fungus had a stronger effect at higher temperatures than at low temperatures. Cox-regression interaction analyses revealed a significant difference (P = 0.045) in the effect of the fungus on survival between the two temperature regimes. The hazard ratios (risk of death) were greater at the higher temperatures of 25 ± 2°C than at the lower temperatures of 21 ± 1°C (Table 2), suggesting that the probability of a mosquito dying soon after infection was greater at higher temperatures than at lower temperatures. Thus, quantitative differences were detected between the two exposure temperatures in all colonies tested (Figures 2 and 3).