Beauveria bassiana for the simultaneous control of Aedes albopictus and Culex pipiens mosquito adults shows high conidia persistence and productivity

This study was conducted to determine the optimal entomopathogenic fungus for the simultaneous control of the adults of two mosquito species, Aedes albopictus and Culex pipiens. The pathogenicity and virulence against the two species of mosquitoes were evaluated by using 30 isolates of Beauveria bassiana, an entomopathogenic fungus isolated from Korea that has high thermotolerance and UV-B tolerance. Regarding pathogenicity, 23 isolates were pathogenic to Ae. albopictus and 12 isolates were pathogenic to Cx. pipiens; Ae. albopictus adults were more susceptible to B. bassiana than Cx. pipiens adults. Among the isolates, 6 isolates that were simultaneously pathogenic to the two species of mosquitoes were used to evaluate virulence and conidia productivity. B. bassiana CN6T1W2 and JN5R1W1 had higher virulence than the other isolates, and they were more virulent in Ae. albopictus than inCx. pipiens. The conidia productivity of B. bassiana JN5R1W1 on millet grain medium was higher than that of B. bassiana CN6T1W2. Based on these results, B. bassiana JN5R1W1 was selected as the most efficient isolate for the simultaneous control of the two mosquito species. B. bassiana JN5R1W1 can be used effectively in the development of fungal insecticides to simultaneously control Ae. albopictus and Cx. pipiens adults with similar distribution areas.


Introduction
Mosquitoes are the most important vectors of severe human viral diseases, such as malaria, dengue, chikungunya, Zika, West Nile, and urban yellow fever, in tropical and subtropical regions (Mayer et al. 2017;Weaver et al. 2018). Mosquitoes are considered to be the most important vector species in the world, not only because of their susceptibility to these diseases but also because they target more than one host during their lifecycle (Fernandes et al. 2018;Harrington et al. 2001). Among the mosquitoes, the Asian tiger mosquito, Aedes albopictus (Skuse) (Diptera: Culicidae), transmits many viral diseases, such as yellow fever, dengue and chikungunya. It is also believed to have the potential for Zika transmission among humans due to the host capacity of the Zika virus (Hochedez et al. 2006;Wong et al. 2013). In addition, the house (or northern house) mosquito, Culex pipiens (L.) (Diptera: Culicidae), is a temperate Eurasian nocturnal mosquito that has established populations throughout temperate North America (Farajollahi et al. 2011). Culex pipiens may be strong nuisance and mediate a number of infectious diseases, such as Japanese encephalitis, West Nile, and Rift Valley fever virus (Farajollahi et al. 2011). Moreover, since these mosquito species are present in areas that contain more than half of the world's population, continuous mosquito control efforts are needed to prevent the prevalence of these diseases (Khaiboullina et al. 2018;Kraemer et al. 2015;Lounibos 2002). In addition, due to recent global climate anomalies, the distribution of these mosquito species and related arboviral diseases are spreading (Shuman 2011). Therefore, these mosquitoes are two major mosquito vectors whose importance is increasing globally. However, since there are no vaccines against these viruses, except for Japanese encephalitis, controlling the mosquitoes that spread them is a method to prevent a pandemic due to these diseases.
There are two methods for controlling mosquitoes: chemical and biological control. Chemical control has been widely used as a major strategy to control mosquito density. However, the side effects of synthetic chemical insecticides in the environment have become widely popular concerns (Hemingway and Ranson 2000;Liu 2015;Moyes et al. 2017). Over the past several decades, the use of synthetic chemical insecticides by agricultural and public health programmes have caused many problems, including insecticide resistance, environmental pollution, and toxicity to humans and non-target organisms (Ariee et al. 2001;Damalas and Eleftherohorinos 2011;Mužinić and Želježić 2018). Therefore, integrated control methods have been attempted to overcome these limitations, but new strategies have been demanded because the integrated methods have not been effective thus far (Patterson 2016). The most attractive method is biological control.
One of the most successful biological control methods is mosquito larval control using Bacillus thuringiensis israelensis (Bti) and Lysinibacillus sphaericus (Berry 2011). Larval control is effective; however, adult control is largely dependent on synthetic chemical insecticides and the control effect has not been achieved. Therefore, it is necessary to develop a new biological control method for adult mosquito control; the representative method is to use entomopathogenic fungi. Entomopathogenic fungi are already widely used in agricultural pest control and are considered harmless and safe to humans and other animals (Lacey et al. 2001(Lacey et al. , 2015. In particular, they have a great advantage of infecting insects, including mosquitoes, through contact with spores compared to other biological control agents that need to be ingested, such as bacteria, viruses and protozoa. Fungal insecticides using various entomopathogenic fungi have been widely developed for agricultural pests, but despite much effort in their development for mosquito control, authentic products have only recently been developed and require more research efforts (Buckner et al. 2017;Snetselaar et al. 2014).
Among the entomopathogenic fungi used in biopesticide development, Beauveria bassiana and Metarhizium anisopliae are the most widely used (Lacey et al. 2015). These fungi have a relatively wide host range and are regarded as environmentally friendly fungi; various biopesticides have been developed with these agents to control agricultural pests and vectors such as ticks, termites and house flies (Lacey et al. 2001(Lacey et al. , 2015. In particular, M. anisopliae has been studied in a variety of mosquito species, but B. bassiana has been rarely studied in mosquitoes (Huang et al. 2017;Scholte et al. 2004). Virulence and persistence together with the productivity of conidia is required to determine the potential of a fungus for further development as a successful biopesticide (Fernandes et al. 2008Lee et al. 2015;Lopez-Perez et al. 2015;Muñiz-Paredes et al. 2017;Samish et al. 2014). In the past, the selection of fungi for the development of biopesticides was mainly focused on virulence, and they was difficult to put into practical use due to problems with the persistence and productivity of fungal conidia. Therefore, recently, a simultaneous evaluation method for virulence, spore persistence and productivity using a large number of fungi has emerged. To date, the selection of fungi against mosquitoes has also been conducted with a focus on virulence, making it difficult to develop successful products.
To develop a successful fungal insecticide for the simultaneous control of Ae. albopictus and Cx. pipiens mosquito adults, our study was conducted to select fungi with high virulence and high conidia persistence and productivity. For this purpose, the virulence against mosquito adults and the conidia productivity of B. bassiana isolates with high conidia persistence among 342 fungal isolates from Korea (Shin et al. 2013) was evaluated. The most effective fungal isolates were finally selected by evaluating the conidia productivity of the selected fungal isolates.

Mosquitoes
Aedes albopictus and Cx. pipiens mosquitoes were obtained from a colony at the laboratory of Insect Ecology and Toxicology, Chungbuk National University, Korea, and have been cultured under laboratory conditions. A single generation of mosquito adults was reared by placing eggs in plastic trays containing water to which 1 g of fish food was added. Larval trays were covered with a netting material. Larvae were fed fish food daily. Pupae were manually removed and placed in 30 cm cubic net cages. Adult males were provided with 10% sucrose solution, while females were fed twice per week on the blood of white mice. All of the mosquitoes were maintained at 28 ± 2 °C and 75 ± 5% RH with a 12 h photoperiod.

Fungi
One hundred nine isolates of B. bassiana from 342 entomopathogenic fungal strains isolated in Korea (Shin et al. 2013) were used. Fungal isolates were cultured on potato dextrose agar (PDA) plates for 14 days at 25 °C. Then, 14-day-old fungal conidia were collected by scraping the fungi from the PDA plates and resuspending the material in a 0.01% Tween-80 (Difco, USA) solution. The conidial suspension was vigorously stirred and filtered through cotton to remove the mycelial debris. The conidial concentration was determined using a haemocytometer.

Thermotolerance and UV-B tolerance for conidial germination
To evaluate the thermotolerance for conidial germination, 100 μL of conidial suspensions (5 × 10 6 conidia/ mL) were transferred to sterile PCR tubes and immediately placed in a thermal cycler adjusted to 45 °C. Aliquots (20 μL) were removed after 4 h and plated (without spreading) on Sabouraud dextrose agar (SDA; Difco TM, USA) with 0.5 μg/mL of benomyl (95% active ingredient) (SDA-B) (Fernandes et al. 2008). Plates were incubated at 25 °C for 24 h before microscopic observation of germination. For the evaluation of conidial germination following UV-B irradiation, 20 μL of a conidial suspension (5 × 10 6 conidia/mL) from each isolate was plated (without spreading) on SDA-B. Conidia were immediately exposed to irradiances of 0.2 J/cm 2 in a UV irradiation chamber (Bio-Link-BLX-E254, Vilber Lourmat, France) at 25 °C . After exposure, the plates were incubated at 25 °C. Control conidial suspensions were not exposed to UV-B but were inoculated on SDA-B at 25 °C. The relative percent germination was calculated by comparing germination to untreated isolates. At least 100 conidia were examined for each treatment in every experiment.

Bioassay
Fourteen-day-old fungal conidia were used for the bioassay. A conidial suspension with more than 90% viability was used for the bioassay. The viability of conidia was determined on SDA-B. The bioassays were designed in two steps. First, the fungal isolates were tested for pathogenicity against mosquito adults. Bioassays were conducted in a plastic cylinder (height 18, diameter 9 cm) capped with netting material. To evaluate the pathogenicity of the fungus, 10 mosquito adults aged 1-2 days were placed in the cylinder, and 1 mL of a conidial suspension (1 × 10 8 conidia/mL) was sprayed into each cylinder using an SD tower sprayer (Shin et al. 2011). Cotton wool freshly soaked with a 10% sucrose solution was placed on the netting. The control group was treated with the same method but without conidia. To determine whether the dead mosquitoes had been infected by the fungus, after 10 days of treatment, all the mosquito cadavers were removed from the cylinder and maintained at over 90% RH in an incubator at 25 ± 1 °C for 7 days. Cadavers with visible fungal growth on their body surface were considered to have died as a result of fungal infection. Additional bioassays were performed to determine the virulence of the fungal isolates. Twenty mosquitoes were used in the bioassay with the same methods described above. Mosquitoes were checked daily for mortality, and the bioassay was repeated three times.

Conidia productivity
Conidia were produced in Petri dishes on Italian millet grains (Setaria italica Beauvois). Millet grains for the production of conidia were prepared according to a previous report (Kim et al. 2011). Briefly, millet grains were placed in a polyvinyl bag and soaked in a half volume of water containing 0.15% citric acid. The bag was held at 95 °C for 15 min and then autoclaved at 121 °C for 30 min. After cooling to ambient temperature, the millet grains (5 g) were divided into the Petri dishes and inoculated with a suspension of 1 × 10 7 conidia (1 mL). The conidia on the millet grains were collected daily and counted using a haemocytometer.

Statistical analysis
The mortality data were analysed by SPSS statistical software ver. 25.0 (SPSS, Inc., Chicago, IL, USA). The lethal time for the treatments was analysed with probit analysis. Data were expressed as the mean ± standard error (SE), and statistical significance was set at the conventional α < 0.05 level.

Evaluation of thermotolerance and UV-B tolerance
Most of the 109 B. bassiana isolates used in this study showed low thermotolerance but high UV-B tolerance (Additional file 1: Table S1). The isolates with more than 50% thermotolerance accounted for 10.1% (11 isolates), and those with 30% or more accounted for only 43.1%. On the other hand, most of the fungal isolates showed more than 50% UV-B tolerance, and 67.9% of the isolates showed more than 70% tolerance. As most of the isolates had high UV-B tolerance, 30 isolates were selected based on thermotolerance and used for the evaluation of pathogenicity against the mosquito adults (Table 1).

Pathogenicity to mosquito adults
The pathogenicity of the fungi to Ae. albopictus and Cx. pipiens mosquito adults was examined using 30 isolates of B. bassiana that had a high thermotolerance and UV-B tolerance. The conidia of the selected isolates were applied to the two species of mosquitoes, and only the isolates that produced confirmed mycosis of the cadavers were considered pathogenic B. bassiana isolates. As a result, typical fungal symptoms were observed over time in dead mosquitoes after conidia treatment (Additional file 2: Figure S1). Twenty-three isolates were found to be pathogenic against Ae. albopictus and 12 isolates were pathogenic against Cx. pipiens adults (Table 2). Among the pathogenic isolates, 6 were determined to be simultaneously pathogenic to Ae. albopictus and Cx. pipiens adults.

Virulence in mosquito adults
The virulence of 6 isolates of B. bassiana that were simultaneously pathogenic to Ae. albopictus and Cx. pipiens adults was evaluated against each mosquito species. The pathogenicity of the 6 isolates was confirmed in both mosquito species; 4 isolates pathogenic to Ae. albopictus and 5 isolates pathogenic to Cx. pipiens showed 100% mortality at 10 days after treatment (Table 3). Moreover, 3 isolates of B. bassiana (CN6T1W2, JN5R1W1 and JN5T2W1) showed 100% mortality against both species of mosquito adults. Regarding lethality, all isolates were found to be slightly more pathogenic to Ae. albopictus than to Cx. pipiens adults. The LT 50 values of B. bassiana CN6T1W2 and

Discussion
and Želježić 2018). Therefore, the research and development of various control methods with relatively low environmental impacts, such as entomopathogenic fungi, has attracted much attention. However, studies on fungi pathogenic to Ae. albopictus and Cx. pipiens are scarcer than studies on other mosquito species. Much of the fungi research has been conducted for the control of Ae. aegypti or Anopheles spp., and there have been many reports on various fungi, including B. bassiana and M. anisopliae, for these mosquito species (Huang et al. 2017;Scholte et al. 2004). There have been some reports of M. anisopliae for the larval and adult control of Ae. albopictus and B. bassiana for adult control using a contamination device, with remarkable results (Buckner et al. 2017;Snetselaar et al. 2014). In the case of Cx. pipiens, the research is relatively limited and only fungal metabolites were considered for larval control (Weiser et al. 1992). Eco-friendly insecticides are mainly used for the control of mosquito larvae, and effective products such as Bti and L. sphaericus have been developed and widely used. However, since the recently developed contamination device has been successful in the control of adult mosquitoes (Buckner et al. 2017;Snetselaar et al. 2014), the selection of a fungus to control mosquito adults is considered to be important. Although mosquito control using fungi has been considered, the fact that only one contamination device has been produced may be due to the ineffective methods of spraying fungal conidia for mosquito adults and treating water containing mosquito larvae. Therefore, it is considered that a treatment method similar to that of the contamination device is the most efficient at present; thus, the proper selection of a fungus to be utilized with this treatment method is becoming important. The fungal isolate selected in this study is expected to be effective for the simultaneous control of Ae. albopictus and Cx. pipiens adults.
In addition, the results of our study showed that 12 isolates and 23 isolates of B. bassiana were pathogenic to Cx. pipiens and Ae. albopictus, respectively, and Ae. albopictus was more susceptible to B. bassiana. However, until now, the pathogenicity of some M. anisopliae (Scholte et al. 2007) and B. bassiana strains (Buckner et al. 2017) has been tested on Ae. albopictus adults, but the pathogenicity of M. anisopliae and B. bassiana has not been reported for Cx. pipiens adults. Therefore, further research on the susceptibility of these mosquito species to B. bassiana using more isolates is needed. In particular, only 6 isolates of B. bassiana were found to be simultaneously pathogenic to these two species of mosquitoes; reports of a fungus that is simultaneously pathogenic to two or more species of mosquitoes are rare. Similar studies have reported simultaneous pathogenicity in Ae. aegypti and Ae. Albopictus, which are of the same genus (Buckner et al. 2017;Scholte et al. 2007). However, there have been no reports on the selection of fungi that are simultaneously pathogenic to two species of mosquitoes from different genera.
In our study, we attempted to increase the industrial usefulness of selected fungi by using only fungal isolates that are expected to be highly persistent in the natural environment. Traditionally, the selection of entomopathogenic fungi for the development of insecticides has focused mainly on their virulence to target insects. However, in terms of industrial usefulness, the persistence of fungicidal agents when applied in outdoor environments or during the distribution process is just as important as virulence Lee et al. 2015). Therefore, in recent years, there have been an increasing number of attempts to evaluate various factors simultaneously and to select suitable fungi for control. Many environmental factors affect the persistence of fungi, but the most significant effects are thought to be thermotolerance and UV-B tolerance (Fernandes et al. 2008Lee et al. 2015).
Thermotolerance is closely related to the production of reactive oxygen species (ROS) (Zhang et al. 2017) and heat shock proteins (HSPs) (Xie et al. 2010). Thermotolerance could be enhanced by the overexpression of HSP25 and pyruvate kinase or superoxide dismutase (Wu et al. 2019;Xie et al. 2010). In addition, it is possible to increase the thermotolerance of fungus by controlling the nutrients or the osmotic pressure of the medium . Such medium conditioning could also enhance UV tolerance. UV tolerance is related to the conidial colour of fungus; darkly pigmented conidia are more tolerant to UV radiation than less-pigmented conidia (Rangel et al. 2006). However, it was also reported that conidial pigmentation may be important to protect from UV radiation, but it is not crucial . In our results, although B. bassiana with white conidia was used, most of the isolates showed high UV-B tolerance with more than 50%. UV-tolerance can be assessed with various UV doses, but we set the standard of 0.2 J/cm 2 for various tolerances of fungi according to the previous report . These various manipulations and controls may enhance the tolerance of fungus, but they may have limitations if the selected fungus does not have high thermotolerance. Therefore, it is important to select and use a fungus with high thermotolerance. In the present study, only B. bassiana isolates with high thermotolerance and UV-B tolerance were used, so the practical use of the finally selected isolate is be expected to be high.
To develop appropriate fungal insecticides, virulence to the target insect, conidia persistence and conidia productivity are important factors (Lopez-Perez et al. 2015;Muñiz-Paredes et al. 2017). Recently, several studies have been conducted on various media that can increase the conidial productivity of fungi (Lopez-Perez et al. 2015;Kim et al. 2014;Safavi et al. 2007;Shah et al. 2005). However, when the basic conidia productivity of the fungus is low, the replacement of the medium is a limitation. Therefore, a comparative evaluation of the conidia productivity of fungi is also important for proper fungus selection. The productivity of conidia as an evaluation factor for the selection of the final fungal isolates was added as a selection index in our study. Conidia productivity was evaluated on millet grain medium, which was reported to be effective for the production and tolerance of conidia (Kim et al. 2011). The conidia productivity of B. bassiana in various solid media was reported with approximately 1.0-2.9 × 10 10 conidia/g on dry substrate, but it was approximately 1.8-2.2 × 10 9 conidia/g on wet substrate as our same condition (Lopez-Perez et al. 2015). Additionally, the conidia productivity in the Italian millet medium used in our study was reported to be approximately 1.5-2.5 × 10 9 conidia/g (Song et al. 2019). This indicated that our finally selected B. bassiana JN5R1W1 has the high conidia productivity with 3.8 × 10 9 conidia/g. The difference in conidia productivity between the fungal isolates in our study indicated that this comparison was necessary to select the optimal fungal isolate. Although our study used only millet grain and evaluated only conidia productivity, it may be beneficial to perform further studies on the virulence and tolerance of fungal conidia produced on various media.
In conclusion, B. bassiana JN5R1W1 was finally selected for the simultaneous control Ae. albopictus and Cx. pipiens adults through evaluations of virulence, persistence and productivity of the fungal conidia. The development of successful fungal insecticides containing this fungal isolate for the simultaneous control of two species of mosquitoes is expected in further studies on optimal formulations and field tests.