Variability of Deltamethrin-Resistant Metarhizium anisopliae Aggressive Strains Group to a Population of the Cossid Moth from Eucalyptus nitens

Random amplified polymorphic DNA-polymerase chain reaction (RAPD-PCR) was used to examine the genetic variability among Metarhizium anisopliae isolates tested to the cossid moth, Coryphodema tristis. All the isolates tightly clustered into one or the other of two groups that diverged at 12%. Results suggested that certain genotypes of the fungus, that grouped together, were able to infect moth larvae while others did not. A fragment of 760 bp, which presents high homology with a host-adaptation related protein coding gene, distinguished between aggressive and non-aggressive isolates. Neither mycelial growth nor sporulation rate or presence of known virulence genes was correlated with mortality values. Some isolates, including the most aggressive isolate ARSEF2518, were compatible with deltamethrin. Deltamethrin treatment killed all the larvae after seven days whereas fungal and mixed treatments respectively reached the same mortality after 28 and 21 days.

and has long been associated with exotic plants such as vines, apples and quince in the Western Cape. It also feeds on multiple native trees. Since 2004, it has been found feeding on Eucalyptus nitens in the Carolina-Badplaas-Lothair area in South Africa, resulting in gregarious feeding behavior and extensive damage (Boreham, 2004). Adult females lay eggs on the bark of trees. As the larvae grow, they bore into the cambium, where they cause tunneling which results in severe damage to the trees which often die. The cossid moth has a two-year life cycle in the Western Cape, but the duration of the life cycle in the summer rainfall area is not yet known. Both the main trunks and branches are attacked.
Neither insecticides nor natural enemies are currently in use or registered for the cossid moth. More effective and sustainable control systems should thus be developed, as for example the recent pheromone determination (Bouwer et al., 2015). This study evaluates twelve Metarhizium anisopliae (Metschnikoff) Sorokin (Hypocreales: Clavicipitaceae) isolates aggressiveness and their mycelial growth, sporulation rate, virulence genes presence, and insecticide-compatibility.
The genetic variability of the assayed population was also determined. A selected isolate, based on the indicated parameters results, was tested in simulated field conditions.

Bioassays
Initially, infectivity assays were performed. Infested E. nitens logs were obtained at Lothair area in South Africa in February 2013, after what C. tristis larvae were collected opening these logs with a log hydraulic chopper. Larvae were pre-sorted into separate Petri dishes (60 mm diam.) each containing four larvae, and held at 10˚C in laboratory one week prior to bioassay. All isolates (Table 1) were cultured on PDA at 20˚C for 2 weeks. Conidia were harvested into 0.01% Tween 80 and concentrations of 0, 10 5 , 10 7 , 10 9 conidia/ml were prepared. Conidial viability > 95% was confirmed by spread-plating 0.1 ml of the 10 5 suspension onto PDA, incubation at 20˚C and germination rate determination after 96 h. In a complete randomized block design at 20˚C and 16:8 h light/dark cycle, each specimen was dipped in the correspondent spore or control suspension decanted into on-ice 5-ml sterile tubes, inverted five times, and transferred to its insect batch into a second series of dishes with 10 ml diet (10 g semi-skimmed power milk, 44 g honey, 20 g agar per 1000 ml distilled water), and sealed with Parafilm (4 larvae/container, 5 replicate containers, 20 larvae/concentration, 4 concentrations, 80 larvae/isolate). Assays were observed every day for 30 days.
At each observation, cadavers were removed from the treated arena and placed into cross-referenced dishes with moist filter paper until infection confirmation by sporulation observation. Experiment was repeated twice using insects collected on two occasions and freshly-produced conidia. Daily recorded mortality was corrected for control mortality and mean cumulative mortality was plotted against 30 days.

Compatibility with Deltamethrin
To evaluate M. anisopliae isolates sensitivity to deltamethrin (Decis 1.5% w/v;

Mycelial Growth, Sporulation Rate, and Virulence Genes Detection
To study mycelial growth, six replicate two weeks-old mycelial plugs (0.5 cm diam.) were placed face down at the centre of 60-mm dish containing 10 ml of PDA. After 15 days at 20˚C and 16:8 h L:D, occupied areas were traced and measured. Sporulation rate was estimated by collecting conidia with 4 ml of 0.01% Tween 80 at 200 rpm for 30 min at 22˚C. As regards some virulence genes detection, PCR amplification of Pr1a protease gene was performed by nested PCR with outer primers METPR1 (5'-CACTCTTCTCCCAGCCGTTC-3'), METPR4 (5'-GTAGCTCAACTTCTGCACTC-3'), and inner primers METPR2 (5'-AGGTAGGCAGCCAGACCGGC-3') and METPR5 (5'-TGCCACTATTGGCCGGCGCG-3'). The template DNA was amplified in a 50 μL PCR reaction volume, consisting of 5 μL of 10X Reaction Buffer, 5 μL of MgCl 2 (25 mM), 5 μL of dNTPs (10 mM), 1 μL of each primer (10 μM), 1 μL of DNA solution and 0.5 μL of Super-Therm Taq polymerase. PCR reactions were performed on a GeneAmp PCR System 9700 (Applied Biosystems) with an initial denaturation step of 4 min at 95˚C. This step was followed by 40 cycles of denaturation at 95˚C (60 s), annealing at 58˚C (60 s) and elongation at 72˚C (2 min). A final extension was conducted for 8 min at 72˚C. The second PCR used 1 μl of the first PCR as template. Products were visualized under UV on a 1% agarose gel stained with Gelred (Biotium), run in a Wide Mini-Sub Electrophoresis System at 90V, 400 mA for 30 min and digitalized in a ChemiDoc Gel Documentation System (BioRad). Pr1b protease gene was amplified using the same PCR and conditions but with primers Pr1B1, Pr1B2, and inner primers Pr1B3 and Pr1B4 (Wang et al., 2002).

Screening of RAPD Markers
To determine the genetic variability of the tested isolates, a total of twenty-one 10-mer random primers (Inqaba Biotech) were screened (Table 2). PCRs were carried out in volumes of 50 μl containing 5 μL of 10X Reaction Buffer, 3 μL of MgCl 2 (25 mM), 4 μL of dNTPs (10 mM), 2.5 μL of primer (10 μM), 1.5 μL of DNA solution and 0.5 μL of Super-Therm Taq polymerase. Conditions were 94˚C for 4 min, followed by 35 cycles at 94˚C (60 s), 47˚C (60 s), 72˚C (2 min), and then 72˚C for 7 min. Electrophoresis was as before but through 2% agarose gels for 90 min. A genetic similarity matrix was calculated using the method of simple matching coefficient. The matrix was used to perform hierarchical cluster analysis based on the un-weighted pair-group method using arithmetic averages using GelQuest and ClusterVis (Sequentix). One diagnostic marker was recovered using the QIAquick gel extraction kit (Qiagen), sub-cloned into the pGEM-T easy cloning vector (Promega), and both stranded sequenced.

Results
Obtained results indicate considerable variation in aggressiveness among isolates, being isolates number 6, 3, 11 and 5 pathogenic whereas the rest of the isolates were not pathogenic. Isolate number 6 (ARSEF2518) presented the highest cumulative mortality value (Figure 1(a)). Mean lethal dose values could not be calculated through probit analysis because pathogenic isolates were only effective at the highest assayed conidia concentration.
Among the twelve isolates, isolates number 2, 3, 6, 8, 9, 10 and 11 growth was not significantly reduced by the tested deltamethrin concentrations ( Figure   1(b)). Deltamethrin treatment killed all the larvae after 7 days. Treatment with ARSEF2518 fungus and mixed treatments respectively reached the same mortality percentage after 28 and 21 days (Figure 1(c)).
Neither mycelial growth (Pearson correlation coefficient (P) = 0.516, F = 0.086) nor sporulation rate (P = 0.360, F = 0.250) data were correlated with cumulative mortality values. Tested virulence genes were present among all the isolates (Table 1).  respectively sprayed with four treatments. See experimental details. Means with the same letter are not significantly different (P > 0.05), by ANOVA followed by Tukey test. Isolates details given in Table 1.
Aggressive isolates grouped together in the topology of the 21 decamer primers (Table 2)-based consensus tree (Figure 2). Cluster analysis of the banding patterns revealed two tightly clustered groups among the M. anisopliae isolates.
Group 1 consisted of six isolates with a mean similarity of 88.6%, and group 2 consisted of six isolates with a mean similarity of 89%. Group 1 and group 2   Table  1.
isolates diverged at the 12% similarity level. One primer with reproducible amplified band of 760 bp unique to the aggressive isolates group was selected.

Discussion
This is the first study investigating the aggressiveness of M. anisopliae to C. tristis. There was considerable variation in aggressiveness among isolates, presenting isolate number 6 (ARSEF2518) the highest cumulative mortality value. It is interesting to note that some Coleoptera or Homoptera-derived isolates showed more aggressiveness than some isolates from Lepidoptera, supporting that the original host is not a reliable indicator of higher virulence in some entomopathogenic fungi (Varela & Morales, 1996;Devi et al., 2008;Romón et al., 2017).
Entomopathogenic fungi and deltamethrin have been used in integrated management research to control ticks (Bahiense et al., 2008) and Triatoma infestans (Forlani et al., 2013), with similar mortality rates among individual and mixed treatments and with no negative effect on certain isolates of the fungus after its  (Bridge et al., 1997), geographical range (Leal et al., 1994), morphology (Bidochka et al., 1994) and/or pathogenicity (Fegan et al., 1993) have been demonstrated previously. In the present study, the nucleotide sequence of the cloned diagnostic RAPD-marker fragment from OPA-08 has been deposited in the GenBank database under accession numbers KJ543536-KJ543539, and it presents 96% homology with a protein mRNA from Metarhizium robertsii related with host adaptation (Hu et al., 2014). Biocontrol agent development requires isolation in pure culture of the microorganism, its identification, efficacy laboratory trials, pilot trials under field conditions, implementation of formulation and biosafety studies. The phase of screening for proper isolates is very time-consuming and expensive and there is a need for faster and easier methods. The development of fingerprinting of higher aggressiveness could be valuable. Our results suggest that the isolate ARSEF2518 appears to have the best potential among the tested isolates for future evaluation and development as fungal biocontrol agent to C. tristis. However, further laboratory and field experiments are needed. More studies should be directed to determine effects of different formulations and storage on conidial viability, and to test its ability and the other aggressive isolates in controlling C. tristis populations in the field.