Susceptibility and Immune Defence Mechanisms of Rhynchophorus ferrugineus (Olivier) (Coleoptera: Curculionidae) against Entomopathogenic Fungal Infections

Insects infected with entomopathogenic fungi, experience physiological changes that influence their growth and immune defence. The potential of nine isolates of entomopathogenic fungi was evaluated after determining percent germination and relative conidial hydrophobicity. However, nutritional indices were evaluated after immersing eighth-instar Rhynchophorus ferrugineus larvae into each isolate suspension (1 × 107 conidia/mL). The results showed that isolates B6884 and M9374 had 44.51% and 39.02% higher conidial hydrophobicity compared with isolate I03011 (least virulent). The results of nutritional index assays revealed a significant reduction in growth indices after infection with different isolates. Compared with control, B6884 and M9374 greatly decreased larval growth by reducing the efficacy of conversion of ingested food (36%–47%) and Efficacy of conversion of digested food (50%–63%). Furthermore, only isolate B6884 induced 100% mortality within 12 days. Compared with control, isolate I03011, possessing the lowest conidial hydrophobicity, only reduced 0.29% of the efficacy of conversion of ingested food (ECI) and 0.48% of the efficacy of conversion of digested food (ECD). Similarly, transcriptomic analysis of genes related to the Red palm weevil (RPW) immune response, including pathogen recognition receptors (C-type lectin and endo-beta-1,4-glucanse), signal modulator (Serine protease-like protein), signal transductors (Calmodulin-like protein and EF-hand domain containing protein) and effectors (C-type lysozyme, Cathepsin L., Defensin-like protein, Serine carboxypeptidase, and Thaumatin-like protein), was significantly increased in larval samples infected with B6884 and M9374. These results suggest that for an isolate to be virulent, conidial hydrophobicity and germination should also be considered during pathogen selection, as these factors could significantly impact host growth and immune defence mechanisms.


Introduction
Red palm weevil (RPW) is an exotic highly destructive pest of palms, particularly Phoenix dactylifera, in different geographical areas worldwide. The legless, creamy white larvae (grubs) of Rhynchophorus ferrugineus (Olivier) (Coleoptera: Curculionidae) are the most destructive stage of the weevil. These insects feed on tender soft palm tissues and move toward the centre of the infested host palm. Under severe attack, such a feeding pattern causes the crown to collapse [1,2].

Quantification of Host Immune Defence-Related Genes Using Quantitative Reverse Transcription Polymerase Chain Reaction (qRT-PCR)
Highly-virulent isolates of entomopathogenic fungi, such as B6884 and M9374, established in the current study greatly induced the expression of immune related genes of red palm weevil larvae. The interaction of the quantitative expression of pathogen recognition receptors (PRRs) from the fat body of red palm weevil larvae with tested fungal isolates was significantly different (F = 357.14; df = 16, 108; p < 0.0001). Overall, C-type lectin (CTL) was significantly expressed compared with other studied PRRs ( Figure 2). However, Beta-glucosidase was not greatly expressed, showing <1% relative fold-expression compared with control. Overall, isolates B6884 and M9374 significantly induced the expression of C-type lectin and endo-beta-1,4-glucanse of red palm weevil larvae ( Figure 2). Isolates I7284, I03011, and M9H755 had the lowest effect on the expression of C-type lectin and endo-beta-1,4-glucanse genes. varied (F = 76.8; df = 8, 36; p < 0.0001). The highest LT50 value, calculated from the infection of isolate I03011 (11.51 days), remained significantly at same level with isolates I7284 (11.21 days) and B3H203 (11.37 days).

Quantification of Host Immune Defence-Related Genes Using Quantitative Reverse Transcription Polymerase Chain Reaction (qRT-PCR)
Highly-virulent isolates of entomopathogenic fungi, such as B6884 and M9374, established in the current study greatly induced the expression of immune related genes of red palm weevil larvae. The interaction of the quantitative expression of pathogen recognition receptors (PRRs) from the fat body of red palm weevil larvae with tested fungal isolates was significantly different (F = 357.14; df = 16, 108; p < 0.0001). Overall, C-type lectin (CTL) was significantly expressed compared with other studied PRRs ( Figure 2). However, Beta-glucosidase was not greatly expressed, showing <1% relative fold-expression compared with control. Overall, isolates B6884 and M9374 significantly induced the expression of C-type lectin and endo-beta-1,4-glucanse of red palm weevil larvae ( Figure 2). Isolates I7284, I03011, and M9H755 had the lowest effect on the expression of C-type lectin and endo-beta-1,4-glucanse genes.  Quantitative expression pattern of signal modulators, such as serine protease-like protein and trypsin-like serine protease, among the tested larvae in response to infections, showed great variations ( Figure 3) The gene expression of the studied signal modulators (F = 1071.42; df = 1, 72; p < 0.0001), upon exposure to different fungal isolates (F = 129.87; df = 8, 72; p < 0.0001) and their interactions (F = 45.73; df = 8, 72; p < 0.0001), showed significant differences. Comparatively, serine protease-like protein was significantly expressed compared with trypsin-like serine protease ( Figure 3). Overall, the most virulent isolate, B6884, induced the highest expression of serine protease-like protein. The lowest fold-expression of serine protease-like protein and trypsin-like serine protease was induced by the least virulent isolates, including I7284, I03011, and M9H755 ( Figure 3). Quantitative expression pattern of signal modulators, such as serine protease-like protein and trypsin-like serine protease, among the tested larvae in response to infections, showed great variations ( Figure 3) The gene expression of the studied signal modulators (F = 1071.42; df = 1, 72; p < 0.0001), upon exposure to different fungal isolates (F = 129.87; df = 8, 72; p < 0.0001) and their interactions (F = 45.73; df = 8, 72; p < 0.0001), showed significant differences. Comparatively, serine protease-like protein was significantly expressed compared with trypsin-like serine protease ( Figure 3). Overall, the most virulent isolate, B6884, induced the highest expression of serine protease-like protein. The lowest fold-expression of serine protease-like protein and trypsin-like serine protease was induced by the least virulent isolates, including I7284, I03011, and M9H755 ( Figure 3).   Quantitative expression pattern of signal modulators, such as serine protease-like protein and trypsin-like serine protease, among the tested larvae in response to infections, showed great variations ( Figure 3) The gene expression of the studied signal modulators (F = 1071.42; df = 1, 72; p < 0.0001), upon exposure to different fungal isolates (F = 129.87; df = 8, 72; p < 0.0001) and their interactions (F = 45.73; df = 8, 72; p < 0.0001), showed significant differences. Comparatively, serine protease-like protein was significantly expressed compared with trypsin-like serine protease ( Figure 3). Overall, the most virulent isolate, B6884, induced the highest expression of serine protease-like protein. The lowest fold-expression of serine protease-like protein and trypsin-like serine protease was induced by the least virulent isolates, including I7284, I03011, and M9H755 ( Figure 3).   The up-regulation of the studied effectors, such as C-type lysozyme, Cathepsin L., Defensin-like protein, Serine carboxypeptidase, and Thaumatin-like protein (F = 1161.49; df = 4, 180; p < 0.0001), among the larvae infected with different fungal isolates (F = 2246.79; df = 8, 180; p < 0.0001) and their interactions (F = 101.06; df = 32, 180; p < 0.0001), significantly differed at twenty-four hours post exposure. Overall, the infection of isolates B6884 and M9374 significantly enhanced the expression of all studied effectors, particularly Defensin-like protein, C-type lysozyme, and Thaumatin-like protein ( Figure 5). The isolates I7284, M9H755, and I03011 did not induce the expression of the studied effectors. The up-regulation of the studied effectors, such as C-type lysozyme, Cathepsin L., Defensin-like protein, Serine carboxypeptidase, and Thaumatin-like protein (F = 1161.49; df = 4, 180; p < 0.0001), among the larvae infected with different fungal isolates (F = 2246.79; df = 8, 180; p < 0.0001) and their interactions (F = 101.06; df = 32, 180; p < 0.0001), significantly differed at twenty-four hours post exposure. Overall, the infection of isolates B6884 and M9374 significantly enhanced the expression of all studied effectors, particularly Defensin-like protein, C-type lysozyme, and Thaumatin-like protein ( Figure 5). The isolates I7284, M9H755, and I03011 did not induce the expression of the studied effectors.

Impact of Entomopathogenic Fungi on the Growth of R. ferrugineus
The efficacy of conversion of ingested food (ECI) of the larvae upon exposure to all tested isolates of entomopathogenic fungi significantly differed (F = 5.22; df = 9, 39; p < 0.0001). Overall, larvae infected with B6884 exhibited the maximum ECI reduction (46.41%) compared with control larvae. A relatively negligible reduction (0.29%) in R. ferrugineus larvae infected with isolate I03011 was observed (Table 3). There were also significant differences in the efficacy of conversion of digested food (ECD) of larvae upon exposure to the tested entomopathogenic fungal isolates (F = 27.48; df = 9, 39; p < 0.0001). Isolate B6884 showed significantly declined ECD, resulting in a tremendous reduction (62.46%) in ECD compared with the control. However, two isolates of I. fumosorosea, I7284 and I03011, and one B. bassiana isolate B3H203 impart a <10% reduction in ECD compared with the control ( Table 3). The approximate digestibility (AD), however, significantly increased upon exposure to different isolates (F = 10.95; df = 9, 39; p < 0.0001). The most virulent isolates B6884 and M9374 significantly enhanced the AD of the infected larvae, increasing the AD 42.78% and 29.49% compared with control larvae, respectively. However, the least virulent isolate I03011, I7284, and B3H203 only increased the AD 0.20%, 2.22%, and 1.80% compared with control larvae, respectively ( Table 3). The reduction in ECI and ECD index directly proportional to the virulence of the isolates. Overall, virulent isolates, such as B6884 and M9374, impart the maximum reduction in the ECI and ECD indexes, whereas the least virulent isolates failed to inhibit growth indices of red palm weevil larvae. In contrary, inversely proportional relationship was observed between AD and virulence of the isolates compared to the ECI and ECD indexes (Table 3).

Impact of Entomopathogenic Fungi on the Growth of R. ferrugineus
The efficacy of conversion of ingested food (ECI) of the larvae upon exposure to all tested isolates of entomopathogenic fungi significantly differed (F = 5.22; df = 9, 39; p < 0.0001). Overall, larvae infected with B6884 exhibited the maximum ECI reduction (46.41%) compared with control larvae. A relatively negligible reduction (0.29%) in R. ferrugineus larvae infected with isolate I03011 was observed (Table 3). There were also significant differences in the efficacy of conversion of digested food (ECD) of larvae upon exposure to the tested entomopathogenic fungal isolates (F = 27.48; df = 9, 39; p < 0.0001). Isolate B6884 showed significantly declined ECD, resulting in a tremendous reduction (62.46%) in ECD compared with the control. However, two isolates of I. fumosorosea, I7284 and I03011, and one B. bassiana isolate B3H203 impart a <10% reduction in ECD compared with the control ( Table 3). The approximate digestibility (AD), however, significantly increased upon exposure to different isolates (F = 10.95; df = 9, 39; p < 0.0001). The most virulent isolates B6884 and M9374 significantly enhanced the AD of the infected larvae, increasing the AD 42.78% and 29.49% compared with control larvae, respectively. However, the least virulent isolate I03011, I7284, and B3H203 only increased the AD 0.20%, 2.22%, and 1.80% compared with control larvae, respectively ( Table 3). The reduction in ECI and ECD index directly proportional to the virulence of the isolates. Overall, virulent isolates, such as B6884 and M9374, impart the maximum reduction in the ECI and ECD indexes, whereas the least virulent isolates failed to inhibit growth indices of red palm weevil larvae. In contrary, inversely proportional relationship was observed between AD and virulence of the isolates compared to the ECI and ECD indexes (Table 3).

Discussion
The pathogenic and growth inhibition activities of naturally-occurring entomopathogenic fungi against insects of different orders have previously been reported [19]. Previous studies have given little consideration to the virulence-determining traits of entomopathogenic fungi and their impact on the growth and immune defence mechanism of an economically important pest of the date palm, R. ferrugineus. The present study showed that microbe virulence factors, including the germination percentage and relative conidial hydrophobicity, significantly affect the virulence of the studied isolates. Entomopathogenic fungal isolates with higher virulence, elicit higher expression of immune related genes, and slowing the larval growth, ultimately leading to higher mortality among infected larvae.
The current findings clearly indicated that conidial germination and hydrophobicity defined the virulence of the isolates. We found variations in germination and conidial hydrophobicity among the studied isolates. Conidial hydrophobicity is specifically provided through a group of small cysteine-rich proteins (hydrophobins), involved in conidial adhesion, virulence, and dispersion. Recently, an empirical evidence of the involvement of three hydrophobin genes Hyd1, Hyd2 and Hyd3 in virulence, hydrophobicity, and conidiation was provided [20]. In another study, the hydrophobin genes, Hyd1 and Hyd2, were implicated in conidial hydrophobicity and, ultimately, the regulation of the level of adhesion and virulence of the strain [21,22]. Based on these findings, we suggested that pathogen virulence is important to invade the host. Conidia with high levels of hydrophobicity resulted in the high mortality of red palm weevil larvae. During fungal pathogen selection, virulence-determining traits, such as conidial hydrophobicity and conidial germination, reported in the current study should be thoroughly considered.
R. ferrugineus larvae infected with isolates B6884 and M9374 gained less weight compared with uninfected (control) larvae. Isolate B6884 remained the most virulent isolate, followed by isolates M9374 > I9602 > M8762 > B03005 > M9H755 > I7284 > B3H203 > I03011. Nutritional indices, particularly the ECI and ECD values of R. ferrugineus larvae, were significantly different, suggesting that virulent isolates tremendously reduced the growth of red palm weevil larvae. The ECI values representing the overall measurement of ingested food used by the larvae for growth were greatly reduced by the infection of isolate B6884 (46.41%), M9374 (36.59%), I9602 (29.25%), M8762 (20.00%), B03005 (14.88%), M9H755 (8.74%), I7284 (4.22%), B3H203 (2.67%), and I03011 (0.29%). The decrease in ECI values compared with the control revealed that most of the ingested food is used for energy to combat the invading pathogen, and less food is used for larval growth. These findings are consistent with a previous investigation in which M. anisopliae (strain 406) significantly reduced the ECI values against all of the studied larval instars of Ocinara varians Walker [23].
ECD, an index to calculate the precise insect biomass from digested food, was also decreased upon infection with the conidia of virulent isolates. This reduction might reflect the fact that digested food is metabolized for energy production, thereby reducing the ECD compared with the control [5]. In another study, similar growth inhibition pattern of ECI and ECD was reported in O. varians Walker larvae in studies of the virulence of entomopathogenic fungal strains [9].
The fungal infection of virulent isolates tremendously increased the AD values of the red palm weevil larvae. The enhanced AD values might reflect the fact that the fungus-inoculated nutrient-deficient larvae require energy for immune defence. This energy could only be obtained through the use of intrinsic abilities to enhance the approximate digestibilities of the limited foodstuff. Similar results were obtained when the pesticidal activity of labramin was evaluated against the Mediterranean flour moth [24].
It is clear from the mortality and nutritional indices results that isolates B6884 and M9374 are the most virulent. This virulence is mainly regulated by the conidial hydrophobicity that enhances the infection by facilitating conidial adhesion with the host cuticle. In case of compatible interaction (infection), the host immune defence mechanism is greatly up-regulated to combat infection and less energy remained for host growth. The host, under such circumstances (low energy), becomes weaker and more vulnerable to the sporulating conidia that starts their growth in vivo.
These findings also showed that isolates B6884 and M9374 enhanced the response of the pathogen recognition receptors, signal modulators, signal transductors and effectors responsible for host immune defence mechanisms. The enhanced activities of PRRs (particularly C-type lectin and endo-beta-1,4-glucanse); signal modulators (particularly Serine protease-like protein); signal transductors (Calmodulin-like protein and EF-hand domain containing protein); and effectors (C-type lysozyme, Cathepsin L., Defensin-like protein, Serine carboxypeptidase, and Thaumatin-like protein) reduced ECI and ECD values, revealing that the host spends most of its energy combating the invading pathogen attack; hence, little nutrition remains for growth and development. The strains with less conidial hydrophobicity and low germination remained less pathogenic and should not be incorporated into the integrated red palm weevil management programme.

Rearing of Red Palm Weevils
RPW adults were collected from infested date palms. The weevils were reared on pineapples with a 16-h light photoperiod at 30 ± 1 • C and 75% ± 5% RH in cages (57.5 × 29 × 58 cm). After hatching, second instar red palm weevil larvae were individually shifted to 400-mL plastic cups (perforated) and provided artificial diets (Table 5) in an incubator under controlled conditions.

Conidial Germination
The conidial germination of the tested entomopathogenic fungal isolates was determined, as previously described [25]. Briefly, a 50-µL suspension of each studied isolate (24-day-old) at a conidial concentration of 1 × 10 7 conidia/mL in 0.05% Tween 80 was used to inoculate Petri dishes (115 mm × 20 mm) containing PDA. The required concentration (1 × 10 7 conidia/mL) for each fungal suspension was prepared using a Neubauer haemocytometer (Wertheim, Germany). The plates were incubated at 25 ± 0.5 • C, 70% ± 5% RH in complete darkness. Each Petri dish was regarded as one replicate. Five replicates were prepared. Percent conidial germination was determined after 18 h by counting total 100 conidia (germinated or ungerminated) from seven different fields of vision in single Petri dish under a compound microscope (×400). Conidium was considered "germinated" if the germ tube was more than half the diameter. The average of seven fields of vision from each plate was considered as one replicate. The data were analysed using one-way analysis of variance (ANOVA) and Fisher's least significant difference (LSD) test for means comparisons [26].

Conidial Hydrophobicity
An aqueous-solvent partitioning assay was used to determine the conidial surface hydrophobicity [27]. Briefly, the final concentration (1 × 10 7 conidia/mL) of each tested isolate with five replicates was prepared using 0.1 M KNO 3 . The optical density (OD) of each experimental unit, referred to as the OD total , was determined at 660 nm using a spectrophotometer. Subsequently, a 6 mL suspension of each experimental unit was shifted to a universal bottle containing 2 mL of hexadecane. After agitation (20 s), the mixture was transferred into a separation funnel for aqueous phase separation. Subsequently, the optical density of the aqueous phase, referred to as the OD aq , was determined at 660 nm using a spectrophotometer. The relative conidial hydrophobicity of all the replicates (five) of each isolate was determined using the following equation: Relative conidial hydrophobicity (%) = 100 {1 − (OD aq /OD total )}

Mortality
Twenty-five newly-molted (8th instar) R. ferrugineus larvae were immersed for 10 s into the 30 mL suspension in glass beaker at concentrations of 1 × 10 6 , 1 × 10 7 , 1 × 10 8 , 1 × 10 9 , and 1 × 10 10 conidia/mL of each studied isolate for use in concentration mortality bioassays under laminar air flow cabinet. Five replicates were prepared with each replicate comprising twenty-five larvae. Each replicate-inoculated larva was individually incubated in separate 400 mL plastic cups (perforated) containing artificial diets (Table 5) to minimize mortality as a result of the cannibalism observed in a preliminary study. The experimental units were incubated under a 16-h light photoperiod at 30 ± 1 • C and 75% ± 5% RH. The bioassays were repeated over time at different occasions.
The LT 50 value of each tested isolate was determined using a single concentration of 1 × 10 7 conidia/mL. Fungus inoculated insects and control (immersed into 0.04% Tween 80) larvae were fed in plastic cups (400 mL) containing diet (Table 5) under 16-h light photoperiod at 30 ± 1 • C and 75% ± 5% RH. Each larva was provided 1.5 g of artificial diet daily. However, the mortality was recorded until 15 days post-inoculation. Each experimental unit was assessed daily to record mortality and replace the old diet with a fresh diet. Five replicates were prepared with each replicate comprising twenty-five larvae. Each larva was individually incubated in separate 400-mL plastic cups (perforated) to minimize the chances of mortality as a result of the cannibalism observed in a preliminary study. The dead larvae were carefully transferred under a laminar airflow cabinet into a sterilized Petri dish containing a dampened filter paper. Outgrowths on the surface of the mycosed larvae were inoculated onto PDA and observed under a microscope to confirm the causal agent of larval mortality. The bioassays were repeated over time at different occasions. Control mortality was adjusted using Abbott's formula [28]. The corrected percent mortality data were angularly transformed. In addition, the corrected angularly transformed cumulative percent mortality data were analysed using repeated measure ANOVA and Fisher's LSD test for means comparison [29]. The lethal time to kill 50% (LT 50 ) for red palm weevil larvae and the lethal concentration to kill 50% (LC 50 ) for red palm weevil larvae was calculated using probit analysis.

Evaluation of the Host Immune Defence Mechanism against Entomopathogenic Fungi through the Quantification of Immune-Related Genes Using qRT-PCR
RPW larvae (8th instar) were infected with each isolate suspension (1 × 10 7 conidia/mL by Tween 80, London, UK) after immersing five larvae for 10 s. Each larva was regarded as one replicate. Infected and control larvae were individually fed an artificial diet in an incubator in separate 400-mL perforated plastic cups under a 16-h light photoperiod at 30 ± 1 • C and 75% ± 5% RH. After 24 h, larvae were dissected in saline to extract total RNA from the fat body using a commercial kit (Cat # 73404; Qiagen, Hilden, Germany). Total RNA was reverse-transcribed using a commercially available kit (Cat # 6110; TaKaRa Clontech, Paris, France). All primers were designed from the gene sequences deposited in NCBI, as listed in Table 6. The quantification was performed, following the protocol of the SYBR ® Premix Ex Taq™ II (Clontech: RR820W) according to the manufacturer for the CFX96 Real-Time System (Bio-Rad, London, UK). The results of each experimental unit were compared with those of the control through the relative fold-expression obtained after transforming the obtained results into absolute values using the 2 −∆∆Ct method [30]. The relative expression of each gene was set to 1 for the uninfected (control) treatment. Due to the stability of the Beta actin, it was used in the current study as house-keeping gene. Two factor completely randomized design and Fisher's LSD test was used to analyse data in SAS (Dubai, UAE) [26].

Impact of Entomopathogenic Fungi on the Growth of R. ferrugineus
Eighth-instar (newly molted) R. ferrugineus larvae were infected with suspension (1 × 10 7 conidia/mL in Tween 80) of each isolate after immersion for 10 s. The infected and control (immersed into 0.04% Tween 80) larvae were separately fed a measured artificial diet in plastic cups and incubated under a 16-h light photoperiod at 30 ± 1 • C and 75% ± 5% RH. After feeding on the diets for 72 h, the final larval weight, remaining diet and frass weight after drying were measured. These weights were used to calculate the AD efficacy of ECD, and efficacy of conversion of ECI on a dry matter basis [5]. Five replicates were prepared, with each replicate comprising 25 larvae incubated in separate perforated plastic cups. The feeding performance bioassays were also repeated over time. The data for the growth indices were analysed using ANCOVA with SPSS version 11.5 (SPSS Inc., Chicago, IL, USA, 2002). In all cases, the treatments (fungal isolate) were considered to be independent variables. The dependent variables and covariates were assigned according to Scott et al. [31]. In the case of AD, the food (artificial diet) consumed by the larvae was considered as the covariate, while the food (artificial diet) consumed minus frass produced by the larva was considered as the dependent variable. For ECI, the mass gained by the larva was considered as the dependent variable, and the food (artificial diet) consumed by the larva was considered as the covariate. For ECD, the mass gained by the larva was considered as the dependent variable, and the food (artificial diet) consumed minus frass produced by the larva was considered as the covariate.

Conclusions
In summary, these results revealed that isolates B6884 and M9374 are highly virulent and impart high mortality, disturb larval growth, and elicit high expression of immune-related genes. Differences in larval growth, immune response, and mortality among infected larvae might be greatly affected through relative conidial hydrophobicity and germination; however, other factors, such as toxin production, might also play an important role. Future research should focus on the host-pathogen genome interaction to fully explore this mechanism.