Aphidius gifuensis Ashmead (Hymenoptera: Braconidae) is a common solitary endoparasitoid of many agricultural and horticultural pest aphids including Myzus persicae (Sulzer), Aulacorthum solani (Kaltenbach) and Sitobion avenae (Fabricius) (Yang et al., 2011; Pan and Liu, 2014). Due to its highly parasitic efficiency, it has already been successfully used to control M. persicae on tobacco in China and been considered to be a potential biological-control agent for effective IPM programs in field and greenhouse (Yang et al., 2011).

Under natural conditions, there are several important determinants restricting the efficiency of the application of A. gifuensis including high temperature and raining (Liu et al., 2016). Generally, the temperatures under greenhouse and field conditions showed irregular cyclic variation daily, and the midday temperature often exceeds 40°C for a period of time. The brief heat stress could cause a degree of physiological and ecological damage to pests and their natural enemies (Dong et al., 2013; Sentis et al., 2015). When temperature was higher than 35°C, no female progeny were produced in Cotesia vestalis (Shi et al., 2013). In Aphelinus asychis, both of the adult survival and longevity are decreased significantly when the temperature was above the 37.5°C (Wang et al., 2016). Furthermore, the decreased mummified aphids and female progeny were also detected at this temperature in A. asychis (Wang et al., 2016). Therefore, in greenhouse, the high temperature in summer is the key limiting factor in the application of natural enemies.

Numerous studies of insect-thermal stress interaction have revealed that insect have evolved complex protective mechanism to protect themselves against the high temperature. Heat-shock proteins (HSPs) and antioxidant enzymes were the most well-known effectors in this process (Yang et al., 2010; King and MacRae, 2015).

HSPs are well known as stress proteins and molecular chaperones participating in protein folding, localization and degradation to influence essential process such as protein synthesis, cell signaling, transcription, and metabolism (Feder and Hoffmann, 1999; Sørensen et al., 2003; King and MacRae, 2015). On the basis of molecular mass and sequence homology, HSPs have been divided into several families including the small heat-shock proteins (sHSPs, the molecular weights ranging from 12 to 43 kDa), HSP60, HSP70, HSP90, and HSP10 (Feder and Hoffmann, 1999; Shi et al., 2013; Nguyen et al., 2016). sHSPs are well distributed across tissues and thought to be the first line of cell defense by preventing irreversible denaturation of substrate proteins under biotic and abiotic stress conditions (Kim et al., 1998; Feder and Hoffmann, 1999; Sørensen et al., 2003). Compared to other HSPs, sHSPs exhibit a greater variation in sequence, structure, size, and function. HSP70s have be divided into inducible (Hsp70) and cognate forms (Hsc70s), which are involved in protein translation, folding, unfolding, translocation, and degradation (Qiu et al., 2006). HSP90s participate in the folding, maintenance of structural integrity, and the proper regulation of a subset of cytosolic proteins, and account for 1% of the soluble protein in most tissues, even in the absence of stress (Picard, 2002).

As stress proteins, HSPs are involved in protecting proteins under oxidation and hypertonic stress, which were induced by extreme temperatures, UV, xenobiotic exposures and parasitoid infestation (Shi et al., 2013; Baruah et al., 2014; Zhang L. J. et al., 2015; Chen et al., 2016). For example, in Harmonia axyridis, sHSPs were thought to play important roles in the cold hardness process (Wang et al., 2017). The gene expression of HSP70 gene family in Rhopalosiphum padi and Melitatea cinxia was significantly induced under thermal stress (Luo et al., 2015; Li et al., 2017). Besides the well investigated stress responses, recently research progresses have revealed that HSPs are associated with diverse molecular and physiological functions such as oogenesis, embryo development, diapause and signal transduction (Sørensen et al., 2003; King and MacRae, 2015). For example, knock down of HSP83 in Acyrthosiphon pisum significantly reduced its fecundity, longevity and the number of viviparous offspring (Will et al., 2017).

Apart from HSPs, antioxidant enzymes in insect including superoxide dismutase (SOD), catalase (CAT), peroxidases (POD) and glutathione-S-transferases (GST) are the other immune system involved in the oxidative damage response (Lopez-Martinez et al., 2008). These antioxidant enzymes can scavenge the thermal stress, UV, xenobiotic exposures and parasitoid infestation induced surplus reactive oxygen species (ROS) (Lopez-Martinez et al., 2008; Yang et al., 2010; López-Martínez and Hahn, 2012; Wang et al., 2012; Zhang S. Z. et al., 2015; Ali et al., 2016). SOD is the most important antioxidant enzymes in the enzyme defense system against ROS. SOD catalyses the disputation of superoxide radicals into oxygen (O₂₎ and hydrogen peroxide (H₂O₂); then H₂O₂ is converted by both CAT and POD into oxygen and water (H₂O).

In previous work, the effect of low temperature (supercooling point and freezing point) on the parasitic potential of A. gifuensis was studied (Liu et al., 2016). In this work, we evaluated the biological performance of A. gifuensis under a short heat stress. Based on our results and previous studies, we could suggest an integrated pest management strategy with complementary diverse natural enemies. In addition, we unraveled the potential contribution of HSPs and antioxidant enzymes to thermal stress in A. gifuensis.

In the application process of natural enemies in eco-agricultural system, thermal stress is one of the key limited factors. In the present work, we not only evaluated the perspective of thermal stress on parasitoid performance but also screened the key potential molecular mechanism involved in this process.

In this work, we found that brief thermal stress negatively affected the parasitoid performance including the hatching rate, survival rate, parasitic capacity, longevity, and female proportion in offspring. Our results showed that female adults were more tolerant to a brief heat stress than male, which was derived from the higher survival rate and longer longevity of female adult in treatment temperatures. The similar results were found in Aphelinus asychis and Aphidius avenae when they were exposed to a brief heat stress (Roux et al., 2010; Wang et al., 2016). When exposed to 40° and 41.5°C, the mean percent survival of A. asychis females was greater than those of males (Wang et al., 2016). However, in A. asychis, the longevity of female adults was significantly longer than that of male adults in all treatment temperatures (Wang et al., 2016). Meanwhile, the emergence rate of A. gifuensis pupae was strongly affected by the heat stress, which was consistent with the depletion of Aphidius colemani and Trichogramma brassicae pupae exposed to a brief heat stress aaand (Hoffmann and Hewa-Kapuge, 2000; Colinet and Hance, 2009). In addition, the parasitic capacity of A. gifuensis female was negatively affected by the heat stress. This phenomenon was also observed in other parasitic wasps. For example, the number of mummified aphids produced by A. asychis females decreased from 413.4 to 71.8 when treatment temperature increased from 25° to 41.5°C (Wang et al., 2016). In this work, our findings highlighted the significance of the negative effects of thermal stress on A. gifuensis performances and recommended the suitable application temperature is under 30°C. In addition, there was a highly survival rate from 30° to 41°C in Propylaea japonica (Zhang S. Z. et al., 2015). Due to the different heat tolerance of A.gifuensis, A. asychis, A. avenae and P. japonica, we constructed an integrated application of these natural enemies of aphid. When temperature under 30°C, we release A. gifuensis, A. asychis and A. avenae to control the aphid in greenhouse and filed. When the temperature is from 30° to 35°C, we could use A. asychis and P. japonica to suppress the quantity of pest aphid. If the temperature above 35°C, P. japonica was the best choice for the biological control of aphid. The integrated application of natural enemies in greenhouse will not only enhance the efficiency of their biological agents but also maximize the benefits through the cost reduction of these biological agents.

Furthermore, we found that at 35°C, A. gifuensis female adults performed better than males, which was consistent with the previous results A. avenae and A. asychis. The potential mechanism of why female exhibited a higher tolerance to thermal stress were: (1) the body size. Numerous studies have revealed that large individuals of both sexes are expected to live longer, to have higher mating success, higher fecundity, produce more daughters, and have a better dispersal ability than small ones (Sagarra et al., 2001; Ellers and Jervis, 2003; Doyon and Boivin, 2005; Santolamazza-Carbone et al., 2007). And, Esperk et al. (2016) have found the body size of Sepsis punctum positively affected heat tolerance. And the body size of female adults were larger than male (Wilbert, 1969; De Block and Stoks, 2003; Teder and Tammaru, 2005; Stillwell et al., 2010). Furthermore, female is more sensitive to environment condition (Teder and Tammaru, 2005); (2) the hydrocarbon, wax layer and lipid of insect cuticle (Denlinger and Hallman, 1998). In Drosophila melanogaster, the rearing temperature influenced the cuticular hydrocarbons profiles and the cuticular hydrocarbons exhibited the sex-specific profiles (Rajpurohit et al., 2017); (3) response genes. In previous work, several HSPs has been identified to be highly expressed in the ovaries and embryonic tissues of females that is absent in male reproductive structures (Palter et al., 1986; Folk et al., 2006; Krebs and Thompson, 2006; Will et al., 2017). In this work, we analyzed the expression profiles of HSPs and antioxidant enzyme genes to investigate their roles in the tolerance to thermal stress or sex-specific tolerance.

The HSPs are molecular chaperones and comprise a large family of proteins involved in the protection against various forms of cellular stress (King and MacRae, 2015). In this work, we identified five HSPs including AgifHSP10, AgifsHSP, AgifHSP70, AgifHSP70-4, and AgifHSP90 based on the transcriptome sequencing (Kang et al., 2017a). In previous work, the vital roles of sHSPs, HSP70s and HSP90s in responding to thermal and pesticide stresses have been well documented in many insect species (King and MacRae, 2015; Sun et al., 2016). For example, in Apolygus lucorum, the expression of AlucHSC70 was significantly induced by cyhalothrin or extremely high temperature whereas it was decreased significantly in treatments of chlorpyrifos or extreme cold temperature (Sun et al., 2016). Similarly, in Leptinotarsa decemlineata, exposure to an extreme temperature of 43°C significantly induced the expression of LdecHSP70 whereas heat stressed larvae of L. decemlineata failed to respond to imidacloprid by producing more HSP70 (Chen et al., 2016). The up-regulation of HSC70 promotes a greater thermal tolerance in Nilaparvata lugens (Lu K. et al., 2016). In this work, we found that both AgifHSP70 and AgifHSP70-4 were significantly induced under the thermal stress. More interestingly, the expression of AgifHSP70-4 in female was higher than that in male at 35°C, which was consistent with the higher survival rate of female adult at this temperature. All of these results suggested that AgifHSP70-4 might be the key factor of temperature resistance in A. gifuensis.

Heat-shock protein 90 (HSP90) is a highly conserved molecular chaperone found in all species except for Archaea, which is required not only for stress tolerance but also for normal development (King and MacRae, 2015). For example, AlucHSP90 was not only an important gene for A. lucorum adults in response to extremely high temperature, but also involved in the resistance or tolerance to cyhalothrin, imidacloprid, chlorpyrifos, and emamectin benzoate, especially for female adults to emamectin benzoate and for male adults to cyhalothrin (Sun et al., 2014). In Acyrthosiphon pisum, ApisHSP83, which was the homologous genes of HSP90, played pleiotropic roles in embryogenesis, longevity, and fecundity (Will et al., 2017). Knocked down of ApisHSP83 resulted in the reduction of adult survival and the number of nymphs born per aphid, which appears to be in striking agreement with the role of the homologous HSP90 in the longevity of D. melanogaster and Tribolium castaneum (Knorr and Vilcinskas, 2011). In D. melanogaster, Hsp83 molecular chaperone complex regulated the nuclear import of methoprene-tolerant (Met), which is required for juvenile hormone signal transduction (He et al., 2014). In this work, the expression of AgifHSP90 was strongly induced by the temperature increasing. And the proportion of female was significantly decreased, which was consistent with Anisopteromalus calandrae (Nguyen et al., 2013). All of these results suggested that AgifHSP90 might be involved in the regulation of longevity and reproduction in A. gifuensis under the thermal stress.

Besides HSP70s and HSP90s, sHSP is another well investigated HSPs in insect, which are assumed to play an important role in the heat stress, metamorphosis, normal development, diapause, and immune responses (King and MacRae, 2015). For example, Csuphsp19.8 and Csuphsp21.7b were both up-regulated dramatically by heat and cold whereas Csuphsp21.5 only be induced by cold stress in Chilo suppressalis (Lu et al., 2014). In Chironomus riparius, small heat shock protein, HSP27 was significantly activated by heat stress and xenobiotic exposures including bisphenol A and CdCl₂ (Martínez-Paz et al., 2014). Additionally, in D. melanogaster, DmelHSP22 up-regulated not only in oxidative stress condition but also during aging (Morrow et al., 2016). In this work, AgifsHSP and Agifsl(2)efl were up-regulated in both sexes under a short heat stress, and the higher expression of Agifsl(2)efl in female was consistent with the higher resistance of female. All of these results suggested that sHSPs might play an important role in environmental stress response not only thermal stress but also xenobiotic exposures. However, the detail functional investigation is still lacking.

Compared to sHSP, HSP70s and HSP90s, to our knowledge, HSP10 in insects has not been structurally and functionally studied in detail (Jia et al., 2011). In eukaryotes, HSP10, originally identified as a mitochondrial chaperone, now is also known to be present in other places such as cytosol, cell surface, and extracellular space (Jia et al., 2011). In this work, we found that the expression of AgifHSP10 increased at 30°C in both male and female whereas it decreased at 35°C. Consistent with this result, in Sitodiplosis mosellana, the expression of SmosHSP70, SmosHSC70, and SmosHSP90 firstly increased and then decreased with the treatment temperature increasing (Cheng et al., 2016). There are similar trends of FoccHSP90, FoccHSC701, FoccHSC702, and FoccHSP60 in Frankliniella occidentalis and McinHSP70-3 and McinHSP70-4 in Melitaea cinxia (Luo et al., 2015; Lu M. X. et al., 2016). This phenomenon implied that exposed at 35°C destructed the A. gifuensis immune system to fail to produce AgifHSP10.

Beside the protection of HSPs, organisms are equipped with a comprehensive antioxidant defense system to relieve oxidative stress and remedy the damaged macromolecules produced by the exposure to xenobiotics or thermal stress. SOD is the most important antioxidant enzyme against ROS. In Panonychus citri and Propylaea japonica, the high temperature exposure increased levels of SOD and GST (Yang et al., 2010; Zhang S. Z. et al., 2015). In this work, all of the expression of AgifSODs in female significantly increased at 30°C whereas there was a relatively poor activation of AgifSODs in male. Furthermore, at 30° and 35°C, the expression of AgifSOD2, AgifSOD3, AgifPOD, and AgifGSTs in female were higher than that in male. And the activities of SOD, POD, and GST showed similarly trends of this expression patterns. In previous work, POD has been identified to be involved in the response to thermal stress (Zhang S. Z. et al., 2015). And GST was thought to be participated in the inactivation of toxic lipid peroxidation products accumulated due to oxidative damage and xenobiotics treatment (Qin et al., 2013; Feng et al., 2015). The higher expression and activity of GSTs in female suggested that female with a stronger ability of inactivation of toxic lipid peroxidation products in A. gifuensis. Combining with the higher resistance of heat stress in female, we synthesized that the higher activities of SOD, POD and GST are the key factors of the sex-specific heat tolerance. Meanwhile, in Propylaea japonica, high temperature exposure also activated the activities of CAT (Zhang S. Z. et al., 2015). Similarly, in this work, the enzyme activity of CAT significantly increased with the increase of temperature, whereas there were no significant difference between female and male at 30° and 35°C, which was consistent with expression profiles of AgifCAT1 and AgifCAT2. In Antheraea mylitta, C. suppressalis and Bombyx mori, CAT activities also presented a positive correlation with the thermal stress (Cui et al., 2011; Nabizadeh and Jagadeesh Kumar, 2011; Jena et al., 2013). All of these results suggested that the antioxidant enzyme systems play an important role in the antioxidant response under high temperature.

In conclusion, the present study not only highlighted the significance of negative effects of thermal stress on A. gifuensis performance but also explored the potential mechanism of antioxidant response in A. gifuensis. We found that the protection of thermal damage is coordinated by HSPs and antioxidant enzymes. Based on the heat tolerance ability of many aphid natural enemies, we suggested an integrated aphid management in summer. The integrated application of natural enemies in greenhouse will not only enhance the efficiency of their biological agents but also maximize the benefits through the cost reduction of these biological control agents.

Z-WK, H-GT, and T-XL designed the research; Z-WK, F-HL, and XL performed research; X-LT, F-HL, W-BY, S-ZZ, and H-GT provided assistance; Z-WK, F-HL, and H-GT analyzed data; Z-WK, X-LT, and H-GT wrote the manuscript; S-ZZ and T-XL edited the manuscript; and Z-WK and X-LT revised the manuscript.