Although the Cry toxin protein possesses specific insecticidal activity against a variety of agricultural and forestry pests, aphids are not sensitive to it . Therefore, the study aiming to uncover the reasons for this insensitivity will give us valuable implications for evaluating and improving its activity against the aphid. We confirmed that Cry1Ab1 toxic core exerted very low ability for killing M. persicae, then we identified the proteins that bind to the Cry1Ab1 toxic core in M. persicae, and analyzed the docking interface between Cry1Ab1 and its three binding proteins, ATP synthase subunit alpha, beta-actin and Gpn2, performed STRING database searching to find their related pathways, all those suggesting that they formed protein complex to disturb the growth and development of the aphid.
ATPase, which is known as the “pH meter” of the cell, regulates pH to modulate enzymatic activity, thereby affecting the metabolic rate of the cell. V-ATPase on the membrane also functions as a receptor to regulate the pH-mediated endocytosis and exocytosis in cells [29, 30]. According to Whyard et al, the consumption of the V-ATPase dsRNA caused the death of Drosophila melanogaster, Tribolium castaneum, A. pisum and Manduca sexta . ATP synthase subunit alpha consists of a soluble domain at the N-terminus in the cytoplasm and a transmembrane region containing 9 α-helices at the C-terminus. Several amino acid residues located at the C-terminus participate in proton transport and the formation of H+ channels . According to the chemical permeation theory, V-ATPase transports protons, and ATP is synthesized using the H+ potential during the process of oxidative phosphorylation. Based on the analysis described above, ATP synthase subunit alpha is presumed to participate in the proton transport required for ATP synthesis mediated by oxidative phosphorylation. Meanwhile, Docking interface analysis indicated that Cry1Ab1 toxic core bound with the functional domain of ATP synthase subunit alpha, Therefore, the interaction between Cry1Ab1 toxic core and ATP synthase subunit alpha might interfere with oxidative phosphorylation, affecting the normal energy supply in the cells and thereby disrupting the metabolic reactions in the cells and even inhibiting the growth and development of aphids.
Actin has been identified as one of the Cry-binding proteins in Mythimna separate , H. virescens , H. armigera , A. aegypti . Vega-Cabrera also observed an interaction between Cry11Aa toxin and actin on the cytoplasmic side of the membrane after toxin penetration . In the present study, β-actin in M. persicae bound to the Cry1Ab1 toxic core. β-actin is involved in a wide range of cellular activities, including the transport of intracellular substances, cellular shape associated with cell movement, endocytosis, cytokinesis and cell adhesion [36–39]. Notably, Cry toxins belong to the pore-forming toxin family (PFT), which is the most widespread group of toxins produced by bacteria . The interaction of PFT and actin not only enhances actin polymerization in vitro  but indirectly promotes toxin oligomerization and endocytosis . Therefore, we speculated that β-actin mediated the endocytosis of Cry1Ab1 located in cell membrane, and β-actin functioned as a cytoskeletal protein that binds to Cry1Ab1 and transports it to an intracellular site as the cytoplasm flows. β-actin is tightly integrated with RNA polymerase III (RNAP III) by interacting with RPABC3 (RPC62), RPABC2 (RPB6) and RPABC3 (RPB8) on RNAP III . Interestingly, RPABC2 and RPABC3, which are common subunits of RNAP Ⅰ, RNAP Ⅱ and RNAP Ⅲ , were detected to be adjacent to each other in the crystal structure of RNAP Ⅱ , and β-actin has been shown to interact with RNAP Ⅰ . Based on the analysis described above, we concluded that β-actin firmly binds to three RNAP by interacting with adjacent RPABC2, RPABC3 and other subunits, indicating that β-actin was quite important in the formation and function of RNAP. Once Cry1Ab1 bound to β-actin, it might hinder the normal assembly of RNAP, affecting the normal transcription and expression of genes, and then blocking the development and metabolism of M. persicae. Captivatingly, actin has also been reported to function as a receptor of beet western yellows virus particles in aphids, and probably played a role in the transcytosis and transmission of the virus . Using a genomic analysis method, Tamborindeguy found that several actin proteins are potentially involved in the transcytosis of viruses in A. pisum . Thus, after digestion by M. persicae, the binding of Cry1Ab1 to actin likely disturbs the interaction of actin and the virus, affecting virus transmission mediated by M. persicae.
The GPN family, including Gpn1, Gpn2 and Gpn3, is known for its invariant glycine–proline–asparagine structural loop . Gpn2 is required for the proper localization of RNAPII and RNAPIII. As shown in the study by Zeng et al, Gpn2 might interact with RNA polymerases , which have important roles in gene expression, cell fate determination and tissue development by catalyzing the synthesis of mRNA and non-coding RNA in eukaryotes [50–52]. Zeng et al confirmed that Gpn2 and Rba50 are directly involved in the assembly of the Rpb3 subcomplex that bridges the Rpb1 and Rpb2 subcomplexes to produce the core of RNAPII, and even subsequent biogenesis of RNAPII . Because β-actin might also bind tightly to RNAP and be involved in the formation and function of RNAP, the interaction of Cry1Ab1 with Gpn2 or β-actin would probably significantly disrupt the assembly and function of RNAP, as well as DNA transcription and expression, subsequently resulting in some damage to the metabolism and development of M. persicae.
From the above analysis, Cry1Ab1 toxic core interacted with Cry-binding protein including ATP synthase subunit alpha, β-actin and Gpn2, it would eventually affect the metabolism and development of M. persicae. Considering the pull down results and docking analysis, it is likely that Cry1Ab1 toxic core bind with ATP synthase subunit alpha and β-actin or ATP synthase subunit alpha and Gpn2 to form protein complex, and the protein complex probably covered the functional domain of three Cry-binding proteins and dragged the ATP synthase subunit alpha, β-actin or Gpn2 to non-functional area of certain Cry-binding protein, thereby potentially interrupting the normal energy supply, blocking the normal transcription and expression of genes and would eventually affect the metabolism and development of M. persicae.
Zhang et al identified Cry1Ac/Ab in unfed Bt coccinellid Propylaea japonica neonates whose parents fed on aphids reared on Bt cotton , suggesting that Cry1Ac/Ab enters the organelles and cell through parental transfer, and Cry1Ab1 protein was unable to be degraded and activated due to the neutral environment in the midgut of aphid . This study proposed a hypothetical model for the molecular mechanism underlying the effect of Cry1Ab1 on aphids (Fig. 7). From the hypothetical model, Cry1Ab1 protein is unable to kill the aphid, but potentially affects development and reproduction. Some studies indicated an effect of Cry proteins on the growth and development of sucking insects. According to Tan et al, the non-target pest Sogatella furcifera (Homoptera: Delphacidae) exhibits a significantly longer developmental duration than the controls after feeding on insect-resistant transgenic indica rice carrying the cry1Ab gene . Azimi reported a delay in the progression of Bemisia tabaci (Hemiptera) fed on Bt cotton expressing the Cry1Ab protein to adulthood later and a lower level of fertility than insects fed on non-Bt cotton . Additionally, Porcar announced that after ingesting Cry1Ab, the growth rates of surviving A. pisum were markedly reduced compared to the control group . Because the capsule method was a short-term bioassay that did not reflect the long-term effects of the Cry toxin on aphids, the Cry1Ab1 toxic core rarely directly kills aphids, but it might affect the development of aphids and the reproduction of populations, similar to Cry1Ac (86.17% of the amino acid sequence is the same as Cry1Ab1), which is less toxic to M. persicae but inhibits the population growth rate . The explanations for the inhibitory effects, however, remain unknown. Our study provided clues for the molecular mechanism of inhibition, namely, the proteins involved in cellular metabolic processes in M. persicae are bound by the Cry1Ab1 toxic core.
Based on the results reported by Pascale et al, actin is involved in the epithelial transcytosis of virus particles in the aphid vector . This finding and the observation that actin is a common Cry-binding protein in our study and other studies prompted us to speculate that the Cry protein disrupt virus transmission mediated by aphids, providing a new strategy for controlling viral diseases mediated by aphids. Additionally, APN and translation elongation factor, which have been identified as other Cry-binding proteins in the aphid (unpublished data from our lab), were also proven to be involved in the interaction with the virus in aphids [58, 59]. APN is the receptor for numerous coronaviruses, such as porcine epidemic diarrhea virus (PEDV) and porcine transmissible gastroenteritis virus (TGEV) . Linz et al recently confirmed that the coat protein (CP) of pea enation mosaic virus (PEMV) binds to its receptor APN in the gut of the pea aphid . The interaction between elongation factor and viral RNA-dependent RNA polymerase (RdRp) or viral genomic RNA promotes virus replication and transmission . Thus, some aphid proteins are simultaneously bound by viral proteins and Cry proteins, further supporting our hypothesis that the Cry protein disrupts virus transmission mediated by aphids.
Last but not the least, based on the long-term feeding effect of Cry protein expressed in plant on aphid population, and the possibility that Cry protein disrupt virus transmission, when we screen anti-aphid Cry proteins, we should not only examine the short-term killing ability but also focus on the long-term effects on the growth and reproduction and virus transmission by aphids.