Discussion
Our results demonstrated that the host specialized aphid can rapidly adapt to new host plant under assistance of phytopathogen. This is a new pathway for insect herbivores adapting to novel host plants. AlthoughA. gossypii born on hibiscus cannot directly colonize cucumber, it can colonize cucumber through the short feeding experience on cucumber infected with a biotrophic phytopathogen. Under natural conditions, possibility of cucumber being infected by phytopathogens always exists, which provides chances for A. gossypii adapting to cucumber. Thus, we have figured out the life history of A. gossypii that infests cucumber (Figure 4).
Cucumber-specialized and cotton-specialized (or hibiscus-specialized) lineages of A. gossypii were the most studied host specialized populations in this species. Liu et al. (2002) found nymphs of A. gossypii from cotton survived for less than 6 d on cucumber, and nymphs of the aphid from hibiscus survived for less than 4 d on cucumber, and nymphs of the aphid from cucumber survived for less than 6 d on cotton either; but nymphs of the aphid from cotton could adapt to hibiscus and vice versa. A study from Australia indicated the alate adults from cotton produced significantly less nymphs on cucumber or pumpkin than on cotton or hibiscus, and the nymphs deposited on cucumber or pumpkin could not develop to maturity; the alate adults from pumpkin produced significantly less nymphs on cotton or hibiscus than on cucumber or pumpkin (Najar-Rodríguez et al. 2009). The existence of fitness trade-offs in host-associated populations of A. gossypii were also reported in several other studies, such as Carletto et al. (2009), Razmjou et al. (2010) and Satar et al. (2013).
Our results were basically in line with the results of those studies. Although the HI lineage in our study did not die out within a few days after being transferred to cucumber, they only maintained very low population growth and their nymphs developed into yellow dwarfs. Even on natal hosts, A. gossypii developed into yellow dwarfs when the population density was high (Watt and Hales 1996). However, the yellow dwarfs of the HI lineage on fresh cucumber were different from those dwarfs under high population density because they did not inflict damage symptom to cucumber plant (Figure 1D). The bad performance of the HI lineage did not improve with the extension of culture time on fresh cucumber. Therefore, the HI lineage used in the present study showed high host specialization as reported in other studies.
Host specialization is widespread in polyphagous insect herbivores (Drès and Mallet 2002). Genetic adaptation and transcriptional adaptation have been proposed to explain how polyphagous insect herbivores adapt to diverse host plants (Mathers et al. 2017). Genetically distinct host-associated populations were reported frequently in aphids. For example, more than 10 host races with genetic differentiation were detected in Western Europe populations of Acyrthosiphon pisum(Peccoud et al. 2009). The cotton-melon aphid A. gossypiicollected from different plant families at large geographical scale were genetically structured by host species (Carletto et al. 2009). In the sugarcane aphid Melanaphis sacchari , host transfer experiments demonstrated the existence of fitness trade-offs, and genetic test revealed genetic structure linked to host plants (Nibouche, et al. 2015). Also, in Myzus persicae , microsatellite DNA analysis revealed genetic divergence between host-associated populations (Margaritopoulos et al. 2007). Genetic basis of host-specialized populations reflected the outcome of long-time coevolution between insects and host plants. However, a recent study showed that M. persicae was able to adapt to diverse host plants in the absence of genetic basis through rapid transcriptional plasticity of genes (Mathers et al. 2017). The HI lineage of A. gossypii was highly specialized on hibiscus and cotton, manifesting as high mortality, very low population growth, abnormal morphology, and more importantly, inflicting no damage symptom to cucumber plant. The feeding experience of as short as two generations on the pre-infected cucumber could make the HI lineage fully adapt to cucumber. We considered that this rapid adaptation was related to transcriptional adaptation rather than genetic adaptation. Some genes related to aphid-plant interaction could be stimulated in this rapid adaptation process.
During feeding, aphid secreted saliva which contained a range of salivary proteins into plant (Will et al. 2007). Some of them acted as effectors that inhibited plant defenses, such as preventing sieve element occlusion (Elzinga et al. 2014; Yates-Stewart et al. 2020; Guo et al. 2020). Meanwhile, plants could recognize some salivary proteins to elicit defense responses, which was thought to determine the host range of aphids (Rodriguez and Bos, 2013; Boulain et al. 2019). Similar to aphids, biotrophic phytopathogens, such as cucumber downy mildew,P. cubensis , also secreted proteins or metabolites into plants during infecting to prevent occlusion or apoptosis of sieve element. Aphids and phytopathogens (especially biotrophic phytopathogens) may have high similarities in suppressing plant immune response by secreting proteins or metabolites into plants. The HI lineage of A. gossypii cannot ingest phloem sap on fresh cucumber but can on pre-infected cucumber. It is possible that the infection of P. cubensis disrupted the immune of cucumber cells which facilitated subsequent phloem sap ingestion of the aphid. That is to say, the HI lineage of A. gossypii hitchhiked P. cubensis in overcoming plant defenses. Whether this is the case remains to be verified.