Exclusively winged gynopara and male in A. gossypii
Insects produce alternative wing morphs in response to environment changing [1], in which winged or wingless morphs were mostly observed in aphids species. Considerable studies on wing phenotypic plasticity mostly have been carried out on pea aphid, the non-host-alternating species, in which two wing morphs environmentally induced or genetically determined, namely WLPF and male, existed [8, 27]. However, three wing morphs existed in cotton aphid, the host-alternating species, in which males were exclusively winged and an extra special wing morph, gynopara, were reported in this species [18]. Distinction of wing dimorphism between host-alternating and no-host-alternating aphid species probably result from their diverse survival strategies. The former always alternates between primary (usually woody) host plants and secondary (herbaceous) hosts, while the latter were usually monophagous [28]. Hence, in host-alternating species A. gossypii, gynoparae and males are exclusively winged to migrate between the primary and secondary hosts. By contrast, not all males were winged in Acyrthosiphon pisum, possibly because this species was not entirely necessary to migrate in winter [8]. Certainly, winged parthenogenetic females were produced for population expansion in both host-alternating and non-host-alternating species in summer [28].
Besides, male wing differentiation in pea aphid were determined by a single locus on the X chromosome called aphicarus [8]. However, molecular mechanism about wing differentiation of gynopara and male in cotton aphid were still not well known. Species exhibiting three alternative wing morphs provide valuable models for studying wing differentiation in aphids and some shared physiological or molecular baselines probably exist. Here, methods for gynopara and male induction in laboratory were established by shortening photoperiod in cotton aphid (Fig. 2). Compare to previous studies [18, 29], one generation induction was enough to produce these two wing morphs with a high effectiveness. Furthermore, morphological characters in body and internal genitals were firstly described and discriminated in three wing morphs (Fig. 3). All these results broaden our understanding about wing phenotypic plasticity in aphid species.
Transcriptomes Comparison And DEGs Analysis
We firstly assembled transcriptomes by WLPF, WPF, gynopara and male in cotton aphid and obtained a total of 46 501 unigenes (Table 1), which was approximated to the number of 44 310 in previous study [20]. After filtering out the repetitive genes and genes without annotation, 37 090 unigenes yielded in our study (Table 1), while only 11 350 unigenes were obtained in Liu’s study. This may result from the difference in wing morphs selection strategies. Only WLPF, gynopara and sexual female were sequenced in Liu’s study, while extra male and WPF were collected in our study meanwhile. Several previously unreported gene transcripts or isoforms probably existed in male and WPF in cotton aphid. In addition, compared to WLPF, gynopara owned 7 270 up-regulated and 3 597 down-regulated DEGs while there were only 741 up- and 879 down-regulated genes in previous study [20]. Different strategies were adopted to eliminate the potential influence of embryos in the mother’s ovaries: embryos were manually removed in Liu’s study while all offspring were born before adult cotton aphid were collected in our study. Compared to these excellent studies performed on cotton aphid wing differentiation, we firstly identified the shared and exclusively DEGs in WPFs, GP, male compared to WLPFs, respectively (Fig. 4).
Pathways potentially involved in wing differentiation of three wing morphs
Several transcriptional studies have focused on aphids to seek signal pathways involved in the wing differentiation. 1 663 DEGs were identified in WPF compared to WLPF in cotton aphid, which were significantly enriched in ribosome, pyruvate metabolism, proteasome, lipid metabolism, protein synthesis and degradation, RNA transport, antigen processing and presentation [29]. Our results were consistent with those finding, in which pyruvate metabolism, antigen processing and presentation were up-regulated in all these three wing morphs compared to WLPF as well (Table S2). 1 620 DEGs were identified in gynoparae compared to WLPF, in which 6 up- and 7 downregulated signaling pathways were enriched, including starch and sucrose metabolism, phototransduction, dorso-ventral axis formation, Wnt, Notch [20]. Likewise, pathway of starch and sucrose metabolism were up-regulated in these three wing morphs (Table S2). This probably results from the indispensability of energy for flight apparatus and flight behavior.
Those studies advanced our understanding of wing dimorphic in cotton aphid. However, Liu’s study mostly was highlighted on reproductive polyphenism while Yang’s study only performed on WPF [20, 29]. In this study, three wing morphs in cotton aphid were firstly compared with each other or to wingless parthenogenetic female. 2 335 shared DEGs including 1 658 up- and 677 downregulated were identified in all three wing morphs compared to WLPF (Fig. 4), which were significantly enriched in 49 up- and 7 downregulated KEGG pathways (Table S2). The upregulated pathways were clustered into categories of signal transduction, lipid metabolism, carbohydrate metabolism, endocrine system (Fig. 5). Energy allocation is important for trade-off between wing morph and wingless morph in aphids [29]. Thus, the upregulated lipid and carbohydrate metabolism in three wing morphs is consistent with the previous observation of the significantly higher triglyceride content in winged morph versus the wingless morph [23]. Besides, compared to WLPFs, all these three alate morphs had up-regulated insulin signaling pathway (Fig. 5), which had been proved regulating wing differentiation and development in several insects including Nilaparvata lugens, Laodelphax striatellus, Sogatella furcifera, Blattella germanica [4, 30, 31]. Considering the importance of insulin in wing determination, relative expression levels of 15 related DEGs in RNA-seq data were validated in all three wing morphs compared to WLPF (Fig. 7), which implied insulin is potentially involved in the wing differentiation of three wing morphs in cotton aphid.
DEGs Associated With Insulin, Flight Muscle And Energy
Insulin receptor 1 (InR1) leads to long-winged morph if active and short-winged morph if active in three planthoppers [4]. Besides, silencing of InR1 disrupts nymph-adult transition of alate viviparous females in A. (Toxoptera) citricidus [26]. InR1 were increased significantly to 6.03, 2.23, 6.70 folds in WPF, gynopara, male compared to WLPF, respectively (Fig. 7, Table S3). This hints the potential importance of InR1 in wing regulation of three wing morphs in cotton aphid. Flightin, a phosphorylated myofibrillar protein essential for thick filament assembly and sarcomere stability in insects flight muscles [32], exhibits the higher transcript accumulation in winged parthenogenetic morphs and male in A. pisum [33], and in winged parthenogenetic females of A. gossypii [29] and Rhopalosiphum padi [34]. The disproportionately high levels of flightin transcript in winged versus wingless morphs likely results from the presence or absence of indirect flight muscles in winged and wingless morphs, which plays a causal role in the morphological divergence of the wing morphs [33]. Particularly, flightin were increased to 2.32, 6.50, 11.75 folds in WPF, gynopara, male compared to WLPF in our results likewise, respectively (Table S3). This imply the potential importance of flightin in flight muscle formation or wing differentiation in three wing morphs in cotton aphid. Besides, phosphoenolpyruvate carboxykinase [GTP]-like (PEPCK) and glycogen phosphorylase-like (glgP) were expressed highly in alate A. citricida adults, and silencing of these two genes individually can all resulted in under-developed wings in WPFs at the rates of 58–79% [35]. Likewise, these two genes were significantly upregulated in all three wing morphs compared to WLPF in cotton aphid (Fig. 7). In addition, odorant receptor co-receptor (Orco) which mediates winged morph differentiation of parthenogenetic female in Sitobion avenae [36], was also higher expressed in all three alate morphs compared to WLPF in cotton aphid (Table. S3).
Taken together, these shared upregulated genes in three wing morphs underline the importance of signaling pathways of insulin, energy generation, orco, flightin in wing differentiation in cotton aphid. The functions of these genes in wing dimorphism in A. gossypii will be confirmed by RNAi strategy in our future study. To our knowledge, this is the first study examining transcriptome-wide patterns of differential transcript accumulation associated with three wing morphs in cotton aphid, which will provide a baseline for future studies on molecular basis of wing differentiation in A. gossypii.