Importin β1 Mediates Nuclear Entry of EIN2C to Confer the Phloem-Based Defense against Aphids

Ethylene Insensitive 2 (EIN2) is an integral membrane protein that regulates ethylene signaling towards plant development and immunity by release of its carboxy-terminal functional portion (EIN2C) into the nucleus. The present study elucidates that the nuclear trafficking of EIN2C is induced by importin β1, which triggers the phloem-based defense (PBD) against aphid infestations in Arabidopsis. In plants, IMPβ1 interacts with EIN2C to facilitate EIN2C trafficking into the nucleus, either by ethylene treatment or by green peach aphid infestation, to confer EIN2-dependent PBD responses, which, in turn, impede the phloem-feeding activity and massive infestation by the aphid. In Arabidopsis, moreover, constitutively expressed EIN2C can complement the impβ1 mutant regarding EIN2C localization to the plant nucleus and the subsequent PBD development in the concomitant presence of IMPβ1 and ethylene. As a result, the phloem-feeding activity and massive infestation by green peach aphid were highly inhibited, indicating the potential value of EIN2C in protecting plants from insect attacks.


Introduction
The gaseous phytohormone ethylene (C 2 H 4 ) regulates plant growth, development, and immunity via a signaling (signal transduction) pathway that uses EIN2 as the central regulator [1][2][3][4][5][6][7][8]. EIN2 is a 141-kD integral membrane protein of 1294 amino acids with structural features suited for ethylene signaling [1]. The N-terminal part of 461 amino acids forms 12 transmembrane domains [1], which enable EIN2 to partake in ethylene signal transduction from the endoplasmic reticulum (ER) membrane towards an intracellular physiological pathway [3]. At the opposite end, the longer C-terminal portion of 833 amino acids [1] contains a putative nuclear localization signal comprising 1262-1269 amino acids, which targets EIN2 or its C-terminal portion into the nucleus [9]. Following translocation, the C-terminal region of EIN2 is sufficient to activate ethylene responses associated with plant growth, development, and immunity [1,7,[9][10][11]. Such a functional C-terminal fragment truncated from the EIN2 sequence was designated as EIN2 CEND [1,12] and is called EIN2C for the sake of convenience hereafter.
EIN2 becomes activated upon ethylene perception by any of the five functionally redundant receptors [4,31,32]. Once ethylene is sensed by an ethylene receptor, EIN2 is dephosphorylated to liberate EIN2Cs [12]. Then, some EIN2Cs move into the nucleus [12] to activate a transcriptional cascade that confers plant immunity and ethylene responses [11,33,34]. EIN2C translocation from the ER into the nucleus is a pivotal step towards the transcriptional cascade. In eukaryotic cells, the import of macromolecules, which are referred to as cargoes and mainly include proteins and RNA, into the nucleus most often requires nuclear import carrier proteins called importins (IMP). IMPs are presented by α and β forms and mediate cargo trafficking most frequently in a complex of either "cargo-IMPβ" or "cargo-IMPα-IMPβ", wherein the IMPα is a substrate adaptor, or, rarely, in a form of "cargo-IMPα", wherein the IMPα directly serves as a transporter [35][36][37]. Arabidopsis genome encodes 8 IMPα and 18 IMPβ proteins [38], but none was associated with the nuclear import of EIN2Cs.
This gap has been bridged by the present study demonstrating the IMP loss-of-function imp Arabidopsis mutants with respect to molecular trafficking engaged in phytohormonemediated basal defense pathways in the plant. By investigating the WT plant, imp mutants including those defected in IMPβ1 (synonym KPNB1), and impβ1-complemented (impβ1/IMPβ1-RFP) transgenic lines, IMPβ1was characterized as an efficient facilitator for the nuclear import of EIN2 454-1294 , which is the first identified EIN2C [1]. The present study was devised to elucidate if IMPβ1-mediated nuclear import of EIN2C also regulates plant PBD against aphid infestations. IMPβ1 and EIN2C were determined to inhibit the phloem-feeding activity of green peach aphid (Myzus persicae Sulzer), a model species of insects attacking Arabidopsis [10,11,25]. Evidence detailed below will demonstrate that IMPβ1 mediates the nuclear transport of EIN2C to activate PBD responses against aphid infestations in Arabidopsis.

IMPβ1 Strongly Affects Arabidopsis Resistance to Green Peach Aphid
This study began with investigating 13 Arabidopsis IMP genes ( Figure 1A), that have loss-of-function mutants generated previously by T-DNA insertion (Supplementary  Table S1). To look for functional links between the IMPs and Arabidopsis defenses against green peach aphid, qRT-PCR was performed to quantify IMP transcript levels in plants of the Arabidopsis ecotype Col-3 during the aphid infestation. Expression of IPT, IMPβ1, IMPα1, and IMPα7 were significantly increased by artificial colonization of the plants with aphids ( Figure 1A). Quantities of IPT, IMPβ1, IMPα1, and IMPα7 transcripts in aphid-colonized plants were 5.8-, 11.0-, 6.6-, and 4.7-times higher than those in aphid-free control plants, respectively. In contrast, the other nine IMP genes did not show significant transcriptional changes in responses to aphid colonization ( Figure 1A). Leaves of 20-day-old plants were artificially colonized with 10-day-old nymphs of green peach aphid (+aphid), while plants in the control group remained free from aphids (−aphid). Two days later, RNA was isolated from leaves and analyzed by qRT-PCR using EF1α as a reference gene. Data shown are mean values ± standard deviation (SD) estimates of results from six independent experiments, each involving 15 plants. Asterisks indicate significant differences between plants with and without aphid infestations. (n = 6, * p < 0.001; Student's t-test.) (B,C) Aphid colony fidelity in two days and productivity rates in five days after placement on leaves of the tested genotypes (listed at bottom). Data shown are mean values ± SD estimates of results obtained from six independent experiments, each involving 200 aphids colonized on 10 plants. The different letters on the graphs indicate significant differences, as assessed using Duncan's new multiple-range test (p < 0.05).
To determine if the IMP genes affect Arabidopsis defenses, colony fidelities and reproduction rates of aphids placed on leaves of the WT and imp plants were conveyed. If the aphid colony fidelity was higher on an imp mutant compared with the WT plant, the IMP was thought to be a positive regulator of Arabidopsis resistance to aphid colonization. The aphid colony fidelity rates on the ipt, impβ1, impα3, impα5, and impα7 mutants were significantly increased compared with the WT plants. Thus, IPT, IMPβ1, IMPα3, IMPα5, and IMPα7 contributed substantially to plant resistance ( Figure 1B). Meanwhile, an IMP was thought inhibitive to aphid reproduction if the reproduction rate was greater on the imp mutant than on the WT plant. The imp mutants that favored the colonization of aphid ( Figure 1B) also supported aphid reproduction ( Figure 1C). The highest rate of aphid reproduction occurred on the impβ1 mutant ( Figure 1C). These data suggested that IPT, IMPβ1, IMPα3, IMPα5, and IMPα7 contribute to Arabidopsis resistance against green peach aphid infestation and that IMPβ1 is the most influential resistance constituent.

IMPβ1 Supports PBD Defense Gene Expression but Does Not Affect Bacterial Infection
Defense response genes PP2-A1, PP2-A2, and GSL5 have been shown to be essential constituents of the PBD [10,11,21,22,25,28]. To determine whether IPT, IMPβ1, IMPα3, IMPα5, or IMPα7 affects Arabidopsis PBD, we conducted qRT-PCR analysis on the leaves of both WT and imp mutant plants. These plants were either free from aphids or artificially colonized with 10-day-old nymphs of green peach aphids. After 24 h, we compared the expression levels of PP2-A1, PP2-A2, and GSL5 in the leaves. The qRT-PCR data indicated that PP2-A1, PP2-A2, and GSL5 were expressed at the steady-state levels in leaves of all plants without aphid colonization, but these genes considerably increased their transcript quantities after aphid colonization in the WT plant ( Figure 2A). Aphid-induced enhancements of PP2-A1, PP2-A2, and GSL5 expression were also found in the ipt, impα3, impα5, and impα7 mutants ( Figure 2A). However, colonization by aphid did not have substantial effects on expression levels of PP2-A1, PP2-A2, and GSL5 in the impβ1 mutant, and did not cause significant changes in transcript amounts of the three genes from the steady-state levels ( Figure 2A).
Similar results were obtained from plants incubated in air and in 10 µL/L ethylene. The ipt, impα3, impα5, and impα7 mutants resembled the WT plants in response to the externally applied ethylene, which highly enhanced expression of PP2-A1, PP2-A2, and GSL5 in leaves of these plants ( Figure 2B). In 24 h after ethylene treatment, expression levels of GSL5, PP2-A1, and PP2-A2 were increased accordingly by five, nine, and six times on average in comparison to the ready-state expression extents. On the contrary, the impβ1 mutant failed to display ethylene-enhanced expression of the PBD response genes. Instead, the expression of PP2-A1, PP2-A2, and GSL5 in leaves of ethylene-treated impβ1 plants remained around the steady-state levels as found in the absence of ethylene treatment ( Figure 2B).
The effects of IMPβ1 on PP2-A1, PP2-A2, and GSL5 expression were confirmed by investigating impβ1-complemented (impβ1/IMPβ1-RFP) transgenic Arabidopsis lines. The impβ1/IMPβ1-RFP lines were generated by transformation of the impβ1 mutant blossoms with the plant binary vector pCAMB1301 [10] that was constructed to carry a recombinant of the native IMPβ1 promoter (IMPβ1P), IMPβ1 coding sequence (IMPβ1), and red-fluorescent protein (RFP) gene (Supplementary Figure S1A). Well-characterized four impβ1/IMPβ1-RFP lines (#1 to #4) shared common characters, resembling the WT plant in growth and development (Supplementary Figure S1B). Here, impβ1/IMPβ1-RFP #1 (simply called impβ1/IMPβ1 hereafter) was compared with the mutant and WT plants in terms of PP2-A1, PP2-A2, and GSL5 expression after plant colonization by aphids or ethylene treatment. Based on qRT-PCR analyses carried out 24 h later, quantities of PP2-A1, PP2-A2, and GSL5 transcripts detected in leaves were significantly increased by aphid colonization or ethylene treatment in impβ1/IMPβ1-RFP plants, as well as in WT plants ( Figure 2C). In both the WT and impβ1/IMPβ1-RFP plants, expression levels of GSL5, PP2-A1, and PP2-A2 gained approximately six-, twelve-, and seven-fold increases by aphid colonization, and about five, nine, and six times by ethylene treatment, respectively. In the impβ1 mutant, however, neither aphid colonization nor ethylene treatment provided evident noticeable in GSL5, PP2-A1, and PP2-A2 expression, which instead remained around the steady-state levels ( Figure 2C).

IMPβ1 Directly Interacts with EIN2C in Plant Nuclei
Our study found that IMPβ1 may be involved in the regulation of PBD by ethylene signaling. EIN2 is a critical factor in the positive regulation of the ethylene signaling pathway. To infer a functional relationship between EIN2 and IMPβ1, the IMPβ1 was subjected to protein-protein interaction assays with EIN2C (EIN2 454-1294 ) using a split-ubiqui- Taken together, these analyses suggest that IMPβ1 is essential, but IPT, IMPα3, IMPα5, and IMPα7 are not, for the PBD response gene expression induced either by aphid colonization or by ethylene treatment in Arabidopsis. In essence, IMPβ1 is likely to partake in ethylene signaling for the PBD regulation. Thus, IMPβ1 is focused in further studies stated hereafter.
We have confirmed that IMPβ1 supports the resistance to aphids in Arabidopsis, and we are interested in exploring whether IMPβ1 affects the resistance to pathogens. To our surprise, IMPβ1 did not affect the plant defense against the bacterial pathogen Pseudomonas syringae pv. tomato (Pst). The WT, impβ1, and impβ1/IMPβ1-RFP plants displayed similar sensitivities to Pst, allowing it to vigorously propagate in leaf tissues and finally cause disease with visible symptoms ( Figure S2A,B). IMPβ1 was also unrelated to the plant resistance against Pectobaterium carotovora subsp. carotovora (Pcc), the bacterial pathogen that causes soft rot in cruciferous plants. Pcc causes necrosis in Arabidopsis leaves after spray inoculation in the absence of water films. Indeed, Pcc bacteria propagated in leaf tissues and caused severe necrosis symptoms irrespectively of the plant genotypes (Supplementary Figure S2C,D). Therefore, IMPβ1 has no visible function in the plant resistance against these virulent bacterial pathogens.

IMPβ1 Directly Interacts with EIN2C in Plant Nuclei
Our study found that IMPβ1 may be involved in the regulation of PBD by ethylene signaling. EIN2 is a critical factor in the positive regulation of the ethylene signaling pathway. To infer a functional relationship between EIN2 and IMPβ1, the IMPβ1 was subjected to protein-protein interaction assays with EIN2C (EIN2 454-1294 ) using a split-ubiquitinbased yeast-two hybrid (SUB-Y2H) system. An interaction occurred specifically between EIN2C and IMPβ1, but not between EIN2C and any of the other IMPs (IPT, IMPα3, IMPα5, and IMPα7) ( Figure 3A). The specificity was further evidenced by the positive control using KAT1 and the negative control using SUC2 ( Figure 3A). To confirm and locate the IMPβ1 and EIN2C interaction, we carried out bimolecular fluorescence complementation (BiFC) assays. IMPβ1 was fused to the N-terminal half of YFP (YFP N ), generating the IMPβ1:YFP N fusion protein, while EIN2N (EIN2N 1−453 ) and EIN2C were fused to the YFP C-terminal half (YFP C ), forming the EIN2N:YFP C and EIN2C:YFP C fusion protein ( Figure 3B). An interaction was observed only between IMPβ1:YFP N and EIN2C:YFP C but not EIN2N:YFP C , and the interaction was found in the nucleus ( Figures 3B and S3). The IMPβ1-EIN2C interaction was corroborated by the luciferase assay performed on leaves of tobacco leaves ( Figure 3C). Clearly, IMPβ1 interaction with EIN2C, creating a molecular basis for the possibility that IMPβ1 mediates EIN2C trafficking into the nucleus.

IMPβ1 Targets EIN2C into Plant Nuclei in Response to Ethylene
Subcellular localization of IMPβ1 protein is affected by ethylene. In the assay, the impβ1/IMPβ1-RFP seedlings were growing continuously in air and were shifted into ethylene, respectively. Twelve h later, leaves were excised from the plants and stained with 4,6-diamidino-2-phenylindole (DAPI), a blue-fluorescent compound that is permeable to membranes, has a high affinity with DNA, and is widely used as a molecular marker of nuclei [39,40]. CLSM clearly visualized the IMPβ1:RFP fusion protein presented in leaf epidermal cells (Supplementary Figure S4A). In those cells, IMPβ1:RFP was found in cytoplasmic and nuclei, before plant treatment with ethylene. After ethylene was applied to the plants, IMPβ1:RFP decreased its localization in the plasma membrane and cytoplasm and more co-localized with DAPI on the nucleus (Supplementary Figure S4A).
Ethylene enhanced the nuclear localization of IMPβ1 protein, which was further confirmed by transient expression assays. In the assay, leaves of WT tobacco plants were transformed with the 35S:IMPβ1:YFP construct. These plants were incubated in air and ethylene, respectively. Leaves of these transformed plants were stained with DAPI, and observed by CLSM. In CLSM imaging, the IMPβ1:YFP fusion protein was localized in cell membranes and nuclei, in the leaves of transformed plants growing in air (Supplementary Figure S4B). In the presence of ethylene, IMPβ1:YFP fusion protein was abundantly expressed, and co-localized with DAPI were colocalized to nuclei in the leaves of transformed plants (Supplementary Figure S4B). Clearly, ethylene facilitates the localization of IMPβ1 in plant nucleus. Consistent with ethylene-facilitated nuclear localization of IMPβ1, EIN2C was found to move into the nucleus in the concomitant presence of ethylene and a functional IMPβ1. In the assay, EIN2C-YFP was linked to P35S ( Figure 4A) and transferred into the WT, impβ1, and impβ1/IMPβ1-RFP plants ( Figure 4B inset 1). These transformed plants were incubated in 10 µL/L ethylene for 40 h. In the subsequent 3 h, EIN2C was constitutively expressed to similar levels in all plants no matter if they were incubated in air or ethylene, but IMPβ1 was expressed only in the WT and impβ1/IMPβ1-RFP plants incubated in ethylene (Supplementary Figure S5). In the WT and impβ1/IMPβ1-RFP plants, the EIN2C-YFP fusion protein started to move into nuclei after 40 min of incubation in ethylene and displayed sharp increases in amounts of the nuclear localization subsequently in 50-170 min ( Figure 4B). Consistently, CLSM of the impβ1/IMPβ1-RFP leaves, the EIN2C-YFP protein was localized to the endoplasmic reticulum (ER) membrane and nuclear envelope, and a minor amount of this protein was found in the nucleus, which was clearly visualized by DAPI, when the plants were incubated in air ( Figure 4B Figure 4B blue curve). We examined the co-localization of Impβ1-RFP and EIN2C-YFP. In air, IMPβ1 and EIN2C were co-located in cell membrane. However, IMPβ1 and EIN2C were co-located in nuclei in plant leaves treated with ethylene (Supplementary Figure S7). Clearly, IMPβ1 targets EIN2C into the nucleus in response to the exogenous ethylene.

IMPβ1-Mediated Nuclear Import of EIN2C Confers PBD Defense Responses
As with the exogenous ethylene, green peach aphid infestation also facilitated the EIN2C-YFP fusion protein localized in the nuclei of WT and impβ1/IMPβ1-RFP plants ( Figure 5A). In WT and impβ1/IMPβ1-RFP plants that were not colonized with aphids, EIN2C-YFP was mostly located in ER membranes, instead of nuclei ( Figure 5A). The nuclear localization of EIN2C-YFP was not detected in impβ1 mutant plants with and without aphid colonization ( Figure 5A). Taken together, these analyses elucidate that IMPβ1 transports EIN2C into the nucleus in response to aphid colonization in Arabidopsis. The aphid-induced nuclear import of EIN2C in the presence of a functional IMPβ1 provides the molecular basis for IMPβ1 to regulate EIN2-dependent insect-deterrent defense responses, which mainly include callose deposition and PBD response gene expression [10,21,22]. We found that these responses were induced in coincidence with constitutive expression of the EIN2C gene and aphid-induced expression of the IMPβ1 gene and the innate EIN2 gene in P35S:EIN2C:YFP-transformed WT and impβ1/IMPβ1-RFP plants. In both plants, the expression of IMPβ1 was significantly increased by artificial colonization with aphids in contrast to the ready-state expression levels detected in the aphid-free control ( Figure 5B). Aphid-induced expression of the EIN2C was significantly inhibited in impβ1 plants compared with WT and impβ1/IMPβ1-RFP plants colonized with aphids ( Figure 5B).
In leaves of the WT and impβ1/IMPβ1-RFP plants, but not in the impβ1 mutant, infestation with green peach aphid induced strong expression of EIN2-dependent insect-deterrent defense response genes ( Figure 5C). The PDF1.2 gene is a molecular marker of the ethylenemediated insect-deterrent plant-defense mechanism [41]. The expression of PDF1.2 is subject to ethylene signaling and provides a broad spectrum of resistance to herbivores, including leaf-eating and sap-sucking insects [11,42]. In contrast, the PBD response genes PP2-A1, PP2-A2, and GSL5 are more specific in function against phloem-feeding insects, typically including aphids [21,22,25,28,41,43]. The aphid-induced expression of these PBD response genes ( Figure 5C) coincided with enhanced callose deposition in leaves of the plants carrying a functional IMPβ1 ( Figure 5D,E). Taken together, these results suggest that the PBD activation is dependent on IMPβ1.

IMPβ1-Conferred PBD Inhibits Phloem Feeding and Massive Infestation by Aphids
The PBD induced by green peach aphid began to impede aphid feeding activities on Arabidopsis ( Figure 6A). Feeding from plants by sap-sucking insects undergoes several phases that can be monitored by an electrical penetration graph (EPG) instrument in real time as distinct waveforms [21,22,25,44,45]. EPG-monitoring over 4 h (Supplementary Table S4) showed that green peach aphid feeding did undergo these major phases ( Figure 6A), and that the aphids fed longer and more often from the phloem of the impβ1 mutant than from the WT and impβ1/IMPβ1-RFP plants ( Figure 6B). Over 4 h, the aphids fed in the phloem phase (Ph2) for a total of 97.4 ± 17.7 min from the impβ1 mutant, but only 11.8 ± 0.8 min on WT and 4.3 ± 0.3 min on the complemented line.
Assessment of aphid colony fidelity ( Figure 6C) and reproductivity rates ( Figure 6D) showed that the WT and impβ1/IMPβ1-RFP plants were more resistant than the impβ1 mutant to infestation by the aphid. Colony fidelity is the number of aphids remaining in the leaf colonies within 24 h, while reproductivity rate is the number of nymphs produced by an adult within five days after artificial colonization [21,22]. The aphid colony fidelity ( Figure 6C) and reproductivity rates ( Figure 6D) were significantly increased in the impβ1 mutant compared with the WT and impβ1/IMPβ1-RFP plants. Over 24 h, the aphid fidelity averaged 72% in the WT, 71% in impβ1/IMPβ1-RFP, and 92% in impβ1 plants, with about 22% higher in the mutant. In five days, totally 26, 23, and 58 nymphs, on average, were produced by an aphid adult individual in leaf colonies of the WT, impβ1/IMPβ1-RFP, and impβ1 plants, respectively. In other words, the presence of a functional IMPβ1 gene in the plants provided >55% resistance against aphid population growth. As a result of altered feeding, colonizing, and reproductive behaviors of the aphids, the impβ1 mutant incurred more severe infestations and displayed leaf chlorosis and necrosis ( Figure 6E). The leaf infestations by aphids caused substantial reductions in net photosynthesis rates shown as A N variant ( Figure 6F) and plant biomass ( Figure 6G). These results demonstrate the critical role of IMPβ1 in the establishment of aphid-induced PBD, which develops in relevance to aphid-induced nuclear import of EIN2, impedes phloem-feeding activities, and reduces infestation of the plant. The EPG waveform patterns shown here represent feeding behaviors of eight aphids placed on each of the three plant genotypes. The non-puncturing phase (NP) indicates that the stylet remained outside the cuticle. Cell puncturing (probe) usually leads to the pathway phase (path) in which the stylet penetrates between cells en route to the vascular tissue [45]. A successful path navigates the stylet to the phloem phase (Ph2) to absorb the phloem sap. In order to prevent protein clogging inside the sieve element, E1 salivation (green lines) first ejects watery saliva [45,46]. Second, E2 saliva (red lines) is added to the ingested sap, thought to prevent phloem proteins, mainly phloem protein 1 (PP1) and phloem protein 2 (PP2), from clogging inside the capillary food canal [44]. During the feeding process, mechanical problems with stylet penetration into the plant tissues, namely, derailed stylet mechanics shown as F, may occur due to a deficient saliva composition [22,47]. Both path and F delay the time to the Ph2 and prevent ingestion of phloem sap [47]. Following a smooth Ph2, the xylem phase (XP) may proceed while aphids try to suck water from the xylem [10,25] to reduce osmotic pressure caused by increased sucrose concentrations in ingested phloem sap [44]. (B) Summation of Ph2 time as mean values ± SDs of results from three independent experiments, each involving eight aphids (n = 24 aphids). (C) Aphid colony fidelity (means ± SDs, n = nine biological repeats). (D) Aphid productivity rates (means ± SDs, n = nine biological repeats). (E) Symptoms of aphid infestation in the different plants 10 days after artificial colonization. Each photo represents nine plants. (F) Leaf net photosynthesis rate (A N ) measurements (means ± SDs, n = six leaves). (G) Plant biomass (means ± SDs, n = 15 plants). In F and G, 30-day-old plants were colonized with or remained free from aphids and the measurements were performed two weeks later. In (B-D,F,G), the different letters on the graphs indicate significant differences, as assessed using Duncan's new multiple-range test (p < 0.05).

EIN2C Complements the ein2-1 Mutant in PBD Responses
To look for the functional connection between IMPβ1-mediated nuclear import of EIN2C and ethylene-induced EIN2-regulated PBD responses, EIN2C was transformed in plants of the Arabidopsis WT, impβ1/IMPβ1-RFP, ein2-1, and impβ1 single mutants, and ein2-1 impβ1 double mutant. These plants were transformed with P35S:EIN2C:YFP or remained untransformed in control and then incubated in air and in 10 µL/L ethylene, respectively. Analyses by qRT-PCR performed 24 h later of the ready-state and ethyleneinduced expression of the genomic EIN2C ( Figure 7A) and total expression of the genomic and introduced EIN2C fragments ( Figure 7B) in untransformed and transformed WT, impβ1, and impβ1/IMPβ1-RFP plants. The qRT-PCR data also confirmed the ready-state and ethylene-induced expression of the genomic EIN2 in untransformed ( Figure 7C) and transformed ( Figure 7D) plants of WT, impβ1, and impβ1/IMPβ1-RFP. In particular, the introduced EIN2C was well expressed in the transformed plants of the ein2-1 and ein2-1 impβ1 mutants as in the other genotypes ( Figure 7B). First, in plants growing under regular conditions, IMPβ1 participates in vegetative growth, possibly by targeting specific cargoes related to growth and development, but this hypothesis remains to be verified. Second, in plants treated with ethylene or incurring aphid infestations, IMPβ1 mediates the nuclear entry of EIN2C to confer ethylene response. Third, IMPβ1-guided nuclear import of EIN2C is also critical for activation of PBD, which effectively inhibits the phloem-feeding activities and massive infestations of aphids on the plants. Question marks (in the nucleus) indicate unanswered questions, namely, to what extent EIN2C is associated with the regulation of expression of ethylene-regulated transcription factors and defense genes [2,3,5,7,11,28,31,48], or with the proteasome activities that degrade ethylene-signaling repressors to facilitate signal transduction [14,21,22,25,28,33,34,49].
Foliar expression levels of IMPβ1 and PBD response genes were quantified. In the absence of ethylene, the expression of IMPβ1 ( Figure 7E) and PBD response genes, including PP2-A1, PP2-A2, and GSL5 ( Figure 7F-H), remained around the steady-state levels all plants. In the presence of ethylene, P35S:EIN2C:YFP-transformed WT, impβ1/IMPβ1-RFP, and ein2-1 plants well supported the PBD-response gene expression ( Figure 7F-H). However, ethylene displayed higher extents to induce the gene expression in the WT and impβ1/IMPβ1-RFP plants than in the ein2 mutant ( Figure 7F-H). Furthermore, the gene expression was considerably inhibited in the impβ1 single mutant and impβ1 ein2-1 double mutant, even in the presence of ethylene ( Figure 7F-H). These differences suggest that transient constitutive expression of EIN2C:YFP temporarily compensates the ein2-1 mutant defects in sufficient expression of the PBD-response genes only when ethylene and a functional IMPβ1 are concomitantly present in the plants. This compensation caused strong inhibitions to the phloem-feeding activities of aphid ( Figure 7I), resulting in marked reductions in the aphid infestations ( Figure 7J) in ein2-1 plants incubated with ethylene in contrast to the incubation in air. In this mutant, the introduced EIN2C provided inhibitions to aphid feeding and population growth as strong as in the WT and impβ1/IMPβ1-RFP plants ( Figure 7I,J). On the contrary, the engineering EIN2C introduced into the impβ1 and impβ1 ein2-1 plants failed to execute any inhibitory effects on aphids, indicating that EIN2C functions downstream of IMPβ1 in ethylene signal transduction towards defense responses.
While aphid infestations substantially reduced leaf photosynthesis rates, photosynthesis was further inhibited by the exogenous ethylene and introduced EIN2C, as evidenced by variations in the A N values detected from the plants with and without ethylene treatment, EIN2C introduction, and aphid infestations ( Figure 7K). In contrast, IMPβ1 did not evidently affected photosynthesis, as evidenced by equivalent A N levels in the plants that carry or lack a functional IMPβ1 ( Figure 7K). While growth of the impβ1 mutant was considerably impaired as compared with that of the WT plant, growth inhibition was further caused by the exogenous ethylene ( Figure 7L). These analyses suggest that de novo expression of EIN2C can complement the ein2-1 mutant also in photosynthesis and vegetative growth in addition to PBD responses.

Discussion
This study has focused on the function of IMPβ1 in nuclear import of EIN2C linked to plant PBD against aphid infestations ( Figure 7M), a distinct model of biological interactions characterized by stylet penetration of plant cells and the cellularly specified responses to this unique stress on plant fitness [21][22][23][24][25][26]. Such a functional relationship between IMPβ1 and EIN2C is also recognized as a fascinating question for the central role that EIN2 bears in ethylene signal transduction towards plant growth, development, and immunity [1,[3][4][5][6][7]. Now the scientific community commonly appreciates that EIN2C liberation from the fulllength EIN2 sequence and subsequent trafficking to the nucleus represent pivotal events in the intricate ethylene signaling networks [2,7,9,12,13].
Since EIN2 454-1294 was demonstrated to be a sufficient regulator of ethylene signaling in Arabidopsis [1], many EIN2C fragments have been characterized [7,12], and some fragments verified to regulate ethylene responses associated with plant growth and development [7,9,[12][13][14][15]. Increasing studies have been devised to elucidate the functional relationship between subcellular localizations and biological performance of the lengthvaried EIN2Cs and even the full-length sequence of EIN2 in plants [7,9,[12][13][14][15]. The present study extends the functional scope of the first identified EIN2C, namely, EIN2 454-1294 [1], to plant defenses against insect infestations. Genetic, molecular, and cytological data obtained in this study clearly demonstrate that this EIN2C is navigated by IMPβ1 into plant nuclei to confer EIN2-dependent PBD responses under induction either by the exogenous ethylene or by aphid infestation ( Figure 7M). Especially, EIN2C can complement the impβ1 mutant, restoring it to the WT in EIN2-dependent PBD responses. Considering these findings, researchers in the scientific community would not find it difficult to comprehend that via nuclear import of EIN2C, IMPβ1 are imported via nuclear transport for regular plant growth and development, physiological responses, and immunological responses, respectively ( Figure 7M). IMPβ1 may be a constituent of normal growth and development (Supplemental Figure S1B), at least affecting vegetative growth. The transition to flowering marks a key adaptive developmental switch in plants which has an impact on their survival and fitness. It is necessary to further identify any other physiological processes affected by IMPβ1. It is particularly necessary to characterize the functional relationship between IMPβ1 or IMPβ1-guided EIN2C trafficking and any regulators of the floral transition pathways that interplay with plant hormones to determine flowering time [50].
Explaining the relationship between subcellular localization and biological performance of EIN2Cs and full-length EIN2 in plants has been a fascinating problem. The different EIN2C species were found in association with the ER, in the nucleus, or over the cytoplasm [9]. By studying five EIN2Cs, either localizing them to the ER membrane or associating them with the nuclear fraction, Wen and colleagues proposed that the ethylene signal promotes the cleavage of the C-terminal portion from ER-located EIN2, and facilitates its nuclear localization to stabilize the EIN3 protein [9]. They determined that the nuclear localization of the EIN2Cs is sufficient to activate EIN3-mediated transcription and ethylene responses [9,51]. Similar results were obtained by Zhang and colleagues. They determined that an EIN2C was cleaved from EIN2 and moved into the nucleus to facilitate jasmonate-induced leaf senescence [15]. Zhang and colleagues proposed "an alternative model of ethylene signaling" [7]. In this model, EIN2 prevented from activation is targeted by its interacting proteins to the 26S proteasome to be degraded in the absence of ethylene; in the presence of ethylene, EIN2 is released from the inhibition to activate ethylene signaling by trafficking from the ET to the nucleus, while EIN2C species assist EIN2 trafficking and the subsequent ethylene responses [7]. Our study demonstrating that IMPβ1 mediates EIN2C entry to the nucleus in response to ethylene treatment or aphid infestation agrees with the recently proposed "alternative model of ethylene signaling".
IMPβ1-mediated EIN2-dependent insect-deterrent responses characteristic of PBD effectively impede massive infestations of aphids in Arabidopsis, indicating the potential value of EIN2C in protecting crops from insect attacks. Coincidently, evidence [52][53][54][55] has shown that some physiological regulators function not only in growth and development but also in defense responses, possibly bridging the problem of growth-defense tradeoffs, or the costs to fitness that accompany defense responses [26,55,56]. While nucleocytoplasmic trafficking is the most important transportation route inside plant cells, the mechanisms supporting this trafficking need further exploration [52][53][54][55][56]. The functions of IMPβ1 in EIN2C trafficking link plant growth and defense regulation ( Figure 7M) now offer a mechanism that could be used to bridge the growth-defense dichotomy. These molecules could be used to concomitantly improve both defense and productivity in plants, especially crops [55,57]. Since the impβ1 mutant is compromised in both growth and defense, IMPβ1 loss-of-function has negative consequences for both sets of processes, rather than creating a trade-off between them. Future studies will focus on engineering the overexpression of IMPβ1 homologs in crops with the aim of synchronized enhancements of crop productivity and immunity.

Plant Material and Growth Conditions
Arabidopsis ecotype Col-3, the ein2-1 single mutant, and 25 imp mutants were previously generated in the Col-3 background. Their seeds were purchased from TAIR (The Arabidopsis Information Resource at www.arabidopsis.org accessed on 12 March 2009). The ein2-1 impβ1 hybrid and impβ1/IMPβ1-RFP transgenic Col-3 lines were generated in the HD lab and used in F6-selfing homozygous generation in this study. Seeds were germinated in flat plastic trays filled with a plant growth substrate. Three days later, germinal seedlings were moved into Φ7-cm pots (1-3 plants per pot) filled with the same substrate. Seed germination and plant growth were accommodated in environmentally controlled plant growth chambers under 24 ± 1 • C, 250 ± 50 µmol quanta/m 2 /s illumination, and a photoperiod circle of 8 h light and 16 h dark.

Aphid Cultures
A single isolate of M. persicae was collected from field-grown radish (Raphanus sativus), near Nanjing in China. A clone of apterous agamic females was obtained by acclimatization in WT Arabidopsis grown in the chamber. The subsequently formed colonies were maintained in nursery Arabidopsis seedlings and were transferred to fresh plants every two weeks in the HD lab located at the Nanjing Agricultural University Weigang Campus (2000-2019) and the Shandong Agricultural University North Campus (since 2018). Uniform 10-day-old aphids were used in this study and were transferred to experimental plants with a fine paintbrush.

Plant Colonization
For Arabidopsis colonization, uniform 10-day-old M. persicae aphids were placed on the top two expanded leaves of plants (10 aphids per leaf). A total of 1200 aphids were monitored in six independent experiments for each genotype and for each single combination of treatment and plant genotype. For each treatment, 200 aphids were placed on 20 leaves (two leaves on 10 plants). Aphid movement from leaf colonies was monitored for five days, and the number of aphids in a leaf colony was scored at 24 h intervals [21,22,25]. The number of nymphs that moved away from colonies was also counted. The proportion of aphids staying in a leaf colony was regarded as colony fidelity. Aphid reproduction was surveyed twice a day by counting newborn nymphs. The reproduction rate was quantified as the ratio between the total number of nymphs produced in five days and the total number of aphid adults that stayed in leaf colonies during the same period [21,22,25].

Gene Expression and PBD Analyses
Total RNA was extracted via the RNA-easy Isolation Reagent Kit (Vazyme, R701-01, Nanjing, China). Plant gene expression was quantified by real-time quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) performed under strict experimental designs (Supplementary Table S2). All the qRT-PCR analyses were conducted using primers (Supplementary Table S3) of high specificities, which were verified by the melting-curve method [22], and using the constitutively expressed EF1α gene as a reference [10]. The qPCR experiments were conducted on the ABI QuantStudio 3a 96 Real-Time PCR system (Thermo Fisher, Waltham, Massachusetts, USA) using the ChamQ TM Universal SYBR ® qPCR Master Mix (Vazyme, Q711-02, Nanjing, China). Relative expression levels of the tested genes (IMPβ1, EIN2C, related variants, and defense response genes, listed in Supplementary  Table S3) were quantified as ratios of their transcript amounts to the EF1α transcript quantity. Callose visualization was performed on leaves as previously described [11].

Ethylene Treatment
Arabidopsis plants were treated with gaseous ethylene in 15 L glass vacuum chambers using the well-established protocol [58,59]. The container has a dome top cover with a valve in its center that allows vacuum removal of air and administration of ethylene.
Before use, all interfaces of the individual components of the container were daubed with petroleum jelly to ensure complete sealing. For treatment, pots containing plants or agar plates containing seeds were placed into the glass container, the container was closed and some air was pumped out. Ethylene gas at the final concentration of 10 µL/L was injected into the container using a syringe and needle through the valve. After ethylene gas was pushed into the container, the valve was left open for a few seconds, allowing the outside air to enter the container so that the total air volume was brought back to the regular level. The valve was closed, and the container was moved into the plant growth chamber to incubate the plants or seeds. Untreated plants were placed in different vacuum containers, but ethylene was not applied.

Aphid Feeding Behavior Monitoring
Aphid feeding activities were observed by the EPG technique [43,44] using the Giga Amplifier system (EPG Systems, Dillenburg, 12, 6703CJ, Wageningen, The Netherlands). Uniform 10-day-old aphids were placed on the upper side of the upper two expanded leaves of an Arabidopsis or wheat plant. The aphids were monitored in three independent experiments. Each experiment involved a total of eight aphids tested with one aphid per leaf using eight plants. Immediately after aphids were placed on leaves, a 20-mm diameter gold wire was attached to the dorsal surface of each aphid's abdomen using silver conductive paint. The other end of the wire was connected to an eight-channel Giga-8 direct current amplifier with eight channels and a 10 9 -Ω input resistance in an electrical circuit that is also connected to the plant via an electrode placed in the soil. The behavior of individual aphids was monitored for 4 h. Voltage waveforms were digitized at 100 Hz with an A/D converter USB device. Waveform patterns were identified according to previously described categories [25,44]. Briefly, the nonpuncturing phase (NP) indicates the stylet is outside the cuticle. Cell puncturing (probe) leads to the pathway phase (Ph2) in which the stylet penetrates between cells en route to the vascular tissue [21,22,25,45]. When the phloem is not a favorite source for feeding, the xylem phase (XP) may be observed while aphids try to suck sap from the xylem [21,22,25].

Gas Exchange Measurements
Gas exchange in the second and third leaves from the top of plants was measured with the LI-6800 photosynthesis system (LI-Corp Biosci, Lincoln, NE, USA). Detailed measurements on single leaves were performed following the manufacturer's instructions and previously described experimental procedures [39,60]. During measurements, relative humidity in the leaf chamber (2 cm 2 for Arabidopsis and 6 cm 2 for wheat) was constantly maintained at 45% and the leaf temperature was kept at 25 • C. CO 2 concentrations at the inlet and outlet of the leaf chamber were monitored by the non-dispersive infrared gas analyzer installed in the system. Photosynthetically active photon flux density was controlled by adjusting intensities of the lamp-house irradiation. Readings of A N were documented automatically by the LI-6800 monitor system integrated into the LI-6800 system.

Genetic Complementation
The genetic complement was constructed in the plant binary vector pCAMBIA1301 [10]. The full-length sequence (1-4634) of the canonical WT IMPβ1 gene was linked N-terminally with its own promoter sequence of nucleotides -2000 to -1 and linked C-terminally with the RFP gene sequence. Plant transformation with the recombinant vector and molecular characterization of transgenic lines were performed by conventional protocols [10,25] and T3 homozygous progenies were used in this study.

Protein-Protein Interaction Assays
As a first step, each of the four IMPs (IMPα1, IMPα7, IMPβ1, and IPT) were cloned into the pGADT7 bait vector, and EIN2C was cloned into the pGBKT7 prey vector of the Dualsystems' split-ubiquitin yeast two-hybrid system (Dualsystems Biotech, Schlieren, Zurich, Switzerland). For YFP BiFC, previously constructed pCAMBIA1301-YFP N and -YFP C plasmid vectors were used in gene recombination. An IMP gene was fused to YFP N between the KpnI and XbaI restriction sites, whereas EIN2C was linked to YFP C using the KpnI and BamHI recognition sites. Similar operations were used in luciferase assay except for replacing YFP N and YFP C with Luc N and Luc C , respectively. The presence and absence of molecular interactions were determined using previously described protocols [34,39].

Subcellular Localization of IMPβ1 and EIN2C
The different chimeric genes were separately constructed in the plant binary vector pCAMBIA1031, transferred into GV3101-bacterial cells, and transiently expressed in plants using a previously described protocol [61]. Alternatively, plants were co-transformed with each of these constructs and the IMPβ1:RFP construct. 48 h after plant transformation, the EIN2C:YFP fusion protein in leaf cells was visualized by CLSM [10,57]. DAPI was used to stain nuclei and applied in an aqueous solution to immerse tested leaves 10 min before observation [61].

Statistical Analysis
Analysis of variance, student's t-tests, and Duncan's new multiple-range tests [62] were performed with GraphPad Prism 8.0 (https://www.graphpad.com/, accessed on 3 August 2019) to determine significance of differences in paired and multiple data from different plants or treatments.