Genomic identification and expression analysis of nuclear pore proteins in Malus domestica

The nuclear pore complex (NPC), comprised of individual nucleoporin (Nup) proteins, controls nucleo-cytoplasmic transport of RNA and protein, and is important for regulating plant growth and development. However, there are no reports on this complex in fruit tree species. In this study, we identified 38 apple Nups and named them based on the known Arabidopsis thaliana homologs. We also completed bioinformatics analyses of the intron and exon structural data for apple Nups. The proteins encoded by the apple Nups lacked a universally conserved domain. Moreover, a phylogenetic analysis separated the apple and A. thaliana Nups into three groups. The phylogenetic tree indicated that MdNup54 and MdNup62 are most closely related to genes in other Rosaceae species. To characterize the 38 candidate Malus domestica Nups, we measured their stage-specific expression levels. Our tests revealed these proteins were differentially expressed among diverse tissues. We analyzed the expression levels of seven apple Nups in response to an indole-3-acetic acid (IAA) treatment. The phytohormone treatment significantly inhibited apple flowering. A qRT-PCR analysis proved that an IAA treatment significantly inhibited the expression of these seven genes. A preliminary study regarding two members of the Nup62 subcomplex, MdNup54 and MdNup62, confirmed these two proteins can interact with each other. A yeast two-hybrid assay verified that MdNup54 can interact with MdKNAT4 and MdKNAT6. On the basis of the study results, we identified apple NPC and predicted its structure and function. The data generated in this investigation provide important reference material for follow-up research.

Scientific Reports | (2020) 10:17426 | https://doi.org/10.1038/s41598-020-74171-0 www.nature.com/scientificreports/ in phenotypes that are similar to those due to mutations to Nup160 and Nup96 15,16 . In addition to altering auxin signal transduction, a mutation to Nup160 also increases the responsiveness of A. thaliana to ethylene, suggesting that it may help mediate the interaction between auxin and ethylene signals 18 . In terms of abiotic stress responses, both Nup160 and HOS1 participated in chilling stress by regulating CBF gene [19][20][21] . Furthermore, The HOS1 protein can specifically mediate the degradation of the ICE1 protein under cold conditions, thereby weakening A. thaliana responses to low temperatures 21 .And the Nup85 mutant reduced ABA and salt stress response in A. thaliana 22 . Additionally, Nup160, Nup96, and HOS1 also affect the flowering time of plants 3 . Specifically, HOS1 interacts with some nuclear genes to regulate its binding to FLC chromatin in flowering plants at low temperatures and weaken the transcriptional inhibition of FLC by HDA6 23 . In A. thaliana, HOS1 interacts directly with CO. In the hos1 mutant, CO accumulates, which inhibits FLC expression and ultimately promotes flowering 24 . Mutations to Nup54, Nup58, Nup62, Nup136, and Nup160 result in an obvious early flowering phenotype in A. thaliana, whereas mutations to Nup62-2 and Nup160-4 lead to dwarfism 1,2 . An investigation of A. thaliana proved that Nup96 promotes the stability of HOS1, which binds to and degrades CO, resulting in delayed flowering. Moreover, HOS1 increases the stability of Nup96, thereby maintaining this regulatory pathway to control flowering time 25 . A lot of research has been done on the model plants, which gives us a certain understanding of the plant NPC 3,4,26 . However, there is no report on NPC research of woody fruit trees. Considering the important function of NPC, it is necessary to carry out the related research on woody fruit trees. Apple is one of the most important fruit tree species worldwide. So we first identified candidate apple Nups based on A. thaliana Nups, after which we characterized the gene structure, protein structure, and tissue-specific expression patterns. We know that most apple species produce relatively few flowers or have stunted flower buds, which seriously affects the apple industry 27,28 . Although previous research has confirmed that some Nups in A. thaliana are involved in the flowering pathway, there have been no similar studies of apple Nups. So we made statistics on the effect of IAA treatment on apple flowering rate and detected the expression of some Nups to preliminarily reveal the effect of apple Nups on flowering. In addition, we conducted a preliminary study on the Nup62 subcomplex of apple. We studied the interaction between MdNup62 and MdNup54 and screened for proteins that interact with MdNup54. To the best of our knowledge, this study is the first comprehensive survey of the apple NPC, and the data presented herein will be useful for future analyses.

Materials and methods
Plant materials and treatments. The roots, stems, leaves, buds, flowers, and fruits of 6-year-old apple trees (Fu ji/T337/Malus robusta Rehd.) were collected for a tissue-specific gene expression analysis. We collected newly grown lateral roots (1-2 mm in diameter), new shoots (2-3 mm in diameter) near the tip, fully expanded leaves near buds, flower buds, blooming flowers, and young fruits, which were immediately frozen in liquid nitrogen and stored at − 80 °C for later use.
Regarding the hormone treatment, 40 apple trees (108° 04′ E, 34° 16′ N) growing in the experimental orchard of the Horticulture College of Northwest A&F University were randomly divided into two groups, which were treated with 300 mg/L IAA or water. During the study, apple leaves were dusted with a low-pressure manual duster, and samples were collected at 30, 50, and 70 days after flowering. The samples were immediately frozen in liquid nitrogen and stored at − 80 °C.
We also investigated the effects of an IAA treatment on the flowering rate of apple trees. Specifically, five similarly growing IAA-and water-treated apple trees were examined. The flowering rate was calculated as previously described 29 .
Identification of apple NPC. To identify apple NPC, we used the 30 identified NPC protein sequences of A. thaliana as queries to search the apple genome database (Malus domestica Genome GDDH13 V1.1, https :// www.rosac eae.org/). The obtained sequences were then used as queries to search the conserved domain database (https ://www.ncbi.nlm.nih.gov/Struc ture/cdd/wrpsb .cgi). The genes lacking the relevant Nup domain were eliminated. All non-redundant putative protein sequences were finally manually checked to confirm the presence of the Nup domain.

Results
Genome-wide identification of NPC in apple. The apple Nups were detected and identified in the GDR database using BlastP. We obtained 38 candidate apple Nups after the genes with incomplete Nuprelated domains and recurring genes were eliminated. The identified apple Nups are Table 1). Figure 1 also shows the gene locus, location, sequence length and other information of apple NUPs.

Gene structures in apple Nups.
To structurally characterize the identified apple Nups, we generated exon-intron diagrams and revealed the coding sequences and untranslated regions ( Fig. 1) Additionally, we also predicted the tertiary structures of the Apple Nups, revealing α helices, β sheets, and random coils in all proteins (Fig. S1).

Expression levels of apple Nups in various tissues.
To functionally characterize apple Nups in apple, we completed a qRT-PCR assay to determine apple Nups expression levels in diverse tissues (flowers, buds, leaves, roots, stems and fruits) (Fig. 5). Because there are 11 pairs of highly similar homologous apple Apple Nups expression patterns in response to IAA treatments during the flower induction period. We investigated the effect of an IAA treatment on apple flower induction. The flowering rate following the IAA treatment was 41.9%, which was significantly lower than the 50.4% flowering rate after the water (control) treatment (Fig. 6). Previous studies have confirmed that some Nups (AtNup54, AtNup62, AtNup96, AtNup160, and AtHOS1) affect flowering time in Arabidopsis thaliana 1,23-25 . And the tissue-specificity analysis found that MdNup54, MdNup62, MdNup133, and MdNup160 were most highly expressed in buds, and MdNup93 was also highly expressed in buds, suggesting that these genes may be involved in the apple flowering pathway. So we analyzed the transcription of these seven candidate MdNups . And the expression of all seven candidate genes was significantly inhibited by the IAA treatment (Fig. 7). Moreover, the transcription of MdNup54, MdNup62, MdNup96, MdNup133, MdNup160 were significantly inhibited at 30, 50, and 70 days after flowering. In contrast, the MdNup93 and MdHOS1 expression levels were not significantly different following the IAA and water treatments at 50 days after flowering, but the expression levels were lower in the IAA-treated samples than in the control samples at 30 and 70 days after flowering. Accordingly, the IAA treatment can significantly inhibit the expression of these genes.
MdNup62 interacts with MdNup54. Previous studies found that Nup54, Nup58, and Nup62 form a complex and function together in metazoan, and the interaction between the other two members in A. thaliana Nup62 subcomplex, AtNup58 and AtNup62, were also reported 31 . And MdNup62 and MdNup54 form the Nup62 subcomplex, we hypothesized that these two proteins interact with each other. To test this hypothesis, we performed a Y2H experiment. First, we observed that the truncated MdNup62 1-507 was self-activating, but MdNup62 508-613 was not (Fig. S2). Therefore, MdNup62 508-613 -pGBKT7 was selected as the bait and was included in a co-transformation of yeast cells along with MdNup54-pGADT7. And they could grow normally on SD/ − Trp/ − Leu medium and SD/ − Trp/ − Leu/ − His/ − Ade/ + X-α-gal medium, and bacame significantly blue on SD/ − Trp/ − Leu/ − His/ − Ade/ + X-α-gal medium. But the co-transformation of MdNup62 508-613 -pGBKT7 and empty-pGADT7 could only grow on the SD/ − Trp/ − Leu medium, and could neither grow nor turn blue on the SD/ − Trp/ − Leu/ − His/ − Ade/ + X-α-gal medium. So the Y2H assay confirmed that MdNup62 can inter- www.nature.com/scientificreports/ act with MdNup54 (Fig. 8). This interaction was further verified in a split-LUC complementation assay. The co-expression of MdNup62-NLUC and MdNup54-CLUC resulted in higher LUC activity than the other combinations (Fig. 8). These results confirmed the interaction between MdNup62 and MdNup54. Thus, we conducted a Y2H assay to explore the biological processes MdNup54 may contribute to. First, we observed that the truncated MdNup54 1-90 was self-activating, but MdNup54 175-339 was not (Fig. S3). Accordingly, MdNup54 175-339 -pGBKT7 was selected as the bait and inserted into yeast cells, which were then transformed with plasmids from the apple bud plasmid library to screen for interacting proteins. Both MdKNAT4 and MdKNAT6 were detected as potential interacting proteins. The MdKNAT4 and MdKNAT6 sequences were cloned and ligated to separate pGADT7 vectors. To conduct a Y2H assay, we co-transformed yeast cells with MdKNAT4-pGADT7 or MdKNAT6-pGADT7 and MdNup54 175-339 -pGBKT7. And these two kinds co-transformed yeast cells could both grow normally on SD/ − Trp/ − Leu medium and SD/ − Trp/ − Leu/ − His/ − Ade/ + X-α-gal medium, and became significantly blue on SD/ − Trp/ − Leu/ − His/ − Ade/ + X-α-gal medium. But the control could only grow on SD/ − Trp/ − Leu medium. The assay results verified that both MdKNAT4 and MdKNAT6 can interact with MdNup54 (Fig. 9).

Discussion
The NPC controls the communication between the nucleus and cytoplasm, with consequences for diverse biological processes that influence plant growth and development. The relatively few studies that have examined plant NPC have been limited to model species, such as A. thaliana. Therefore, only the A. thaliana NPC has been systematically identified. We know very little about the corresponding apple genes. Thus, we identified and attempted to functionally characterize the apple NPC.

Genome-wide identification and characterization of Nups in apple.
We identified 38 candidate apple Nups and more than 30 A. thaliana genes. Almost all of the A. thaliana Nups had a corresponding homologous sequence in apple, suggesting these are conserved plant genes. However, apple homologs of the A. thaliana Nup58 gene was not detected (Fig. 10). A. thaliana Nup58 belongs to Nup62 subcomplex, which means that lacking Nup58 of the apple Nup62 subcomplex must be functionally different from that of A. thaliana. The lack www.nature.com/scientificreports/ of Nup58 gene is bound to lead to differences in the function of the apple and A. thaliana. NPC. Additionally, because apple is a woody plant species, its genome is more complex than that of A. thaliana, with 11 genes (Nup35, Nup50, Nup93, Nup96, Nup98, Nup107,Nup136, Sec13, Seh1, NDC1, and RAE1) having two alleles each, whereas only Nup98 and Nup50 had two alleles in A. thaliana 2 . Although there are some differences between the apple and A. thaliana NPC, the basic structures of the encoded proteins are the same, including the Nup62, Nup93, and Nup107-160 subcomplexes as well as other Nups inside and outside the nuclear pore 1,2 . These findings imply the apple and A. thaliana NPC are functionally similar. The details regarding the introns and exons of the apple Nups suggested there is a lack of similar structures among the genes, including between the genes within the same subcomplex, further demonstrating the relative functional independence of the apple Nups. Apple Nups encode a class of compounds with no conserved domain among all members, unlike the members of other apple gene families (e.g., IDD, GRF, GASA and SBPbox genes) 27,28,32,33 . Only some of the apple Nups encode a conserved domain. Therefore, these proteins may have similar functions. Conserved domains were not detected among the other apple Nups, implying a lack of functional redundancy.
We also analyzed the phylogenetic relationships between Nups. First, we constructed a phylogenetic tree based on the A. thaliana and apple Nups and divided 69 Nups into three groups, consisting of 20, 30, and 19 members (Fig. 3). The apple Nups are most closely related to the corresponding A. thaliana Nups (e.g., MdNup96a/b and MdCG1 are most closely related to AtNup96 and AtCG1, respectively). These results suggest Nups are conserved and may have similar functions in diverse species.
The Nup62 subcomplex occupies an important position in the nuclear pore (i.e., central pore channel). An analysis of the evolutionary relationships involving MdNup54 and MdNup62 indicated that both genes are closely related to genes in other Rosaceae species, especially to genes in Pyrus communis, which is in the same subfamily as apple.
Apple Nups expression patterns. We performed a qRT-PCR assay to study the expression levels of apple  Arabidopsis thaliana studies have confirmed that some Nups (AtNup54, AtNup62, AtNup96, AtNup160, and AtHOS1) affect flowering time 1,23-25 . However, little is known about their potential roles in apple-induced flowering. Therefore, we investigated the expression patterns of MdNup54, MdNup62, MdNup93, MdNup96, MdNup133, MdNup160, and MdHOS1 to preliminarily explore whether they are associated with IAA-mediated flowering. We determined the flowering rates of 'Nagafu No. 2' treated with IAA. Our data indicated that the IAA treatment significantly inhibited flowering, which was consistent with the results of previous studies on apple and other species 34,35 . We subsequently performed a qRT-PCR assay to quantify the expression of these genes. The expression levels of the IAA-treated plants were significantly lower than those of the controls, suggesting these seven apple Nups are responsive to the application of exogenous IAA. Earlier investigations revealed that A. thaliana Nup62, Nup96, and Nup160 genes are involved in the auxin signaling pathway 36 , which is consistent with the results of this study. However, we cannot determine whether apple Nups are involved in the IAA treatment resulting in the reduction of flowering rates. And we only speculate that apple Nups are involved in the IAA regulation of apple flowering pathway. But it is certain that IAA treatment will reduce the expression of apple Nups and flower rates. And the relationship between apple Nups and flower rates needs to be verified by follow-up experiments.
Preliminary functional characterization of the Nup62 subcomplex in apple. The apple Nup62 subcomplex has only two members (MdNup54 and MdNup62), whereas the corresponding complex in A. thaliana has three members (AtNup54, AtNup58, and AtNup62) 1,2 . And we confirmed that MdNup54 and MdNup62 interact in apple. In other words, the complete biological functions of MdNup54 and MdNup62 may require the interaction between these two proteins. Additionally, because of a lack of Nup58, the apple Nup62 subcomplex and the corresponding A. thaliana subcomplex may be functionally diverse.
We conducted a Y2H assay to verify the interactions between MdNup54 and two members of the KNOX family (MdKNAT4 and MdKNAT6). To the best of our knowledge, this is the first study to reveal an interaction between an NPC and members of the KNOX family. The KNOX family members have crucial functions  www.nature.com/scientificreports/ related to plant hormone signaling [37][38][39] , as well as leaf 40,41 and flower development 42 . The apple MdKNAT4 and MdKNAT6 genes are homologs of the A. thaliana AtKNAT4 and AtKNAT6 genes, respectively. Earlier studies proved that AtKNAT4 influences seed dormancy 43 , while AtKNAT6 plays an important role in maintaining meristem integrity and flowering 42 . Apple MdNup54 may also affect similar pathways. Notably, the KNOX family is involved in cytokinin and gibberellin signaling pathways [37][38][39] . These two hormones are closely related to apple flowering 27,44,45 , suggesting that apple MdNup54 may indirectly affect cytokinin and gibberellin signaling pathways by controlling the transport of KNOX genes into the nucleus, thereby regulating apple flowering. But previous studies have shown that selective transport between the nucleus and cytoplasm depends on nuclear transport receptors (importin and exportin), which bind cargos and interact with the NPC selective barriers for cargo transport 3,4,46 . Recently, researchers have found that Nup85 and MED18 can interact directly with each other. And the two mutants had the same abiotic stress phenotype 22 . Therefore, the relationship between Nup54 and KNOX genes may provide a hypothesis for NPC studies that Nups may interact directly with transcription factors to control their nuclear transport.