Introduction

Apple (Malus × domestica Borkh.) is a widely cultivated and economically important fruit crop in temperate regions worldwide owing to its high nutritional value, good storage, and lengthy supply period. And Fuji apple is the main cultivar in China, but there are cultivation and production problems, including flowering difficulties and severe alternate bearing [1, 2]. However, with global warming, an increase in the average temperature in winter will result in earlier apple flowering [3, 4], and if there is cold weather in early spring, then significant flower and fruit losses will result. Additionally, at present, extreme hot weather occurs frequently in summer, causing other problems, such as growth impairment and production decline [5, 6], which have seriously affected the development of the apple industry in China.

Floral induction pathways have been extensively studied, and there are six signalling pathways in the model plant Arabidopsis thaliana, including photoperiodic, vernalization, autonomic, gibberellin, temperature-sensitive, and age pathways[7,8,9]. In apple, the functions of some key flowering-related genes have been well studied in recent years, such as APETALA1 (AP1), LEAFY (LFY), FLOWERING LOCUS T (FT), and TERMINAL FLOWER 1(TFL1). For instance, overexpression of MdMADS5, a putative homolog of AP1, leads to significant early flowering in Arabidopsis [10]. Apple anti-TERMINAL FLOWER 1 transgenic lines flower significantly earlier than the WT, with the earliest flowering at 8 months, while the WT did not flower for 6 years [11]. Through transcriptome analyses, the induction of apple flower buds was found to be regulated by sugar and hormone signalling pathways [12]. Other omics studies have revealed the molecular mechanisms involved in responses to exogenous treatments, such as sugar [13], 6-benzylaminopurine [14], and gibberellins [15], and their effects on the flowering of apples. However, research on apple flowering is still relatively limited.

A nuclear pore complex (NPC) is composed of a class of nucleoporins (Nups) located in the nuclear pore [16]. More than 30 Nups have been identified in Arabidopsis and 38 members have been identified in apple [16, 17]. Some Nups interact and form three subcomplexes: Nup62, Nup93, and Nup107–160 [16, 18]. Nups control the transport of substances, such as RNA and proteins, between the nucleus and cytoplasm [19, 20], and play important roles in regulating plant growth and development, as well as biotic and abiotic stresses [19, 21, 22]. For example, HOS1, Nup96, Nup54, Nup58, Nup62, Nup136, and Nup160 are important for plant flowering [16, 23,24,25,26]. HOS1, Nup85, Nup96, and Nup133 participate in abiotic stress pathways [18, 20, 27,28,29]. MOS7, Nup96, Nup160, and Sec1 play important roles in plant immunity [30,31,32], and Nup96, Nup160, and TPR affect hormone signalling pathways [33,34,35,36,37].

Heat shock factors (HSFs) are important components of signal transduction and play important roles in diverse stress pathways [38]. The HSF family in plants has more members (21 HSFs in Arabidopsis) and more complex regulatory mechanisms [39, 40] than in vertebrates (4 HSFs) or Drosophila (only 1 HSF). On the basis of their structural differences, HSFs may be divided into three classes, A, B, and C [39]. Class A has the C-terminal short peptide AHA domain, which has an activator function, while the B and C classes lack this domain [41]. HSFs specifically identify and bind heat shock elements (HSEs), which contain nGAAnnTTCn or nTTCnnGAAn in the downstream target genes’ promoters [42]. Class A members (HSFA1a, HSFA1b, HSFA1d, HSFA1e, HSFA2, and HSFA3) positively regulate plant heat tolerance [43,44,45,46,47], while, in contrast, Class B HSFs (HSFB1 and HSFB2b) negatively regulate heat-induced HSFs and plant heat tolerance [48]. In addition to responding to heat stress, some HSFs (HSFA2, HSFA1E, and HSFA4C) appear to be involved in plant flowering pathways [49, 50].

Currently, there are no reported functional studies of Nups in apple. Nup62 is a member of the Nup62 subcomplex in the central core of the nuclear pore [16, 17], and nup62 A. thaliana mutants have been reported to flower early, indicating Nup62’s involvement in flowering pathways [25]. In this study, we characterized apple Nup62, which showed a high transcription level at the flower bud developmental stage and was responed to high temperature. The overexpression of MdNup62 in Arabidopsis resulted in earlier flowering compared with WT. Moreover, The overexpression of MdNup62 in Arabidopsis and tomato both reduced heat resistance. Further, we performed a yeast two-hybrid (Y2H) sieve library experiment to screen for proteins that interact with MdNup62, and the interactions between MdNup62 and the MdHSFs were confirmed. And the overexpression of MdHSFA1d and MdHSFA9b independently in Arabidopsis resulted in earlier flowering and enhancing heat resistance. Thus, MdNup62 and the MdHSFs regulate flowering and respond to temperature changes. These results provide a theoretical reference for managing the impact of global warming on the apple industry.

Results

Apple NPC structure and composition, and its expression patterns

Compared with vertebrate, apple NPC consists of 38 Nup proteins, but missing some Nups, such as Nup153, Nup358, Pom121, etc. Refer to the structure of vertebrate NPC [16], we devided apple NPC into five parts: Cytoplasmic filaments (Nup214 and Nup88), Cytoplasmic and Nuclear ring (Nup98, RAE1, and Nup107-160 Subcompex), Scaffold and central channal (GP210, NDC1, Nup62 Subcompex, and Nup93 Subcompex), Nuclear basket (Nup50 and Nup136), and Distal ring (Tpr/NUA), as well as GLE1, ALADIN, CG1, and HOS1 also participate in NPC constitution (Fig. 1a). Additionally, MdNup62 interacts with MdNup54 on Y2H and LUC experiment, and they might forming the central apple NPC channel involving in nucleocytoplasmic transport (Fig. S1) [17].

Fig. 1
figure 1

The nuclear pore complex (NPC) structure and composition in Vertebrate and Malus. a A schematic of the nuclear pore with the cytoplasmic side at the top and the nuclear basket at the bottom for Vertebrate (left) and Malus (right). b Tissue specific expression patterns of apple NPC components by RNA-seq. The full names of the different abbreviations are as follows, ‘Nagafu No.2’ long branches flower buds (FLB), ‘Nagafu No.2’ short branches flower buds (FSB), ‘Nagafu No.2’ axillary buds (FYB), ‘Qinguan’ axillary buds (QYB), ‘Nagafu No.2’ fruit (FR), ‘Yanfu No.3’ stem tip (YF3J), ‘Yanfu No.6’ stem tip (YF6J), and ‘M9-T337’ root (T337R). Each number after the abbreviation represents a biological repetition

Full size image

We examined the expression patterns of NPC components in different tissues of several apple varieties (Fig. 1b; Table S1). The expression levels of MdNup62 as central channel component showed significantly higher in buds, stem, roots than in fruit of apples, but other channel component MdNup54 showed significantly low expression levels in all tissues compared with MdNup62, indicating that MdNup62 play a key role in regulation of growth and stress response by controlling nucleocytoplasmic transport in apple.

Feature, expression, and subcellular localization analyses of MdNup62

We initially performed a simple bioinformatics analysis of MdNup62. A phylogenetic tree of Nup62 from six Rosaceae plants (Rosa chinensis, Pyrus communis, Prunus persica, M. domestica, Rubus occidentalis, and Fragaria vesca) was constructed using MEGA-X. MdNup62 was most closely related to the Nup62 of pear (Fig. 2a). The aligned protein sequences revealed a conserved Nsp1_C domain (Fig. 2b). The subcellular localization of MdNup62 was determined by introducing 35S::MdNup62-GFP into tobacco leaves (Fig. 2c). Tobacco leaves transformed with the empty vector 35S::GFP were used as controls. In the tobacco leaves expressing 35S::MdNup62-GFP, the GFP signal was observed only in the nuclear pore, while the GFP signal was detected throughout the control tobacco leaf cells, indicating that MdNup62 localized to the nuclear pore.

Fig. 2
figure 2

Identification and analysis of MdNup62. a Phylogenetic analysis of Rosaceae Nup62. b The conservative domain of Rosaceae Nup62. c Subcellular localization of MdNup62. The upper panel shows 35S::EGFP, and the lower panel shows 35S::MdNup62-EGFP. d and e Analyses of MdNup62 expression levels in diverse ‘Nagafu No. 2’ apple tissues d and in different flower bud developmental stages of ‘Nagafu No. 2’ e. (f) The phenotype of ‘Nagafu No. 2’ tissue-cultured seedlings (upper panels) and the in situ accumulation of superoxide radical (O2−) at 0, 1, 3, and 6 h under heat treatment conditions (lower panels). Bar = 1 cm. g MdNup62 expression levels in ‘Nagafu No. 2’ tissue-cultured seedling leaves at 0, 1, 3, and 6 h under heat-treatment conditions. Each sample was analysed with three biological replicates, each comprising three technical replicates. Means followed by different lowercase letters are significantly different at the 0.05 level. The same below

Full size image

The transcript levels of MdNup62 in different tissues were determined using qRT-PCR (Fig. 2d). The highest expression level was in flower buds. An MdNup62 expression analysis during the flower bud developmental stages revealed that the expression level was stable at 30 to 60 days after flowering and reached its highest level at 70 days after flowering (Fig. 2e). Thus, MdNup62 maintained a high expression level during flower bud induction, indicating that it may be related to bud differentiation in apple.

We exposed apple tissue-cultured seedlings to a heat treatment. The reactive oxygen species (ROS) accumulation in leaves increased from 0 to 6 h under heat-treatment conditions (Fig. 2f). Moreover, the expression level of MdNup62 was determined at different times during the high-temperature treatment (Fig. 2g). MdNup62 was significantly induced by high temperature, and its expression level was highest at 1 h after exposure to the high temperature. Thus, MdNup62 may be involved in the heat-resistance pathway of apple.

Overexpression of MdNup62 promotes flowering

To confirm MdNup62's role in flowering, we performed an Agrobacterium-mediated genetic transformation of MdNup62 into A. thaliana. We found that OE-MdNup62 lines flowered significantly earlier than WT (Fig. 3a). Additionally, OE-MdNup62 lines had significantly fewer rosette leaves than WT during bolting (Fig. 3b). The presence of the transgene in OE-MdNup62 lines was confirmed using genomic PCR (Figure S2a), semi-quantitative RT-PCR (Fig. 3c), and qRT-PCR (Fig. 3d). The transcript levels of flowering-related genes were analysed by qRT-PCR (Fig. 3e). The expression levels of AtFT, AtLFY, and AtAP1 significantly increased in OE-MdNup62 lines compared with WT. This demonstrated that the overexpression of MdNup62 promoted flowering in Arabidopsis.

Fig. 3
figure 3

MdNUP62 promotes flowering in Arabidopsis. a Phenotype of the MdNUP62-overexpression Arabidopsis line for flowering time. Bar = 2 cm. b Statistical analysis of rosette leaves of Arabidopsis thaliana during bolting. Asterisks denote significant differences as determined by a t-test (*P < 0.05).The same below. c Semi-quantitative RT-PCR analysis of MdNup62 expression in Arabidopsis samples. d qRT-PCR analysis of MdNup62 expression in Arabidopsis samples. e Relative expression levels of flowering genes (AtFT, AtLFY, AtSOC1, and AtAP1) in WT and MdNup62-overexpression lines

Full size image

Overexpression of MdNup62 reduces high-temperature resistance

Because MdNup62 was induced by high temperature, we investigated the high-temperature resistance function of MdNup62. OE-MdNup62 Arabidopsis lines were subjected to a high-temperature (45 °C) treatment (Fig. 4a). Additionally, the survival rate of transgenic Arabidopsis was significantly lower than that of WT (Fig. 4b). We also performed a qRT-PCR analysis of A. thaliana HSPs (AtHSP101, AtHSP22-ER, AtHSP21.0, and AtHSP70T-2) (Fig. 4c). Their expression levels in transgenic Arabidopsis were reduced under high-temperature conditions. Consistently, after the heat treatment, the ROS accumulation in leaves was clear greater in OE-MdNup62 lines compared with WT (Fig. 5a). In addition, the malondialdehyde and H2O2 levels were significantly greater than in WT (Fig. 5b, c). Moreover, the superoxide dismutase, peroxidase, and catalase activities were lower in OE-MdNup62 lines than in WT (Fig. 5d–f). High-temperature resistance assays were carried out in transgenic tomato plants (Fig. 6a). As in transgenic A. thaliana, the survival rate of transgenic tomato was significantly reduced compared with WT (Fig. 6b). The presence of the transgene in OE-MdNup62 lines was confirmed by genomic PCR, and qRT-PCR (Fig. 6c,d). The expression levels of HSPs (HSP101, HSP22-ER, HSP21.0, and HSP70T-2) in transgenic tomato were significantly reduced under high-temperature conditions compared with under normal growth conditions (Fig. 6e). These results indicate that MdNup62 reduces plant high-temperature resistance.

Fig. 4
figure 4

MdNup62 reduced high-temperature resistance in Arabidopsis. a Phenotype of the MdNup62-overexpression Arabidopsis line for high-temperature resistance. Bar = 1 cm. b Survival rates of WT and MdNup62-overexpression Arabidopsis lines after the high-temperature treatment. c Relative expression levels of high-temperature resistance-related genes (AtHSP101, AtHSP22.0-ER, AtHSP21, and AtHSP70T-2) in WT and MdNup62-overexpression lines at the normal (22 °C) temperature and 1 h after exposure to the high-temperature (37 °C) treatment

Full size image
Fig. 5
figure 5

Changes in the level of accumulated ROS and activities of ROS-scavenging enzymes in OE-MdNup62 and WT Arabidopsis leaves under heat-stress conditions. a In situ accumulations of superoxide radicals (O.2−) before (upper panels) and after (lower panels) heat treatment as revealed by nitro blue tetrazolium staining. Changes in the level of accumulated ROS b and c Quantitative measurement of H2O2c– and malondialdehyde concentrations in Arabidopsis leaves treated with and without the high temperature. d–f Activities of superoxide dismutase (SOD), peroxidase (POD), and catalase (CAT) at 6 h after the heat treatment

Full size image
Fig. 6
figure 6

MdNup62 reduced high-temperature resistance in tomato. a Phenotype of the MdNup62-overexpression tomato line for high-temperature resistance. Bar = 5 cm. b Survival rates of WT and MdNup62-overexpression tomato lines after the high-temperature treatment. c qRT-PCR analysis of MdNup62 expression levels in tomato samples. d Genomic PCR analysis of MdNup62 transgenic tomato lines. e Relative expression levels of high-temperature resistance-related genes (SlHSP101, SlHSP22.0-ER, SlHSP21, and SlHSP70T-2) in WT and MdNup62-overexpression lines at the normal temperature (22 °C) and 1 h after exposure to the high-temperature (45 °C) treatment

Full size image

MdNup62-interacting protein screening

To further reveal the function of MdNup62, we conducted a Y2H sieve library experiment using a MdNup62 truncated body (MdNup62508–613-pGBKT7) that is not self-activated. We identified 62 putative MdNup62-interacting proteins (Table S3). Some transcription factors were identified, such as HSFs (MdHSFA1d, MdHSFA1e, MdHSFA9, MdHSF30, MdHSF1, and MdHSF8), as well as MdMYB21, MdMYC2, MdGATA11, and MdBAK1. In addition, some enzymes and other functional genes were found. Because transcription factors that have transcriptional regulatory functions must be transported into the nucleus, and because MdNup62 has regulatory effects on the transport of the proteins, we hypothesized that MdNup62 interacts with these MdHSFs and controls their transport.

MdNup62 interacts with MdHSFs

We cloned parts of the MdHSFs (MdHSFA1a/b/d/e and MdHSFA9a/b) independently into the pGADT7 vector and then cotransformed each with MdNUP62508–613-pGBKT7. MdNup62 interacted with these MdHSFs (Fig. 7a). Additionally, we used MdHSFA9b in pull-down assays. The recombinant MdNup62-HIS fusion protein was purified with MdHSFA9b-GST, but not with GST alone (Fig. 7b). The split-LUC complementation assay revealed that the co-expression of MdNup62-NLUC with MdHSFA1d-CLUC or MdHSFA9b-CLUC resulted in a higher LUC activity than the other combinations (Fig. 7c–e). These results confirmed the interaction between MdNup62 and both MdHSFA1D and MdHSFA9b.

Fig. 7
figure 7

MdNup62 interacts with MdHSFs. a Interactions between MdNup62508−613 and MdHSFs (MdHSFA9a/b and MdHSFA1a/b/d/e) in Y2H assays. The MdNup62508−613 truncated sequence was cloned into pGBKT7, whereas MdHSFAs (MdHSFA9a/b and MdHSFA1Da/b/d/e) were cloned independently into the pGADT7 vector. Empty pGADT7 plus MdNup62.508−613-pGBKT7 was used as the control. b Interactions between MdNup62 and MdHSFA9b in the pull-down assay. Western blotting with a GST antibody revealed that MdNup62-HIS was pulled down by MdHSFA9b-GST. c–e Interactions between MdNup62 and both MdHSFA9b and MdHSFA1d in a luciferase (LUC) complementation experiment. Empty NLUC and empty CLUC, MdNup62-NLUC plus empty CLUC, empty NLUC plus MdHSFA9b, and MdHSFA1d-CLUC were used as controls

Full size image

Feature, expression, and subcellular localization analyses of MdHSFA9b and MdHSFA1d

Phylogenetic tree analysis showed that Apple and Arabidopsis HSFs were divided into four groups (I, II, III, IV), with MdHSFA1a/b/d/e in groupII, and MdHSFA9a/b in groupI (Fig. 8a). We also examined the expression patterns of MdHSFs in different tissues of several apple varieties (Fig. 8b; Table S2). And the expression levels of MdHSFA1a/b/d/e.

Fig. 8
figure 8

Subcellular localization and expression analyses of MdHSFA9b and MdHSFA1d. a Phylogenetic analysis of HSFs in Malus and Arabidopsis. b Tissue specific expression patterns of apple MdHSFs by RNA-seq. The full names of the different abbreviations are as follows, ‘Nagafu No.2’ long branches flower buds (FLB), ‘Nagafu No.2’ short branches flower buds (FSB), ‘Nagafu No.2’ axillary buds (FYB), ‘Qinguan’ axillary buds (QYB), ‘Nagafu No.2’ fruit (FR), ‘Yanfu No.3’ stem tip (YF3J), ‘Yanfu No.6’ stem tip (YF6J), and ‘M9-T337’ root (T337R). c Subcellular localizations of MdHSFA9b and MdHSFA1d. The upper panel shows 35S::EGFP, the middle panel shows 35S::MdHSFA9b-EGFP, and the lower panel shows 35S::MdHSFA1d-EGFP. d Analyses of MdHSFA9b and MdHSFA1d expression levels in diverse apple ‘Nagafu No. 2’ tissues and in different flower bud developmental stages of ‘Nagafu No. 2’

Full size image

and MdHSFA9a/b showed significantly higher in buds.

The subcellular localizations of MdHSFA9b and MdHSFA1d were studied by independently introducing 35S::MdHSFA9b-EGFP and 35S::MdHSFA1d-EGFP, respectively, into tobacco leaves (Fig. 8c). Tobacco leaves transformed with the empty vector 35S::EGFP served as controls. In the tobacco leaves expressing 35S::MdHSFA9b-EGFP and 35S::MdHSFA1d-EGFP, the GFP signals were observed in both the nucleus and cytoplasm, while the GFP signal was detected throughout the control tobacco leaf cells, indicating that MdHSFA9b and MdHSFA1d localized to both the nucleus and cytoplasm.

A tissue-specific expression analysis revealed that MdHSFA1d was expressed highest in flower buds and stems. The highest expression level of MdHSFA9b was in stems, but the expression levels in the other tissues were also high. Subsequently, the expression levels of MdHSFA9b and MdHSFA1d remained high during the flower bud developmental stages, while the highest was at 70 days after flowering (Fig. 8d). These results indicated that MdHSFA9b and MdHSFA1d maintained high expression levels during flower bud induction, suggesting that they may be involved in the bud differentiation of apple.

Overexpression of MdHSFA9b and MdHSFA1d promotes flowering

To verify the flowering phenotype of HSFs, we performed Agrobacterium-mediated genetic transformations of MdHSFA9b and MdHSFA1d into A. thaliana. Like OE-MdNup62, OE-MdHSFA9b and OE-MdHSFA1d lines flowered significantly earlier than WT (Figs. 9a and S3a). Additionally, they also had significantly fewer rosette leaves than WT during bolting (Figs. 9b and S3b). We also performed genomic PCR (Figure S2b, c), semi-quantitative RT-PCR (Figs. 9c and S3c), and qRT-PCR (Figs. 9d and S3d) to confirm the presence of the transgene in the OE-MdHSFA9b and OE-MdHSFA1d lines. The transcript levels of AtFT, AtLFY, and AtSOC1 were significantly increased in OE-MdHSFA9b and OE-MdHSFA1d lines compared with WT (Figs. 9e and S3e).

Fig. 9
figure 9

MdHSFA9b promotes flowering in Arabidopsis. a Phenotype of the MdHSFA9b-overexpression Arabidopsis line for flowering time. Bar = 2 cm. b Statistical analysis of rosette leaves of Arabidopsis thaliana during bolting. c Semi-quantitative RT-PCR analysis of MdHSFA9b expression in Arabidopsis samples. (d) qRT-PCR analysis of MdHSFA9b expression in Arabidopsis samples. e Relative expression levels of flowering genes (AtFT, AtLFY, AtSOC1, and AtAP1) in WT and MdHSFA9b-overexpression lines

Full size image

Overexpression of MdHSFA9b and MdHSFA1d enhances high-temperature resistance

To study the high-temperature resistance phenotypes of MdHSFA9b and MdHSFA1d, we also exposed OE-MdHSFA9b and OE-MdHSFA1d transgenic plants, respectively, to high-temperature (45 °C) conditions (Figs. 10a and S4a). The survival rates of OE-MdHSFA9b and OE-MdHSFA1d lines were significantly greater than that of WT (Figs. 10b and S4b). Consistently, the ROS accumulation in leaves decreased in transgenic plants after the high-temperature treatment (Figure S5a, b). We also performed a qRT-PCR analysis of A. thaliana HSPs (AtHSP101, AtHSP22-ER, AtHSP21.0, and AtHSP70T-2) (Figs. 10c and S4c), and their expression levels in transgenic A. thaliana increased under high-temperature conditions compared with under normal growth conditions. These results indicated that MdHSFA9b and MdHSFA1d enhance plant high-temperature resistance.

Fig. 10
figure 10

MdHSFA9b enhanced high-temperature resistance in Arabidopsis. a Phenotype of the MdHSFA9b-overexpression Arabidopsis line for high-temperature resistance. Bar = 2 cm. b Survival rates of WT and MdHSFA9b-overexpression Arabidopsis lines after the high-temperature treatment. c Relative expression levels of high-temperature resistance-related genes (AtHSP101, AtHSP22.0-ER, AtHSP21, and AtHSP70T-2) in WT and MdHSFA9b-overexpression lines at the normal temperature (22 °C) and 1 h after exposure to the high-temperature (37 °C) treatment

Full size image

Discussion

Plant flowering has always been an important topic in crop and horticultural sciences, and issues with apple flowering have long hindered the development of the apple industry in China [1, 2]. The Nups control protein transport between the nucleus and cytoplasm, and they participate in a variety of biological processes, including flowering [19, 20]. In A. thaliana, Nup96 promotes the stability of HOS1, and HOS1 conjugates and degrades CO, then promotes FLC expression, leading to delayed flowering. In addition, HOS1 increases the stability of Nup96 and thus maintains this regulatory pathway to control the flowering time [23, 26]. Mutations in Nup54, Nup58, Nup62, Nup136, and Nup160 have resulted in a prominent earlier flowering phenotype compared with WT [16, 25]. In the present study, MdNup62 maintained a high expression level during flower development. To verify the flowering function of MdNup62, we determined the flowering phenotypes of OE-MdNup62 A. thaliana lines. Interestingly, the phenotypes of the overexpression lines were consistent with Arabidopsis deletion mutants and showed obvious early flowering. Previous studies found that both Nup62 deletion mutants and overexpression strains of Arabidopsis have increased the sensitivities to auxin, indicating that the overexpression does not result in a functional gain, but rather a functional loss, like the mutant [51]. Therefore, the overexpression of MdNup62 in this study may also result in a functional loss. However, MdNup62 is involved in the flowering pathway.

With global warming, extreme high-temperatures will occur more frequently, which will seriously affect the normal growth and development of plants [5, 6]. And Nups are involved in temperature-stress responses. HOS1 and Nup160 were reported to be involved in cold resistance [27, 28]. Nup85 and Nup133 control mRNA output only under warm conditions and are more sensitive to transcription factor localization at warm temperatures [20]. In this study, MdNup62 responded to high-temperature stress in apple. However, OE-MdNup62 lines had reduced high-temperature resistance in both Arabidopsis and tomato. By analysing the relative expression levels of HSPs (HSP101, HSP22-ER, HSP21.0, and HSP70T-2) in transgenic plants, we found no obvious correlations between OE-MdNup62 lines and WT at a normal growth temperature, but OE-MdNup62 lines had significantly lower HSP expression levels than WT under high-temperature conditions.

In plants, Nup-interacting proteins have been studied [17, 18, 26], and some potential Nup85-interacting proteins have been identified by immunoprecipitation and subsequent mass spectrometry in Arabidopsis, such as the Nup107–160 subcomplex (Nup160, Nup133, Nup43, Nup96, Nup107, Seh1, and Sec13), several mediator subunits (MED16, MED14, and MED18), HOS1, and Sec13A. The interactions between Nup85 and three proteins, HOS1, Sec13A, and MED18, have been confirmed. Additionally, a direct interaction between Nup96 and HOS1 in Arabidopsis has also been reported [26]. In our previous study, the interaction between MdNup54 and MdNup62 was confirmed in apple [17]. However, there are no reports of direct interactions between transcription factors and Nups in plants. We previously identified an interaction between apple MdNup54 and MdKNAT4/6 using a yeast double-hybridization test, but further verification is needed [17]. In this study, we verified direct interactions between MdNup62 and MdHSFs, indicating that the Nups may directly recognize related transcription factors and thus regulate their transport. This provides a new direction of study for Nups.

Because of the early flowering of OE-MdNup62 Arabidopsis lines, MdHSFs that interact with MdNup62 may be also involved in the flowering pathway. Consistent with this conjecture, some HSFs are associated with flowering [49, 50]. HSFA1E and HSFA4C directly target and positively regulate the flowering gene SOC1 in lettuce [49]. Arabidopsis HSFA2 directly targets and promotes the expression of REF6, and the REF6–HSFA2 loop directly targets and activates HTT5, which coordinates early flowering [50]. In this study, we found that MdHSFA9b and MdHSFA1d maintained high expression levels during flower bud induction. Additionally, OE-MdHSFA9b and OE-MdHSFA1d Arabidopsis lines flower significantly earlier than WT. This suggests that MdHSFA9b and MdHSFA1d promote plant flowering. MdNup62, MdHSFA9b, and MdHSFA1d share the same flowering phenotype, possibly because the overexpression of MdNup62 fosters HSF accumulation in the nucleus, promoting the expression of downstream flowering-related genes and advancing flowering.

HSFs play important roles in regulating plant resistance to high temperatures. HSFA1 positively regulates the heat tolerance of tomato, the expression of HSFA2 is dependent on HsfA1, and the thermotolerance of the posttranscriptional silencing of the HsfA1 gene in protoplasts can be restored by plasmid-borne HsfA2 [52]. HSFA1d and HSFA1e activate HsfA2 transcription, and a double knockout of HSFA1d and HSFA1e impairs tolerance to heat-shock stress [43]. In Medicago truncatula, HSFA9 plays important roles in thermotolerance [53]. In the current study, we obtained similar results for MdHSFA9b and MdHSFA1d. The expression levels of HSPs in the two overexpression Arabidopsis lines were significantly greater than in WT, and both lines had enhanced high-temperature resistance levels. Like the flowering and auxin phenotypes [51], the opposite phenotypes between OE-MdNup62 and OE-MdHSFA9b, OE-MdHSFA1d indicates that the overexpression of MdNup62 may also result in a lack of function under heat-stress conditions. Similar to the results of this study, Zhang et al. (2020) found that nup85 and nup133 increase the ubiquitous protoplast (nucleus and cytosol) signals of IAA17 and PIF4 at 28 °C compared with at 22 °C. Furthermore, the nup96 and hos1 mutants show significant increases in the ubiquitous localizations of IAA17 and PIF4 signals at 28 °C (72% and 66%, respectively) compared with 22 °C (40% and 49%, respectively)[20]. Thus, the nuclear accumulations of the IAA17and PIF4 proteins in nup85, nup96, nup133, and hos1 are reduced compared with WT, and the defects are more severe at 28 °C. Therefore, we hypothesized that the transport of MdHSFA9b, MdHSFA1d, and other MdHSFs is inhibited in OE-MdNup62 lines at high temperatures, resulting in the inhibition of the transcription of downstream HSPs, which further reduces high-temperature resistance.

On the basis of these findings, we constructed a hypothetical model of MdNup62-related pathways involved in high-temperature resistance (Fig. 11). At normal temperature, apple MdHSFs were not induced, and not much transported into nucleus that cannot lead to up-regulate expression of MdHSPs in WT and OE-MdNup62. However, at high temperature, apple MdHSFs were significantly induced, and then transported into the nucleus through NPC channels to promote the expression of MdHSPs in WT, in which enhanced high-temperature resistance. But for OE-MdNup62 lines, the structure of the apple NPC changed, and blocked the transport of high temperature induced MdHSFs into the nucleus that cannot induce much MdHSPs expression causing heat injuring (Fig. 11). Additionally, OE-MdNup62, OE-MdHSFA9b and OE-MdHSFA1d lines showed significant early flowering phenotype compared with WT (Fig. 3, 9; Figure S3).

Fig. 11
figure 11

Model of MdNup62 interactions with MdHSFs involved in heat-stress tolerance in apple

Full size image

In conclusion, temperature is an important factor affecting flowering. With global warming, apple flowering will occur earlier, increasing the risk of chilling-related injury. Moreover, extreme hot weather is also occurring frequently. Both climatic conditions seriously affect the development of the apple industry. MdNup62 interacts with MdHSFs to regulate flowering and heat-resistance pathways in plants. Thus, both MdNup62 and the MdHSFs regulate flowering and respond to temperature changes. This research provides a theoretical reference for managing the impact of global warming on the apple industry.

Materials and methods

Plant materials and growth conditions

The plant materials were 6-year-old apple trees (‘Fuji’ /T337/Malus robusta Rehd.) growing in the experimental orchard of the Horticulture College of Northwest A & F University (108°04′ E, 34°16′ N). We collected new shoots (2–3 mm in diameter) near the tips, fully expanded leaves near buds, flower buds, blooming flowers, and young fruit, which were immediately frozen in liquid nitrogen and stored at − 80 °C for later use.

The ‘Fuji’ plants were grown on MS medium containing 0.1 mg·L−1 indolebutyric acid and 0.6 mg·L−1 6-benzylaminopurine under long-day conditions (16 h-light/8 h-dark) at 24 °C and were subcultured every 45 days. Arabidopsis plants(‘Columbia’) were grown under long-day conditions (16 h-light/8 h-dark) at 22 °C. Tomato plants (‘Ailsa Craig’) were grown under long-day conditions (16 h-light/8 h-dark) at 25 °C. And the arabidopsis and tomato seeds were previously preserved in our laboratory.

Heat map, protein alignment, and phylogenetic analysis

Based on RNA-seq data of our laboratory, the heat map of apple different tissues was constructed using MeV (Multiple Experiment Viewer) software. A protein sequence alignment of Nup62 from six Rosaceae plants was performed using DNAMAN software. The Nup62 protein sequences were obtained from the GDR database (https://www.rosaceae.org/). The phylogenetic tree was constructed using MEGA-X software.

RNA extraction and qRT-PCR analysis

Total RNA was extracted from apple trees, Arabidopsis seedlings, tomato seedlings, and apple seedlings using an RNA Plant Plus Reagent Kit (TIANGEN, Beijing, China). The RNA was used as the template to synthesize cDNA with a PrimeScript RT Reagent Kit (Takara, Shiga, Japan). The qRT-PCR analysis was conducted on a StepOnePlus Real-Time PCR System (Thermo Fisher Scientific, USA). The reaction solution contained 10 μL SYBR Green I Master Mix (CWBIO, Beijing, China), 0.5 μmol·L−1 primers (SANGON BIOTECH, Shanghai, China), and 1 μL each template in a total volume of 20 μL. The PCR program was as follows: 95 °C for 3 min; 40 cycles of 94 °C for 15 s, 60 °C for 20 s, and 72 °C for 15 s. All the samples were analysed with three biological replicates, each comprising three technical replicates. Relative gene expression levels were calculated in accordance with the 2−ΔΔCt method [54]. The primers used for qRT‐PCR (Table S4) were synthesized by the Sangon Biotechnology Co. Ltd. (Shanghai, China).

Subcellular localization

The open reading frames (ORFs) of the MdNup62, MdHSFA1d, and MdHSFA9b genes were inserted independently into the pCAMBIA2300-EGFP vector to generate the 35S::MdNup62-EGFP, 35S::MdHSFA1d-EGFP, and 35S::MdHSFA9b-EGFP recombinant plasmids, respectively. These recombinant plasmids were inserted independently into Agrobacterium tumefaciens strain GV3101 cells. The GV3101 cells containing these recombinant plasmids were then infiltrated into tobacco leaves. GV3101 cells containing the pCAMBIA2300-EGFP vector (35S::EGFP) served as the control. After an additional 3 days of growth in the dark, green fluorescent protein (GFP) signals in transformed tobacco leaves were detected using a Leica TCS SP8 SR Laser Scanning Confocal Microscope (Leica, Germany).The primers used are listed in Table S5.

Genetic transformation

The genetic transformations were performed in accordance with published methods for Arabidopsis [55] and tomato (‘Ailsa Craig’) [56] plants. The transgenic Arabidopsis and tomato lines were selected on MS plates supplemented with 50 mg·L−1 and 100 mg·L−1 kanamycin, respectively.

Yeast two-hybrid (Y2H) assay

The MdNup62508–613 truncated sequence was cloned into the pGBKT7 vector to generate the MdNup62508–613-pGBKT7 recombinant plasmid. The MdHSFAs’ ORFs were inserted individually into the pGADT7 vector to generate the MdHSFAs-pGADT7 recombinant plasmids. The recombinant plasmids were inserted into Gold Yeast Two-Hybrid cells, which were then grown on a selective medium. The primers used are listed in Table S5.

Split luciferase (LUC) complementation

The full-length MdHSFA1d and MdHSFA9b coding sequences were cloned independently into the CLUC vector, while MdNup62 was cloned into the NLUC vector. The split-LUC complementation assay was performed with tobacco leaves. The reconstituted LUC activity was detected in the dark using a Princeton Lumazone Pylon 2048B cooling camera (Princeton, USA). The LUC activity was quantified using the Dual-Luciferase Reporter Assay System (Promega, USA). The primers used are listed in Table S5.

Pull-down assays

The ORFs of MdNup62 and MdHSFA9b were cloned into the pET-28a and pGEX-6p-1 vectors, respectively, and subsequently overexpressed independently in Escherichia coli BL21(DE3) (Transgene). The pull-down assays were conducted using the His-Tagged Protein Purification Kit (Clontech) in accordance with the manufacturer’s instructions. The primers used are listed in Table S5.

Heat-tolerance assays

The ‘Fuji’ plants at 30 days after propagation were used for the 45 °C heat treatment. We collected leaf samples before and at 1, 3, and 6 h after the treatment. The samples were immediately frozen in liquid nitrogen and stored at − 80 °C for later use.

Two-week-old transgenic Arabidopsis and 3-week-old transgenic tomato were used for the heat treatment in an artificial climate chamber. OE-MdNup62 A. thaliana lines were subjected to 45 °C for 12 h, and OE-MdHSFA9b and OE-MdHSFA1d A. thaliana lines were subjected to 45 °C for 16 h. OE-MdNup62 tomato lines were subjected to 45 °C for 14 h.

Evaluation of stress tolerance

The superoxide dismutase, peroxidase, and catalase activities and the malondialdehyde and H2O2 levels were detected using the corresponding Suzhou Comin Biotechnology test kits (Suzhou Comin Biotechnology Co., Ltd, Suzhou, China). The presence of O2− in leaf samples was determined by staining with nitro blue tetrazolium.

Statistical analyses

Statistical analyses were performed using SPSS software. Data are reported as means ± SDs. Asterisks (*) indicate significant differences between treatments as assessed by Student’s t-test at P < 0.05 (*) and P < 0.01 (**). Different lowercase letters above the bars indicate significant differences (P < 0.05, Tukey’s test).