Interaction among homeodomain transcription factors mediates ethylene biosynthesis during pear fruit ripening

Abstract Fruit ripening is manipulated by the plant phytohormone ethylene in climacteric fruits. While the transcription factors (TFs) involved in ethylene biosynthesis and fruit ripening have been extensively studied in tomato, their identification in pear remains limited. In this study, we identified and characterized a HOMEODOMAIN TF, PbHB.G7.2, through transcriptome analysis. PbHB.G7.2 could directly bind to the promoter of the ethylene biosynthetic gene, 1-aminocyclopropane-1-carboxylic acid synthase (PbACS1b), thereby enhancing its activity and resulting in increased ethylene production during pear fruit ripening. Yeast-two-hybrid screening revealed that PbHB.G7.2 interacted with PbHB.G1 and PbHB.G2.1. Notably, these interactions disrupted the transcriptional activation of PbHB.G7.2. Interestingly, PbHB.G1 and PbHB.G2.1 also bind to the PbACS1b promoter, albeit different regions from those bound by PbHB.G7.2. Moreover, the regions of PbHB.G1 and PbHB.G2.1 involved in their interaction with PbHB.G7.2 differ from the regions responsible for binding to the PbACS1b promoter. Nonetheless, these interactions also disrupt the transcriptional activation of PbHB.G1 and PbHB.G2.1. These findings offer a new mechanism of ethylene biosynthesis during climacteric fruit ripening.


Introduction
Fruit ripening involves a multitude of metabolic changes that impact various sensory traits such as color, f lavor, texture, and aroma.These modifications are not only advantageous for seed dispersal, but also provide essential dietary components, including nutrients and fiber, for both human and animal diets [1][2][3].Therefore, elucidating the regulatory mechanisms underlying fruit ripening holds significant importance in enhancing breeding strategies approaches and optimizing cultivation practices, ultimately resulting in improvements in fruit quality.
In climacteric fruits, ripening is characterized by a selfamplifying increase in respiration and ethylene production [4,5], which is preceded by the biosynthesis of S-adenosyl-Lmethionine (SAM) and its conversion to 1-aminocyclopropane-1-carboxylate (ACC) by ACC synthase (ACS), and subsequent conversion of ACC to ethylene by 1-aminocyclopropane-1carboxylate oxidase (ACO).It is reported that these gene activities are directly regulated by multiple transcription factors (TFs), including Basic helix-loop-helix (bHLH), Apetala2/Ethylene responsive factor (AP2/ERF), Auxin response factor (ARF), MADSbox, NAM/ATAF1/2/CUC2 (NAC), and Homeobox (HB) [6][7][8][9][10].In these TFs, in apple, MdERF3 induces MdACS1 transcription to increase ethylene production, and the transcriptional activation of MdERF3 is repressed by the interaction between MdERF3 and MdERF2 [11].MdERF2 suppresses the transcription of MdACS1 to impede ethylene biosynthesis, and the transcriptional repression of MdERF2 is disrupted by the interaction between MdERF2 and MdMYC2 (a bHLH TF) [9].In banana MaERF11 suppresses the transcription of MaACO1 to negatively affect ethylene biosynthesis during fruit ripening, and this repression is strengthened by the interaction between MaERF11 and MaHDA1 (a histone deacetylase) [12].Theoretically, TFs involved in the regulation of ethylene biosynthesis may also have interacting partners.
HB TFs constitute a diverse superfamily encompassing various families such as SAWADEE, DDT, KNOX, BELL, WOX, PHD, and HD-ZIP [13,14].Multiple studies have demonstrated the involvement of HB TFs in diverse biological processes including leaf development, inf lorescence architecture, embryogenesis, asexual reproduction, and fruit replum formation [15][16][17][18].Notably, several HB TFs are also implicated in phytohormone biosynthesis and signal transduction.For instance, in potato, POTH1 decreases gibberellin (GA) accumulation in leaves by suppressing the expression of GA20-oxidase1 (a key gene for gibberellin biosynthesis) [19].In rice, OsIPT2 and OsIPT3 decrease GA biosynthesis by suppressing the expression of GA20-oxidase genes, while simultaneously promoting CK biosynthesis by inducing the expression of cytokinin (CK) biosynthesis genes [20].In Arabidopsis, ATHB5 inhibits seed germination and seedling growth in responsive to abscisic acid [21], while ATHB52 negatively modulates primary root elongation by responding to ethylene and impairing auxin distribution and gravitropism [22].In peach, PpHB.G7 promotes ethylene biosynthesis by enhancing the expression of both PpACS1 and PpACO1 [23], whereas tomato LeHB-1 plays a positive role in fruit ripening [24].These instances indicate that HB TFs may be involved in regulating ethylene biosynthesis and consequently inf luence fruit ripening processes.
Pear is a widely consumed fruit globally and belongs to the category of climacteric fruit, which exhibit a sharp increase in ethylene production during ripening.To gain insights into the genetic regulation of fruit ripening, in this study, three HB genes, namely PbHB.G7.2, PbHB.G1, and PbHB.G2.1, were identified, and their molecular regulation on ethylene biosynthesis was elucidated.These results uncover a new mechanism underlying ethylene biosynthesis in pear fruit.

PbACS1b, PbACO1, and PbHB.G7.2 are associated with pear fruit ripening
To identify the ACS, ACO, and HB genes involved in pear fruit ripening, the transcriptome data from cvs.Housui, Cuiguan, and Xueqing fruits, as reported in a previous study [25], were used to perform differential expression analysis of these genes between enlarging and ripening stages (Table S1, see online supplementary material).A total of four ACS, four ACO, and 78 HB genes were detected in the fruit samples from the three pear cultivars (Table S2, see online supplementary material).The detected ACS and ACO genes were designated based on their orthologs in phylogenetic trees (Fig. S1, see online supplementary material), while the HB genes were designated according to a prior study conducted on peach fruit [23].Among the detected 85 genes, only PbACS1b, PbACO1, PbHB.E10.2, and PbHB.G7.2 exhibited higher expression levels in ripening fruit compared to enlarging fruit across all three cultivars (Fig. 1a).However, quantitative real-time PCR (qRT-PCR) analysis showed that PbHB.E10.2 did not display differentially expressed between cv.Housui fruits at 104 and 157 days after f lower blooming (DAFB) nor between cv.CG fruits at 96 and 136 DAFB (Fig. 1b).In contrast, PbACS1b, PbACO1, and PbHB.G7.2 were more highly expressed in ripening fruit than in enlarging fruit across all three cultivars (Fig. 1b).These results indicate a potential involvement of PbACS1b, PbACO1, and PbHB.G7.2 in pear fruit ripening.

PbHB.G7.2 mediates ethylene biosynthesis in pear fruit callus by inducing the PbACS1b expression
To test the potential role of PbHB.G7.2 in mediating the expression of PbACS1b and PbACO1, an over-expression experiment was conducted using pear fruit callus.The result showed that ethylene production was increased in the pear fruit calli overexpressing PbHB.G7.2, compared to the fruit calli expressing the empty vector pSAK277 (Fig. 2a).Moreover, PbHB.G7.2 and PbACS1b were more highly expressed in the fruit calli overexpressing PbHB.G7.2 than in the fruit calli expressing the empty vector pSAK277, while the expression level of PbACO1 remained relatively unchanged (Fig. 2b).These results indicate that the over-expression of PbHB.G7.2 promotes ethylene biosynthesis by inducing the expression of PbACS1b, while exerting minimal inf luence on the expression of PbACO1.
To test the ability of PbHB.G7.2 to enhance the activity of the PbACS1b promoter, 2000-bp sequences upstream of PbACS1b were inserted into the pGreenII 0800-LUC vector as a reporter construct (Fig. 2c).The report construct was then infiltrated into tobacco leaves along with the effector construct containing the overexpression vector of PbHB.G7.2.Subsequently, the dual-luciferase assay showed that the LUC activity driven by the PbACS1b promoter (Pro-PbACS1b) was significantly increased in the presence of the 35S::PbHB.G7.2 effector construct compared to the empty vector (Fig. 2d), while the LUC activity driven by the PbACO1 promoter was hardly changed (Fig. S2, see online supplementary material).This result suggests that PbHB.G7.2 can enhance the activity of the PbACS1b promoter in tobacco leaves.
Because the binding sites of HB TFs remain unknown, to determine the binding region of PbHB.G7.2 within the PbACS1b promoter, the promoter was divided into four fragments, P1 (from −2000 bp to −500 bp), P2 (from −2000 bp to −1000 bp), P3 (from −1500 bp to −1 bp), and P4 (from −1000 bp to −1 bp; Fig. 2d), and used as individual reporters in a dual-luciferase assay.The result showed that PbHB.G7.2 enhanced the LUC activities driven by either P3 or P4 fragment, while exhibiting no effect on the LUC activities driven by either P1 or P2 fragment (Fig. 2d).This result indicates that the PbHB.G7.2 binding region within the PbACS1b promoter is likely located within the upstream region spanning from −500 to −1 bp of the initiation codon of PbACS1b.Subsequently, the PbACS1b promoter was further divided into four additional fragments, P5 (from −2000 bp to −400 bp), P6 (from −2000 bp to −300 bp), P7 (from −2000 bp to −200 bp), and P8 (from −2000 bp to −100 bp; Fig. 2e), and used as individual reporters in a dual-luciferase assay.The result showed that PbHB.G7.2 enhanced the LUC activities driven by either P7 or P8 fragments, but while having no impact on the LUC activities driven by either P5 or P6 fragments (Fig. 2e).These results provide compelling evidence that the PbHB.G7.2 binding region within the PbACS1b promoter is located within the upstream region spanning from −300 bp to −200 bp of the initiation codon of PbACS1b.
To test the direct binding capability of PbHB.G7.2 to the PbACS1b promoter, the entire sequences (P) and fragment P9 (from −300 bp to −200 bp; Fig. 2f) of the PbACS1b promoter were individually inserted into the pAbAi vector.The yeast-one-hybrid (Y1H) assay showed that both the P and P9 fragments of the PbACS1b promoter exhibited binding affinity towards PbHB.G7.2 (Fig. 2f), suggesting a potential physical interaction between PbHB.G7.2 and the PbACS1b promoter.To validate the Y1H result, a 34-bp fragment within the P9 sequences was synthesized as a biotinylated probe, while recombinant PbHB.G7.2 was generated in Escherichia coli (Fig. S3, see online supplementary material).The electrophoretic mobility shift assay (EMSA) showed that the biotinylated probe derived from the PbACS1b promoter could be specifically bound by the recombinant PbHB.G7.2, and the strength of the binding signal corresponded inversely with the concentration of the non-labeled competitor probe (cold probe; Fig. 2g).These results definitively demonstrate the physical interaction between PbHB.G7.2 and the PbACS1b promoter.

Pear fruit ripening is regulated by ethylene
To test the role of ethylene in the regulation of pear fruit ripening, ethephon and an ethylene inhibitor, 1-methylcyclopropene (1-MCP), were used to treat on-tree fruits of the cv.CG (Fig. 3a).As a result, fruit ripening was delayed by 1-MCP treatment and was accelerated by ethephon treatment.Moreover, soluble solid content and ethylene production decreased in the fruit treated with 1-MCP and increased in the fruit treated with ethephon, compared to the control fruit (Fig. 3b and c).These results are consistent with the positive regulatory effect of ethylene on pear fruit ripening.In addition, the expression of PbACS1b and PbACO1 were down-regulated in the fruit treated with 1-MCP and upregulated in the fruit treated with ethephon, compared to controls (Fig. 2d).However, the expression of PbHB.G7.2 remained largely unchanged in response to any treatment or the control condition (Fig. 2d), indicating that ethylene stimulation did not induce the expression of PbHB.G7.2 in pear fruit.

PbHB.G7.2 mediates ethylene biosynthesis during pear fruit ripening
To clarify the function of PbHB.G7.2 in ethylene biosynthesis and fruit ripening, transient over-expression and silencing experiments were performed on cv.Sucui No.1 fruit.As a result, the pericarp of the fruit over-expressing PbHB.G7.2 presented a yellow color, in contrast to the fruit expressing the empty vector pSAK277 (Fig. 4a).Conversely, the pericarp of the fruit with silenced PbHB.G7.2 presented a green color compared to the fruit expressing the empty vectors pTRV1 and pTRV2 (TRV; Fig. 4b).This result suggests that PbHB.G7.2 positively contributes to pear fruit ripening.Moreover, ethylene production increased in the fruit over-expressing PbHB.G7.2, while it decreased in the fruit with silenced PbHB.G7.2, compared to the controls (Fig. 4c).The result suggests that PbHB.G7.2 has a positive effect on ethylene biosynthesis.In addition, the expression levels of PbACS1b and PbHB.G7.2 were up-regulated in the fruit over-expressing PbHB.G7.2, whereas they were down-regulated in the fruit with silenced PbHB.G7.2, in comparison to the control fruit (Fig. 4d).These results suggest that PbHB.G7.2 positively affects ethylene biosynthesis by inducing the expression of PbACS1b.
These results clearly indicate that fragment D of both PbHB.G1 and PbHB.G2.1 interacts with the combined fragments D and C of PbHB.G7.2.

Both PbHB.G1 and PbHB.G2.1 physically interact with the PbACS1b promoter
To identify the binding regions of PbHB.G1 and PbHB.G2.1 within the PbACS1b promoter, a dual-luciferase assay was performed using fragments P1, P2, P3, and P4 of the PbACS1b promoter in conjunction with PbHB.G1 and PbHB.G2.1 constructs.The result showed that PbHB.G1 enhanced the LUC activity driven by either P1 or P2 fragments (Fig. 6c), indicating that the binding region of PbHB.G1 within the PbACS1b promoter is located within the upstream region from −2000 bp to −1000 bp of the initiation codon of PbACS1b.PbHB.G2.1 enhanced the LUC activity driven by either P1 or P3 fragments (Fig. 6c), indicating that the binding region of PbHB.G2.1 within the PbACS1b promoter likely positioned at the boundary between fragments P2 and P3.
To refine the identification of interaction regions, the PbACS1b promoter was further divided into additional fragments P10 (from −1600 bp to −1 bp), P11 (from −1700 bp to −1 bp), P12 (from −1800 bp to −1 bp), P13 (from −1900 bp to −1 bp), P14 (from −1060 bp to −1 bp), and P15 (from −2000 bp to −940 bp), and used for dual-luciferase assay (Fig. 6d).The result showed that PbHB.G1 enhanced the LUC activity driven by either P12 or P13 fragments, indicating that the binding region of PbHB.G1 within the PbACS1b promoter is located within the upstream region from −1800 bp to −1700 bp of the initiation codon of PbACS1b.Moreover, PbHB.G2 enhanced the LUC activity driven by P14 and P15 fragments (Fig. 6d), indicating that the binding region of PbHB.G2.1-Pro-PbACS1b within the PbACS1b promoter is located within the upstream region from −1060 bp to −940 bp of the initiation codon of PbACS1b.
To test whether PbHB.G1 and PbHB.G2.1 directly bind to the PbACS1b promoter, fragments P16 (from −1800 bp to −1700 bp) and P17 (from −1060 bp to −940 bp; Fig. 6e) were employed in a Y1H assay.The result showed that PbHB.G1 exhibited binding affinity towards both the P16 and P sequences of the PbACS1b promoter, while PbHB.G2.1 displayed binding affinity towards both the P17 and P sequences of the PbACS1b promoter (Fig. 6e).This result suggests that PbHB.G1 and PbHB.G2.1 may physically bind to the PbACS1b promoter.
To confirm the Y1H result, 33-bp and 40-bp sequences from fragments P16 and P17, respectively, were synthesized as biotiny- lated probes.Additionally, PbHB.G1 and PbHB.G2.1 were recombined in E. coli (Fig. S3, see online supplementary material), and the recombinants were used in an EMSA.The results showed that the recombinants of PbHB.G1 and PbHB.G2.1 could effectively bind to the respective biotinylated probes derived from the PbACS1b promoter, and the binding signals were observed to be inversely correlated to the concentration of the cold probe (Fig. 6f).Consequently, these findings provide further evidence that both PbHB.G1 and PbHB.G2.1 physically bind to the PbACS1b promoter.

PbHB.G1 and PbHB.G2.1 were lesser expressed in ripening fruit than PbHB.G7.2
The presence of PbHB.G7.2, PbHB.G1, and PbHB.G2.1 was found to enhance the activity of the PbACS1b promoter, and their transcriptional activations were repressed by the interactions of PbHB.G7.2 with PbHB.G1 and PbHB.G2.1.It has confused the specific HB gene(s) responsible for playing a more important role in the PbACS1b transcription and ethylene biosynthesis.For this reason, expression levels of PbHB.G7.2, PbHB.G1, and PbHB.G2.1 were examined in enlarging and ripening fruits of cvs.Housui, Cuiguan, and Xueqing.The result showed that PbHB.G7.2 had higher levels of expression than both PbHB.G1 and PbHB.G2.1 in ripening fruits, but not in enlarging fruits, across all three pear cultivars (Fig. 7a).Simultaneously, ethylene production was significantly increased in ripening fruits compared to developing fruits in all three cultivars (Fig. 7b).Moreover, to test the prevalence of the expression profile of the three HB genes in ripening fruits across various cultivars, ripening fruits from 10 additional cultivars were collected for the qRT-PCR analysis.The result showed that PbHB.G7.2 had higher levels of expression than PbHB.G1 and PbHB.G2.1 in all 10 cultivars (Fig. 7c).These results suggest a more significant role of PbHB.G7.2 in the transcriptional regulation of PbACS1b compared to PbHB.G1 and PbHB.G2.1.In addition, to gain further insight into the involvement of PbACS1b, PbACO1, and PbHB.G7.2 in fruit ripening, enlarging fruits from 10 additional cultivars were analysed by qRT-PCR.The result showed that PbACS1b, PbACO1, and PbHB.G7.2 were more highly expressed in ripening fruits compared to enlarging fruits across all 10 cultivars (Fig. 7d; Fig. S5, see online

PbHB.G7.2 mediates ethylene biosynthesis in pear fruit by inducing the PbACS1b expression
The ripening mechanism in climacteric fruits has been widely studied in apple [9][10][11]26], banana [12,27,28], peach [23,29], and kiwifruit [30,31].In contrast, the understanding of genes involved in ethylene biosynthesis and signaling transduction in pear fruit is limited [32][33][34].To explore the ripening mechanism in pear fruit, we treated pear fruits (while still on the tree) with exogenous ethephon and 1-MCP and confirmed the positive mediation of ethylene on pear fruit ripening (Fig. 2a).This result is supported by previous reports that demonstrated a delay in pear fruit ripening with 1-MCP treatment [32].Coincidently, two ethylene biosynthetic genes, PbACS1b and PbACO1, are positively correlated with fruit ripening and ethylene production (Fig. 2d), suggesting the involvement of both PbACS1b and PbACO1 in ethylene biosynthesis during fruit ripening.Of these two genes, PbACO1, an ortholog of apple MdACO1 (Fig. S1b, see online supplementary material), is identical to the temporarily designated PbACO54 in a previous study [32].PbACS1b, an ortholog of peach PpACS1 (Fig. S1a), differs from the previously reported PuACS1a in Pyrus ussuriensis [33].The identification of these two genes lays the groundwork for future research on the transcriptional regulation of ethylene biosynthesis during pear fruit ripening.
In perennial fruit trees, ethylene biosynthesis genes are regulated by PpHB.G7 and PpERF.A16 in peach fruit [23,35]; MdERF3, MdARF5, and MdMYC2 in apple fruit [9][10][11]; and AdNAC2 and AdNAC6/7 in kiwifruit [30,36].However, only PbERF24 has been reported to regulate the expression of PbACO1 in pear fruit [32].In this study, we found that PbHB.G7.2 promotes ethylene production during pear fruit ripening by inducing the expression of PbACS1b (Figs 3 and 4).Interestingly, PbHB.G7.2 belongs to the HD-ZIP II family and is an ortholog of PpHB.G7, which is known to be involved in peach fruit ripening [23].This result indicates that the role of HB TFs in group G7 in ethylene biosynthesis during fruit ripening is similar across different fruit trees.The only distinction lies in the fact that PpHB.G7 can bind to both ACS and ACO promoters, whereas PbHB.G7.2 can only bind to the ACS promoter.
Transcription activation of PbHB.G7.2 is weakened by its interaction with PbHB.G1 and PbHB.G2.1 Increasing evidence has shown the inf luence of the interacting partner on the transcriptional activation or repression of the TFs involved in fruit ripening [10][11][12]28].The detected proteinprotein interactions can either enhance or weaken the transcriptional impact of TFs on the expression of downstream genes.Within the HB family of TFs, interactions have been observed between KNOX proteins and BELL proteins [37,38].In this study, we found that PbHB.G1 and PbHB.G2.1 interact with PbHB.G7.2 (Fig. 5).These two HB TFs also belong to the HD-ZIP II family [23] and, like PbHB.G7.2, directly interact with the PbACS1b promoter to enhance its activity (Fig. 7).Interestingly, the transcriptional activation of these three HB TFs was disrupted by the interactions of PbHB.G7.2 with PbHB.G1 and PbHB.G2.1 (Fig. 6b).This is similar to a previous report where protein-protein interactions between TFs weakened their transcriptional effect on the expression of downstream genes.For example, the interaction between MdERF3 and MdERF2 repressed the transcriptional activation of MdERF3 [11], while the interaction of MdERF2 with MdMYC2 repressed the transcriptional activation of MdERF2 [9].Aux/IAA TFs can also repress the transcriptional activation of ARF TFs through protein-protein interactions involving their C-terminal domains [39,40].
In addition to the detected ERF and HB TFs, protein-protein interactions between two members were also reported in the NAC [41,42], MADS-box [43,44], bHLH [45], and WRKY [46] families.However, little is known about the temporal features of these TFs when expressed in tissue.The transcriptional activations of PbHB.G7.2, PbHB.G1, and PbHB.G2.1 on the PbACS1b promoter contradict their protein-protein interactions in the present study, making it uncertain which HB TF plays a crucial role in ethylene biosynthesis and fruit ripening.To address this, the expression levels of PbHB.G7.2, PbHB.G1, and PbHB.G2.1 were analysed in pear fruits.We found that PbHB.G1 and PbHB.G2.1 were expressed at a lower level in ripening fruits compared to PbHB.G7.2 (Fig. 7).This result suggests that PbHB.G7.2 may have a more important role in ethylene biosynthesis than PbHB.G1 and PbHB.G2.1.Moreover, the lesser expression of PbHB.G1 and PbHB.G2.1 in ripening fruit could weaken the interactions of PbHB.G7.2 with PbHB.G1 and PbHB.G2.1, thereby enhancing the transcriptional activation of PbHB.G7.2 on the expression of PbACS1b and increasing ethylene production during pear fruit ripening.

Conclusion
Our result indicates that the identified PbHB.G7.2 positively affects ethylene biosynthesis by inducing the expression of PbACS1b, independent of ethylene signaling.However, in enlarging fruits, the low expression of PbHB.G7.2 results in its interaction solely with PbHB.G1 and PbHB.G2.1, leading to a decrease in ethylene production.In ripening fruit, the high expression of PbHB.G7.2 allows for its interaction with PbHB.G1 and PbHB.G2.1, enabling binding to the PbACS1b promoter and subsequently increasing ethylene production.Both PbHB.G1 and PbHB.G2.1 are unresponsive to ethylene signal (Fig. S6, see online supplementary material) yet exhibit lower expression levels compared to PbHB.G7.2 in ripening fruit.These findings differ from previous reports on the molecular regulation of ethylene biosynthesis during fruit ripening.

Exogenous treatments of ethephon and 1-methylcyclopropene
Exogenous treatments were performed in cv.Cuiguan fruit at approximately 80 DAFB.At least 40 fruits on the tree were soaked in 300 μL l −1 ethephon or 1.5 μL l −1 1-methylcyclopropene (1-MCP) solutions for 5 min.To block light and limit gas exchange, the treated fruits were bagged with black airtight papers that were removed after 24 h.The fruits treated with ethephon were collected at the ripening stage, while the fruits treated with 1-MCP were collected on the same day as the control fruits at the ripening stage.At least six fruits were included in each of six biological replicates.The sampling regime was identical to a previous report [32].

Measurements of soluble solids and ethylene production
Soluble solids were measured as described in a previous study [32].Ethylene production was measured using a gas chromatographic system (GC2010, Shimadzu, Japan) with a 1000-μL Rheodyne injector (Gaoge, Shanghai, China) according to a previous report [47].In brief, the fruits in each replicate were weighed and sealed in a 3-L airtight chamber.A 0.5-mL gas sample was collected from the chamber by a 0.5-mL syringe and then injected into a gas chromatograph fitted with a 30-cm glass column (3.2 mm ID).

Transformation in pear fruit
The fruit of cv.Sucui No.1, which has yellow skin at ripening stage [48], was used for transient transformation.The over-expression vector of PbHB.G7.2 was constructed by inserting the full-length coding sequences of PbHB.G7.2 into the pSAK277 vector.
The constructed vector was transformed into Agrobacterium tumefaciens strain EHA105 by electroporation.The incubation and injection conditions were identical to a previous report [32].Agrobacteria containing the construct or an empty vector were infiltrated into 40 fruits.
For gene silencing, a 520-bp fragment of PbHB.G7.2 was inserted into the pTRV2 vector, and then introduced into A. tumefaciens strain GV3101.All incubations and transient transformation conditions were identical to a previous report [35].Agrobacteria containing pTRV1 and the construct were co-infiltrated into 40 fruits, while Agrobacteria containing the pTRV1 and pTRV2 empty vectors were co-infiltrated into the other 40 fruits.At least five fruits were weighted and sealed in a 3-L airtight jar, and four replicates were prepared.Ethylene production was measured when the phenotype of the fruit over-expressing or silencing PbHB.G7.2 was different from the control.Each sample of fruit f lesh covered by bacteria was collected for qRT-PCR analysis.
Moreover, the Agrobacterium harboring the over-expression vector of PbHB.G7.2 or the empty vector pSAK277 was used to infect the fruit calli of 'Clapp's Favorite' as shown in a previous report [49].All of the primer sequences are listed in Table S4 (see online supplementary material).

Yeast-one-hybrid assay (Y1H)
The full-length coding sequences of PbHB.G7.2, PbHB.G1, and PbHB.G2.1 were individually inserted into the pGADT7 vector.The 2000-bp sequences upstream of PbACS1b and the P9, P16, and P17 sequences were individually inserted into the pAbAi vector.The Y1H assay was performed using the Matchmaker Gold Yeast One-Hybrid Library Screening System (Clontech, Palo Alto, CA, USA).All of the primer sequences are listed in Table S4 (see online supplementary material).

Electrophoretic mobility shift assay (EMSA)
The full-length coding sequences of PbHB.G7.2, PbHB.G1, and PbHB.G2.1 were inserted into the pCold-TF expression vector and then introduced to E. coli strain BL21.Extraction and purification of the recombinant proteins were identical to a previous study [35].Moreover, based on the results detected by the dualluciferase and Y1H assays, the sequences of the biotin-labeled probes were 5 -CTTATGGGTCGGTTTTTCTACATTTGAATTTTGA-3 in the P9 sequences, 5 -TTATATGTTGTATCTTATTTCATCGTTGA TTAC-3 in the P16 sequences, and 5 -CCTCAAGGAAATAGTAATAC CCTTTAGATTTGAAGTCCTA-3 in the P17 sequences.An EMSA was performed according to a previous study [35].All of the primer sequences are listed in Table S4 (see online supplementary material).

Yeast-two-hybrid assay (Y2H)
A cDNA library was constructed with the mRNA extracted from the ripening fruits of cv.Cuiguan using Make Your Own 'Mate & Plate' Library System (Clontech).The Y2H screening was performed using the Matchmaker Gold Yeast Two-Hybrid System (Clontech).The full-length coding sequences of PbHB.G.7.2 were inserted into the pGBKT7 vector, and then used as bait for Y2H screening.

Quantitative real-time PCR and statistical analysis
Quantitative real-time PCR (qRT-PCR) was performed as described in a previous study [32].All of the primer sequences are listed in Table S4 (see online supplementary material).Means and standard errors were calculated using ANOVA.The significance at P < 0.05 was displayed by the lowercase letters (such as a, b, and c) or asterisk.

Figure 1 .
Figure 1.Involvement of PbACS1b, PbACO1, and PbHB.G7.2 in pear fruit ripening.a Three heatmaps revealed the expression profiles of ACS, ACO, and HB genes between the ripening and enlarging fruits in three pear cvs, Housui, Cuiguan, and Xueqing.b Expression analyses of PbACS1b, PbACO1, PbHB.G7.2, and PbHB.E10.2 in the enlarging and ripening fruits of three pear cultivars: 157 DAFB is the cv.Housui fruit at ripening stage, while 82 and 104 DAFB are the fruits at enlarging stages; 136 DAFB is the cvs.Cuiguan and Xueqing fruit at ripening stage, while 76 and 96 DAB are the fruits at enlarging stages.In cv.Housui, 157/104 represents the fold-change in RPKM value between 157 and 104 DAFB, while 157/82 represents the fold-change in RPKM value between 157 and 82 DAFB.In cvs.Cuiguan and Xueqing, 136/96 represents the fold-change in RPKM value between 136 and 96 DAFB, while 136/76 represents the fold-change in RPKM value between 136 and 76 DAFB.

Figure 2 .
Figure 2. Mediation of PbHB.G7.2 on ethylene biosynthesis and the PbACS1b expression in pear fruit callus.a Ethylene production was measured in the fruit calli that were either over-expressing PbHB.G7.2 or harboring an empty vector.b The qRT-PCR analysis showing the expression levels of PbHB.G7.2, PbACS1b, and PbACO1.c The constructed reporters and effectors for the dual-luciferase assay.d The PbACS1b promoter (P) was divided into four fragments (P1, P2, P3, and P4; Left panel) and then used for the dual-luciferase assay to identify the PbHB.G7.2 binding region (Right panel).e The PbACS1b promoter (P) was further divided into four additional fragments (P5, P6, P7, and P8; Left panel) and then used for the dual-luciferase assay to narrow the PbHB.G7.2 binding region (Right panel).f Reporters harboring the PbACS1b promoter (P) or the fragment P9 were screened in yeast cells by 50 ng/mL Aureobasidin A (AbA) (Left panels).The Y1H assay revealed the binding of the fragment P9 by PbHB.G7.2 (Right panel).PC and NC represent positive and negative controls, respectively.g A 34-bp fragment within the PbACS1b promoter was used as a biotin-labeled probe (Top panel).EMSA showed that PbHB.G7.2 bound to the biotin-labeled probe (Bottom panel).Cold probe concentrations were 10-fold (+) and 100-fold (++) of labeled probes.'+' and '-' indicate the interaction and non-interaction in the Y1H assay and indicate the presence and absence of cold probe, biotin-labeled probe, and recombinant PbHB.G7.2 in the EMSA, respectively.

Figure 3 .
Figure 3. Ethylene positively regulates pear fruit ripening.a Left panel shows that the pear fruit was treated by ethephon at 80 DAFB and was harvested at 104 DAFB.Right panel shows that the pear fruit was treated by ethephon at 80 DAFB and was harvested at 110 DAFB.Control and treatment represent untreated and ethephon/1-MCP-treated fruits, respectively.b Soluble solids was measured in ethephon−/1-MCP-treated and control fruits.c Ethylene production was measured in ethephon−/1-MCP-treated and control fruits.d Expression analyses of PbACS1b, PbACO1, and PbHB.G7.2 in ethephon−/1-MCP-treated and control fruits.