Sand pear (Pyrus pyrifolia Nakai. ‘Whangkeumbae’) fruit were collected as described in the study (Shi et al., 2019) from Hebei Agricultural University (Baoding, China). The pulp of these pears was collected for further study. Shoots were collected at 5 days old, young stems were collected at 10 days old, and young leaves were collected at 5 days old. Age (day old) was calculated from the starting time when buds that had just emerged in pear trees could be observed by the naked eye. All samples including petals and anthers were prepared as described in Shi et al. (2019).

The pear fruit with uniform size, free from mechanical damage, diseases and pests were screened at 150 days after full bloom (DAFB) and stored at 4°C for two months and then transferred to shelf life at room temperature for 36 h. Fruits were divided into two groups. One group was soaked in 200 μM SA (Solarbio, Beijing, China) for 12 h. The second group was soaked in distilled water for 12 h as control. The SA-treated and control fruit were air-dried and stored at room temperature for 12 h and 24 h. All fruit were stored at room temperature for 36 h and sampled every 12 h. For each sampling point, three biological replicates with five fruit each were used for analysis.

Pulp samples at one week after fruit harvest were separately soaked in 2, 20, 200, and 2000 μM SA (Solarbio, Beijing, China) for 12 h at room temperature, and the control was soaked in distilled water for 12 h (Shi et al., 2019). One week after harvest, fruit pulp samples were soaked in 1 μL/L ethephon (Solarbio, Beijing, China; Zhang et al., 2013) and 200 μM IAA (Solarbio, Beijing, China) for 6, 12, and 24 h (Shi and Zhang, 2014). Concurrently, the controls of the ethephon and IAA treatments were soaked in distilled water for 6, 12, and 24 h, respectively. The pear fruit pulp was treated with 15% glucose (Solarbio, Beijing, China; Shi et al., 2019) for 6, 12, and 24 h. The pulp was soaked concurrently in ddH₂O for 6, 12, and 24 h as corresponding controls. All samples were frozen as described in the previous study (Shi et al., 2019).

Total RNA was extracted from pear fruit that were collected at 150 d after full bloom. Complementary DNA (cDNA) library was constructed as described in primary study (Shi and Zhang, 2012) for sequencing. One clone that encoded EIN3 was identified and designated as PpEIN3a.

Total RNA was isolated from the shoots, young stems, young leaves, petals, anthers, developing pear pulp, and pulp that was treated with SA/ethephon/IAA/glucose by a previously described method (Li et al., 2005).

The PpEIN3a expression profiling was performed using real-time quantitative PCR via a Magic SYBR mixture according to the manufacturer’s instructions (CoWin Biosciences, China) in the detection system (Mastercycler ep realplex 4, Eppendorf AG, Hamburg, Germany). A pear actin gene was used as a standard control for the real-time quantitative PCRs. A two-step RT-PCR procedure was carried out in all experiments via a previously described method (Li et al., 2005). The gene-specific primer sequences for PpEIN3a were 5’-GGAGTTGATGATGGGCAGAAAATG-3’ and 5’-GGTTCAGACATGTTGATGTTGCAT-3’. The relative value for the expression level of PpEIN3a gene was calculated by Y=10^DCt/3.75*100% (DCt is the differences of Ct between the control actin products and the target PpEIN3a products, i.e., DCt=Ct _PpEIN3a -Ctactin, and 3.75 is a parameter for the cycle number which represents a 10-fold expression difference between the target gene and control gene).

The sequence analyses of the isolated pear EIN3a gene (cDNA) and its deduced protein were done with DNA STAR software (DNA STAR Inc., Madison, WI, USA). The conserved domain was determined by National Center for Biotechnology Information (NCBI) Conserved Domain Search (http://blast.ncbi.nlm.nih.gov/Blast.cgi). The protein multiple sequence alignment analysis was performed with Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/), and protein motif analysis was performed using Motif Scan (http://myhits.isb-sib.ch/cgi-bin/motif_scan). The evolutionary relationships among twenty-nine EIN3 protein sequences from different plants were determined by MEGA7 software (Kumar et al., 2016) based on the neighbor-joining method (Saitou and Nei, 1987). The genomic DNA sequence of PpEIN3a was analyzed according to genome information of sand pear (Gao et al., 2021) by using the GDR database (https://www.rosaceae.org/).

High performance liquid chromatography (HPLC) was used to determine the content of SA (Zhang et al., 2003). The frozen sand pear pulp was ground into good powder in liquid nitrogen, and for the extraction of SA, a 0.8 g sample was soaked with 1 mL 90% (v/v) methyl alcohol in the dark for 12 h at 4°C. Then, the sample was centrifuged at 8,000 g for 10 min at 4°C. The supernatant was transferred into a clean 10-mL centrifugal tube and stored for 2 h at 4°C. The remaining sediment was re-suspended in 0.5 mL 90% methyl alcohol and re-extracted for 2 h at 4°C. The above sample was centrifuged again. The two resulting supernatants were mixed together and dried with N₂ to approximately 0.3 mL aqueous solution. Then, 20 μL of 1 mg/mL trichloroacetic acid was added. After shaking, 1 mL cyclohexane and ethyl acetate mixture (V/V=1/1) was added for extraction, which was repeated twice. The upper organic phase was transferred to a clean 10-mL centrifugal tube and dried with N₂. A 0.5 mL methyl alcohol was added to dissolve and filter, and the obtained solution represented the free SA in the sample.

A 0.5 mL of 2 M HCl was added into the remaining aqueous phase. After shaking, the mixture was heated in 80°C-water bath for 1 h and, after cooling, was extracted twice with cyclohexane and ethyl acetate (V/V=1/1). The rest procedure is the same as above, and the obtained solution represented the conjugated SA in the sample.

The contents of free SA and conjugated SA in sand pear pulp were determined by high performance liquid chromatography (HPLC, Agilent 1260, USA). Agilent Eclipse Plus C18 column (250 mm × 4.6 mm, 5 μm; Agilent, USA) was used to elute the SA with 65% acetonitrile and methyl alcohol. The column temperature was 25°C, 20 μL of sample was used for detection, the flow rate was set at 1 mL/min, and the detection wavelength was 280 nm.

Each experiment was repeated independently three times, and the data were analyzed using GraphPad Prism 8. Independent t tests for equality of means showed very significant differences (** p value < 0.01) between the controls and treated fruit.

A 2 g frozen sand pear pulp was ground into powder with liquid nitrogen. It was transferred to a clean 10-mL tube and added 4 mL of 95% ethanol. The mixture was boiled and extracted for 20 min. After cooling, the mixture was centrifuged for 15 min at 10,000 g. Then, the supernatant was transferred into a clean 10-mL tube. The remaining sediment was re-suspended with 4 mL 80% ethanol and re-extracted for 30 min at 70°C. The above sample was centrifuged again. Two supernatants were merged together and evaporated to dry by a rotary evaporator (IKA RV10, GER) at 40°C. Then 2 mL of distilled water was added to evaporated residue. The evaporated residue was shaken and dissolved. The preparation solution was obtained and stored at −20°C for further analysis (Lizarda and Yang, 1979).

The ethylene production in frozen samples was detected following the methods of Fan et al. (1998) with some modifications. A 1 mL of the preparation solution was taken and put into a 10 mL sample bottle. Then 40 μL of 25 mM/L HgCl₂ was added into the bottle. The reaction sample bottles were then sealed with rubber stopper and kept in ice for 10 min. Then 0.2 mL of precooled mixture of 5% NaClO and NaOH (V/V=2/1) was injected into the reaction vessel by l-mL syringe. The mixture was shaken rapidly for 5 s and put into the ice for 5 min. A 1 mL of gas from the headspace was taken to measure the ethylene production by a gas chromatography (Agilent 7820A, USA) with an FID detector. Agilent GS-Alumina column (30 m × 0.53 mm) was used to elute the ethylene production. The column temperature was 80°C and the injection volume was 1 mL (Lizarda and Yang, 1979).

The method of Chen et al. (2003) for extraction of IAA was modified. A 2 g frozen sand pear pulp was ground into good power with liquid nitrogen. A 2 mL of pre-cooled 80% methyl alcohol was added into the power and the sample was transferred to a clean 10-mL tube. Following the remaining power was washed with 2 mL of pre-cooled 80% methyl alcohol and transferred to the clean 10-mL tube, which repeated twice. The mixture in the 10-mL tube was added with 500 μL of butylated hydroxytoluene (BTH) and soaked in the dark for 12 h at 4°C. In dark, the samples were shaken by ultrasonication and vortexing (Kunshan Schwimmer ultrasonic cleaner, KQ-100DE, China) for 1 h, and centrifuged at 12,000 g for 15 min at 4°C. The supernatant was transferred into a clean 10-mL tube and dried with N₂ to 1-2 mL aqueous solution, and the pH was adjusted to 8.0 by 1 M dipotassium hydrogen phosphate and 1% acetic acid. Following 0.4 g polyvinylpyrrolidone (PVPP) was added to remove impurities such as phenols. The mixture was shaken for 0.5 h by ultrasonication and then centrifuged at 12,000 g for 5 min at 4°C. The supernatant was transferred into a clean centrifuge tube, and the pH was adjusted to 3.0 via 1 M citric acid and 1% NaOH. Following 3 mL ethyl acetate was added into the aqueous phase, shaken for 0.5 h, and placed for 2 h at 4°C. The upper solution was transferred to a clean 10-mL tube and dried with N₂.

A C18 solid phase column was activated by adding, successively, 4 mL of methyl alcohol and 2 mL of sterile water. A 2 mL of 20% methyl alcohol was added into the 10-mL tube to dissolve the purified and dried supernatant. The mixture was poured into the activated C18 column. Following the column was washed with 2 mL of 10% methyl alcohol. Finally, 1 mL of pH 8 methyl alcohol was used to elute IAA from the column and through a 0.22 μm filter.

The determination of IAA was performed by HPLC (Agilent 1260, USA) and the C18 column (250 mm × 4.6 mm, 5 μm; Agilent, USA) was used to elute the IAA. Mobile phase A was sterile water, mobile phase B was acetonitrile, mobile phase C was 0.7% acetic acid, and mobile phase D was methyl alcohol. The column temperature was 30°C, 10 μL of sample was used for detection, the flow rate was set at 0.7 mL/min, and the detection wavelength was 254 nm.

Each experiment was repeated independently three times, and the data were analyzed using GraphPad Prism 8. Independent t tests for equality of means showed very significant differences (** p value < 0.01) between the controls and treated fruit.

The frozen sand pear pulp was ground into homogenate with 20 mL sterile water for determination of glucose content. A 0.5 g sample was transferred to a clean conical bottle and was heated in a 70-80°C water bath for 3 h. After cooling, the sample was centrifuged at 8,000 g for 10 min. Then, the supernatant was transferred into a clean 10-mL tube. A 1 mL of the final supernatant was filtered through a 0.22 μm filter before analyses (Wang, 2013).

The measurement of glucose content was performed by HPLC (Agilent 1260, USA) and chromatographic column (Shodex Asahipak NH2P-50 4E, 4.6 mm × 250 mm, 5 μm) was used to elute the glucose. Solvent was 25% sterile water and 75% acetonitrile. The flow rate was set at 0.6 mL/min, the column temperature was 32°C, and 5 μL of sample was used for detection.

The fruit were characterized after the 200 μM SA treatment by measurements of fruit firmness and soluble solids contents (SSC) in fruit samples. Firmness was measured on both sides of the fruit after peeling, using a hardness tester (GY-1, China) that was fitted with a 7.5 mm diameter probe at a rate of 10 mm/s. The data were recorded as kg-force and converted to Newtons (Huang et al., 2021). The SSCs of fruit were determined from juice samples from both sides of the fruit after peeling with a digital refractometer (“Pocket” PAL-1, 0-53%, Atago, Japan). Each experiment was repeated independently three times, and the figures were made with GraphPad Prism 8. Independent t tests for equality of means showed significant differences (*p value < 0.05) or very significant differences (** p value < 0.01) between the controls and treated fruit.

After measurement of firmness and SSC content, fruit slices of pulp without skin from same batch of fruit were frozen in liquid nitrogen and stored at −80°C until determination of hormone and glucose contents.

Each experiment was repeated independently three times, and the data were analyzed using SPSS. Independent t tests for equality of means showed significant difference (*p value < 0.05) or very significant difference (**p value < 0.01) between the control and treated fruit.

One cDNA encoding the EIN3 protein was identified by screening the sand pear fruit cDNA pool (Shi and Zhang, 2012). The cloned cDNA was designated as PpEIN3a, accession number in GenBank: KT726838. The genomic DNA sequence of PpEIN3a was analyzed according to sand pear genome database (Gao et al., 2021). PpEIN3a cDNA encodes an EIN3 homolog that consists of 604 amino acids and shares relatively high homology with pear PbEIL (Pyrus x bretschneideri, XP_009355060; 97.52% identity) and PuEIL (Pyrus ussuriensis x Pyrus communis, KAB2627254; 97.52% identity) at the amino acid level. The PpEIN3a protein contains a conserved ethylene insensitive 3 (EIN3) domain ( Figure 1 ). All three sequences contain a conserved ethylene insensitive 3 (EIN3) domain, which suggests that PpEIN3a may have similar function with PbEIL and PuEIL.

As shown in Figure 1 , twelve substitutions and two deletions at the amino acid level occurred in PpEIN3a compared with other fruit EIN proteins. Among these, two substitutions may be crucial for the function of PpEIN3a. Namely, the nonpolar Pro was replaced by a polar Ser at position AA571. At position AA572, the nonpolar Ile was substituted by a polar Asn. Moreover, at positions AA599 and AA600, PpEIN3a has deleted the two amino acids (e.g., Asp and Ala).

The analysis of the evolutionary relationships among PpEIN3a and other EIN3/EIL proteins, which was previously reported in plants, is shown in Figure 2 . The evolutionary tree was divided into two subgroups. The PpEIN3a protein has the closest evolutionary relationship with pear PbEIL (Pyrus x bretschneideri, XP_009355060); pear PuEIL (Pyrus ussuriensis x Pyrus communis, KAB2627254); wild apple MbEIN3 (Malus baccata, TQD73050) (Chen et al., 2019); and apple MdEIL1 (Malus domestica, XP_008346133) among all EIN3/EIL proteins. PpEIN3a also has a close evolutionary relationship with sand pear PpEIN3b (Pyrus pyrifolia, KT726839) (Shi et al., 2019). However, PpEIN3a has relatively distant evolutionary relationships with the six EIN3/EILs that occupy another clade of the tree (subgroup II, Figure 2 ). These results suggest that PpEIN3a might have diverged earlier from these EIN3s during evolution.

Real-time quantitative PCR analysis was performed to explore the PpEIN3a gene expression pattern. PpEIN3a was preferentially expressed in sand pear anthers. It exhibited a moderate expression activity in young stems and young leaves. A relatively weak expression signal of PpEIN3a was detected in the pulp, shoots, and petals ( Figure 3A ).

Further analysis of the expression pattern of PpEIN3a during fruit development was performed to investigate whether the expression of PpEIN3a is regulated during fruit development. The experimental results ( Figure 3B ) revealed that PpEIN3a exhibited the highest expression activity in 30-d-after-bloom fruit during whole-fruit development, and the relative value of this expression was 49.36. With fruit enlargement, the expression of the PpEIN3a gene gradually decreased to a relatively low level (e.g., relative value of 9.60) at 120 d after full bloom fruit. With fruit ripening, the expression of the PpEIN3a gene was nearly constant (e.g., 145 d after full bloom to 5 d after harvest). Interestingly, during fruit senescence, the expression of the PpEIN3a gene increased to its highest level (e.g., a relative value of 30.34) at 10 d after harvest fruit during fruit ripening and senescence. These results suggested that the PpEIN3a gene might play crucial roles during pear fruit senescence.

The above results showed that the PpEIN3a expressions were regulated during fruit senescence. SA is an important plant hormone that delays sand pear fruit senescence (Hassan et al., 2007). Therefore, SA treatments were performed to discover whether SA regulated PpEIN3a expression. As shown in Figure 4 , PpEIN3a expression was significantly downregulated by all SA concentrations. Moreover, there were very significant differences between the control and 20/200/2000 μM SA groups.

To explore the regulation patterns of PpEIN3a by ethylene, auxin, and glucose, the pulp from sand pear fruit at one week after harvest was treated with 1 μL/L ethephon, 200 μM IAA, and 15% glucose. The real-time quantitative PCR results showed that the transcript accumulations of PpEIN3a were significantly induced by the 1 μL/L ethephon treatment at 6, 12, and 24 h ( Figure 5A ). The PpEIN3a expressions were also markedly upregulated by the 200 μM IAA treatment at 6, 12, and 24 h ( Figure 5B ). Interestingly, the expression of PpEIN3a was induced by the 15% glucose treatment at 3 and 6 h ( Figure 5C ). These results showed that SA played an antagonistic role toward ET, IAA, and glucose in regulating the expression of PpEIN3a to delay pear fruit senescence.

Free SA contents in control fruit increased at shelf life of 12 h and then decreased, and in SA-treated fruit increased at shelf life of 12 h and 24 h, while decreased at shelf life of 36 h. However, compared with the controls, the contents of free SA in sand pear obviously increased after SA treatment at shelf life of 12 h and 24 h ( Figure 6A ). Moreover, there was no obvious change on conjugated SA contents of control fruit during shelf life, while the conjugated SA contents in sand pear significantly increased after SA treatment at shelf life of 24 h and 36 h ( Figure 6B ). The results indicated SA could induce the free and conjugated SA contents in sand pear during shelf life.

As shown in Figure 6C , ethylene production increased at shelf life of 12 h, 24 h, and 36 h in both control and SA-treated fruits. However, the ethylene production was significantly decreased in SA-treated fruits compared with that in controls during all the shelf life ( Figure 6C ). IAA contents of control fruit increased during the shelf life at 24 h and 36 h. While the content of IAA dramatically decreased after SA treatment compared with control at the same shelf-life time points, 24 h and 36 h ( Figure 6D ). The results showed that SA inhibited the increase in ET and IAA content of sand pear fruit. Interestingly, the contents of glucose in control fruit increased during the late shelf-life (24 h and 36 h), and in SA-treated fruit increased during all the shelf life. Moreover, the content of glucose increased when fruit were treated by SA for 12 h and then there were no differences between SA treatment and control fruit during the shelf life ( Figure 6E ). These results showed that SA played an antagonistic role toward ET and IAA during fruit senescence.

Additionally, sand pear fruit firmness in both the control and SA-treated samples showed a downward trend with longer shelf times. However, the SA-treated fruit maintained much higher fruit firmness than the controls ( Figure 7A ). This result showed that SA delayed the decrease in shelf-life sand pear fruit firmness to make fruit fresh. Moreover, the soluble solids content (SSC) in both the control and SA-treated fruit remained identical ( Figure 7B ), which suggested that the fruit quality was maintained in the SA-treated fruits.