FaPAO5 regulates Spm/Spd levels as a signaling during strawberry fruit ripening.

Abstract Polyamines are important for non‐climacteric fruit ripening according to an analysis of the model plant strawberry. However, the molecular mechanism underlying the polyamine accumulation during ripening has not been fully elucidated. In this study, an examination of our proteome data related to strawberry fruit ripening revealed a putative polyamine oxidase 5, FaPAO5, which was localized in the cytoplasm and nucleus. Additionally, FaPAO5 expression levels as well as the abundance of the encoded protein continually decreased during ripening. Inhibiting FaPAO5 expression by RNAi promoted Spd, Spm, and ABA accumulation while inhibited H2O2 production, which ultimately enhanced ripening as evidenced by the ripening‐related events and corresponding gene expression changes. The opposite effects were observed in FaPAO5‐overexpressing transgenic fruits. Analyses of the binding affinity and enzymatic activity of FaPAO5 with Spm, Spd, and Put uncovered a special role for FaPAO5 in the terminal catabolism of Spm and Spd, with a K d of 0.21 and 0.29 µM, respectively. Moreover, FaPAO5 expression was inhibited by ABA and promoted by Spd and Spm. Furthermore, the RNA‐seq analysis of RNAi and control fruits via differentially expressed genes (DEGs) indicated the six most enriched pathways among the differentially expressed genes were related to sugar, abscisic acid, ethylene, auxin, gibberellin, and Ca2+. Among four putative PAO genes in the strawberry genome, only FaPAO5 was confirmed to influence fruit ripening. In conclusion, FaPAO5 is a negative regulator of strawberry fruit ripening and modulates Spm/Spd levels as a signaling event, in which ABA plays a central role.

The known fruit ripening-related tomato mutants have enabled researchers to use tomato as a model species for studying the effects of PA on ripening (Dibble et al., 1988;Gapper, McQuinn, & Giovannoni, 2013;Hao et al., 2018;Liu et al., 2018;Mattoo et al., 2006Mattoo et al., , 2007Mehta et al., 2002;Mutschler, 1984;Saftner & Baldi, 1990;Sharma, Pareek, Sagar, Valero, & Serrano, 2017;Tassoni, Watkins, & Davies, 2006;Tsaniklidis et al., 2016). An increase in Put contents may prolong the tomato fruit ripening process (Dibble et al., 1988;Saftner & Baldi, 1990;Tassoni et al., 2006;Tsaniklidis et al., 2016). In transgenic tomato plants, the overexpression of the yeast S-adenosylmethionine decarboxylase (SAMDC) gene (ySAMdc) that functions in the production of decarboxylated S-adenosylmethionine (dcSAM) for both Spd and Spm biosynthesis, results in the ripening-specific accumulation of Spd, Spm, and ethylene, which promotes the lycopene accumulation and quality of fruits (Mehta et al., 2002). Analyses of the ySAMdc-overexpressing (OE) transgenic tomato plants and metabolic activities revealed that PAs function as anti-apoptotic regulatory molecules to revive tomato fruit metabolic memory (Mattoo et al., 2006(Mattoo et al., , 2007. Although PA biosynthesis in tomato fruits is committed to Arg or Orn, the induction of the Orn pathway decreases in the later ripening stages (Lasanajak et al., 2014). In transgenic tomato fruits expressing the mouse ornithine decarboxylase (ODC) gene, the Put, Spd, and Spm contents reportedly increase, while ethylene production, the respiration rate, and water loss decline, ultimately resulting in a considerably delayed fruit ripening (Pandey, Gupta, Chowdhary, Pal, & Rajam, 2015). Spatial and temporal analyses of PA biosynthesis and the expression of metabolism-related genes, including ODC, ADC (arginine decarboxylase), CuAO (copper-containing amine oxidase, also known as DAO), SPDS (Spd synthase), SPMS (Spm synthase), SAMDC, and PAO (polyamine oxidase), indicated that these genes more or less take part in PA biosynthesis and metabolism during the early fruit development stage.
Moreover, the high CuAO and SPMS transcription levels during the later fruit development stage are consistent with a sharp increase in fruit size. Furthermore, SPDS1 is the most highly expressed of these genes during the fruit ripening process, and most of these genes are highly expressed in fast-growing tissues, with CuAO playing an especially important role in ripening fruits (Tsaniklidis et al., 2016). A recent study proved that the SlPAO2, 3, and 4 expression levels are upregulated, whereas SlPAO1, 5, 6, and 7 expression levels are downregulated in developing tomato fruits .
Grape and strawberry are typically used for investigating non-climacteric fruit ripening. Compared with the climacteric fruit ripening involving ethylene, non-climacteric fruit ripening is an ethylene-independent process with only slight changes in ethylene emission (Alexander & Grierson, 2002). Although Put and Spd contents sharply decrease, while Spm is maintained at stable levels during grape berry ripening (Shiozaki, Ogata, & Horiuchi, 2000), PA catabolism indeed plays a vital role in berry enlargement and aroma production (Agudelo- Romero et al., 2014). The OsPAO5 transcript level gradually increases during rice seed germination, but this increase is inhibited by 5 mM guazatine (Chen et al., 2016). An increase in PAOderived hydrogen peroxide (H 2 O 2 ) levels is terminated by the addition of the PAO-specific inhibitor guazatine in Arabidopsis (Toumi et al., 2019). Exogenous ABA enhances the expression of the maize polyamine oxidase gene, which contributes to the ABA-induced cytosolic antioxidant defense via H 2 O 2 (Xue, Zhang, & Jiang, 2009  . In comparison to more progress in the role PAs in climacteric tomato fruit ripening, a role of PAs in non-climacteric fruit ripening is just onset , many important issues remain to be solved, especially a role of PA metabolism in strawberry fruit ripening. In the present study, on the basis of our proteome data, a probable polyamine oxidase 5, FaPAO5, was screened, after which a similar trend in the transcript and protein levels during ripening was confirmed. Pharmacological, physiological, biochemical, and molecular analyses were completed, including a subcellular localization as well as an examination of expression-influencing elements, transient transgenic fruits, and RNAi (RNA interference) and transgenic fruit RNA-seq data, in addition to an evaluation of enzymatic activities based on HPLC and isothermal titration calorimetry (ITC). We determined that FaPAO5 negatively modulates strawberry fruit ripening by regulating Spm/Spd levels.

| Plant material
Strawberry (Fragaria × ananassa cultivar "Zhangji") plants were cultivated in a greenhouse at 23-28°C with 60%-70% relative humidity in the spring of 2017 and 2018. On the basis of a previous report , the following seven strawberry fruit developmental stages were analyzed: SG, LG, DG, Wt, IR, PR, and FR, which corresponded to 7, 13, 17, 22, 24, 26, and 28 DAA, respectively. Twenty fruits of a uniform size were sampled at each stage (n = 20; each fruit representing a biological replicate) and then immediately frozen in liquid nitrogen and stored at − 80°C until analyzed.

| Proteome analysis of strawberry fruit ripening
A combination of tandem mass tag quantitative proteomics and liquid chromatography tandem mass spectrometry analyses were completed by Jingjie PTM Biolab Co. Ltd. (Hangzhou, China). Specifically, strawberry fruits from the LG, Wt, IR, and PR stages were examined. The extracted proteins were separated by 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis to ensure the quality of the extractions was appropriate. Protein concentrations were determined with a BCA kit (Biyuntian, China). Differentially abundant proteins were defined as those with abundances that were >2-fold (p < .05) different among fruit development stages. Proteins were functionally classified based on Gene Ontology terms from the following three main categories: biological process, cellular component, and molecular function. Additionally, the enriched KEGG pathways (http://www.genome.jp/kegg/) among the differentially abundant proteins were identified based on a two-tailed Fisher's exact test.
The GO/KEGG annotations with a corrected p-value < 0.05 were considered significant. This analysis was repeated three times. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD018073.

| Subcellular localization
The FaPAO5 coding sequence was cloned into the pSuper1300-GFP vector at the SpeI and KpnI restriction enzyme sites (Primers used to amplify FaPAO and vector construction see Table S3). The FaPAO5-GFP recombinant plasmid was inserted into tobacco (N. tabacum) leaf cells via an A. tumefaciens (strain GV3101) mediated infiltration.
At 72 hr after the infiltration, the inferior tobacco leaf epidermis was removed and the GFP signal was observed with the LSM 710 META confocal laser-scanning microscope (Zeiss, Germany). The analysis was repeated three times.

| Determination of PA, anthocyanin, and soluble sugar contents
Three randomly selected fruits were analyzed by HPLC to quantify the PA, anthocyanin, and soluble sugar contents as previously described . The HLPC was completed with the following columns: ZORBAX Eclipse XDB-C18 column (4.6 × 250 mm, 5 µm; Agilent) at 30°C for detecting PAs; the ZORBAX Eclipse XDB-C18 column (4.6 × 150 mm, 5 µm; Agilent) for detecting anthocyanins; and the Agilent Technologies 1,200 Series Sugar-Pak™ column (6.5 × 300 mm; Waters, Milford, MA, USA) for detecting soluble sugars. The analysis was repeated three times.

| Determination of fruit firmness and soluble solid concentrations
Three fruits were evaluated to determine the fruit firmness and soluble solid concentrations. Fruit firmness was measured with the FHM-5 fruit hardness tester (Takemura Electric Works Ltd, Tokyo, Japan). The soluble solid concentrations in receptacles were determined with the MASTER-100H sugar analysis instrument (ATAGO, Tokyo, Japan) as described by Jia et al. (2013). The analyses were repeated three times.

| RNA isolation, cDNA synthesis, gene cloning, and qPCR
Total RNA was extracted from 3 g receptacle tissue from three randomly selected fruits with the OMEGA RNA extraction kit (OMEGA Biotek, Norcross, GA, USA). The purity and integrity of the RNA were assessed by gel electrophoresis and A260/A230 and A260/A280 ratios. The RNA (0.4 µg) was used as the template to synthesize the first-strand cDNA with the Trans kit (TransGen, Beijing, China).
Details regarding the primers used for the qPCR assay and FaPAO5 cloning are listed in Tables S4 and S5. The qPCR analysis was repeated three times.

| Construction of RNAi and OE plasmids and fruit agroinfiltration
The Gateway system was used for constructing the RNAi and OE plas-

| In vitro fruit disk incubation
Three randomly selected white strawberry fruits were cut into 1-mm disks. The fruit disks were treated with 100 µM Put, Spd, Spm, ABA, guazatine, or fluridone. The basic treatment solution (i.e., without these compounds) was used as a control. Samples were incubated for 2 hr as described by Jia et al. (2013). A qPCR analysis of FaPAO5 expression was completed as described above. The analysis was repeated three times.

| Isothermal titration calorimetry assay
The final concentration of the recombinant FaPAO5 fusion protein was adjusted to 10 µM in ITC buffer (200 mM NaCl, 20 mM Tris, 0.5 mM EDTA, and 10 mM maltose, pH 7.4). The ITC analysis was conducted with 10 µM recombinant FaPAO5 and 100 µM ligands (Put, Spd, and Spm; Sigma-Aldrich) in an ITC200 calorimeter (MicroCal, Northampton, MA, USA) at 30°C. The protein samples were titrated with a 2-μL injection of ligands every 5 min as described by Wild et al. (2016). Data were fitted with the ORIGIN 7.0 software (MicroCal). The assay was repeated three times.

| HPLC analysis of FaPAO5 activity and its products
To analyze the products of PA oxidation as described by Wang & Liu (2016), 10 μg purified FaPAO5 was incubated with 150 mM Spd in 100 mM phosphate buffer (pH 7.5) at 38°C or with 150 mM Spm in 100 mM phosphate buffer (pH 6.5) at 36°C. Samples were analyzed at different time-points. The reaction products were extracted with precooled perchlorate (5% volume fraction), which was followed by the derivation and benzoylation of the extracted PAs and Dap. The substrates and products were examined by HPLC with the ZORBAX Eclipse XDB-C18 column (4.6 × 250 mm, 5 μm; Agilent) at 30°C.
The mobile phase was methanol:water (v:v, 64:36) for 20 min at a flow rate of 0.7 ml/min. The effluent absorbance was monitored at 230 nm. The injection volume was 10 ml. The HPLC experiment was repeated three times.

| RNA-sequencing of RNAi and control fruits
To further clarify the negative regulatory function of FaPAO5 during strawberry fruit ripening, the RNA-seq analysis of FaPAO5-RNAi and control fruits was performed. Three fruits were randomly sampled for an RNA isolation and cDNA synthesis, which were completed as previously described (Wang et al., 2017). Total

| Expression pattern and subcellular localization of FaPAO5 based on proteomic data
An analysis of our proteome data for large green (LG), white (Wt), initially red (IR), and partially red (PR) fruits around the onset of strawberry fruit ripening uncovered three polypeptides related to a putative polyamine oxidase 5 in the National Center of Biotechnology Information (NCBI) database (https://blast.ncbi.nlm.nih.gov/Blast. cgi) that continually decreased in abundance (log2 fold-change ≥ 1) (Fig. S1a). The screened protein, which was named FaPAO5, comprised 552 amino acids and was encoded by a 1,659-bp coding sequence. In developing fruits, a quantitative real-time PCR (qPCR) analysis indicated that the FaPAO5 expression level was continually downregulated from the small green (SG) to the LG, de-greening (DG), Wt, IR, and PR fruit stages, but increased slightly from the PR to FR fruit stages (Fig. S1b). These results suggested that FaPAO5 might negatively regulate fruit ripening.

| Factors influencing FaPAO5 expression detected in the fruit disk incubation test
To explore the effects of Spm, Spd, ABA, guazatine, and fluridone on FaPAO5 expression, fruit disk and pharmacological tests were performed. The FaPAO5 expression level increased in response to Spd, Spm, and fluridone (an ABA biosynthesis inhibitor), decreased after an exposure to ABA and guazatine, and was unaffected by Put (Figure 2a and b). These results indicated that ABA and guazatine inhibit FaPAO5 expression, whereas Spd and Spm have the opposite effect. Therefore, FaPAO5 expression may be influenced by Spm/Spd/ABA, but not by Put. Accordingly, Spm/Spd may be a positive regulator, whereas ABA is a negative regulator for FaPAO5 expression.
To confirm that FaPAO5 is a negative regulator of strawberry fruit ripening, Agrobacterium tumefaciens (strain GV3101) cells con-  Figure 3d). Finally, the ratios of dark red: red: chimeric fruits in the RNAi, OE, and control groups were 18:2:0, 0:1:19, and 3:17:0, respectively. The qPCR data confirmed that the FaPAO5 transcription levels were significantly downregulated by an average of 57% and upregulated by an average of 156% in the RNAi and OE fruits, respectively, relative to the control levels (Figure 4a). These data indicated that FaPAO5 serves as a negative regulator of strawberry fruit ripening.

| Manipulation of FaPAO5 expression affects ripening-related processes at physiological and molecular levels
To characterize the role of FaPAO5 in the regulation of strawberry fruit ripening, the transgenic dark red RNAi and chimeric OE fruits

| Analysis of FaPAO5 enzymatic activity
To investigate the FaPAO5 enzymatic activity, the corresponding coding sequence (from 1 to 1,659 bp) was expressed in Escherichia coli cells and purified. A 105-kDa recombinant protein (1.5 mg/ml) was purified via MBP resin (Fig. S2a) and identified in a western blot with an anti-His tag antibody (Fig. S2b). Additionally, an ITC titration was completed with 20 µM recombinant protein and 100 µM Put, Spd, or Spm, with a 2-μL injection of PAs every 5 min. The data proved that the binding of FaPAO5 to Spm and Spd followed a saturation kinetics curve, with a dissociation constant (K d ) of 0.21 µM ( Figure 6a) and 0.29 µM (Figure 6b), respectively, and the stoichiometry (N) approached a 1:1 FaPAO5 to Spm/Spd-binding ratio.
However, there was no specific curve for the binding of FaPAO5 to Put (Figure 6c). These data demonstrated that FaPAO5 binds tightly to Spm/Spd, but not to Put.
To further determine the enzymatic nature of FaPAO5, it was incubated with Spm/Spd/Put and the substrates and metabolic products were analyzed by HPLC. The HPLC results indicated that at the 0 min time-point (i.e., before the reaction between FaPAO5 and the polyamines), only Spd/Spm/Put were detected. At 30 min after initiating the reaction, in addition to the substrates, 1,3-diaminopropane (Dap) was detected as a product of the Spm/Spd reactions. The detected Dap signals increased at 60 min in a time-dependent manner (Figure 7a and b). In contrast, Dap was undetectable in the Put reaction ( Figure 7c). These results implied that FaPAO5 specifically degrades Spm/Spd via terminal metabolism (Cona, Rea, Angelini, Federico, & Tavladoraki, 2006). This observation was confirmed by the production of H 2 O 2 in the reaction between FaPAO5 and Spm/ Spd/Put at different pH and temperature conditions. Specifically, both pH and temperature affected H 2 O 2 production, but not for Put ( Fig. S3).

| Transcriptome analysis of RNAi and control fruits
To clarify the negative regulatory function of FaPAO5 during strawberry fruit ripening, the RNAi and control fruits underwent an RNAseq analysis as previously described (Wang et al., 2017 Table S2). These results suggested that sugar, phytohormones (especially ABA), and Ca 2+ might be involved in the FaPAO5mediated regulation of strawberry fruit ripening.

| Global analysis of the polyamine oxidase family
Plant PAOs generally belong to multiple gene families, with 11 members in the rice genome and at least five members in the Arabidopsis genome Moschou, et al., 2008;Chen et al., 2016). To determine the functions of PAO family members during strawberry fruit ripening, we comprehensively examined our transcriptome data for the LG, Wt, IR, and PR fruit developmental stages (Bioproject accession in GenBank: PRJNA438551; Guo et al., 2018). We detected four PAO genes in the strawberry genome (PAO1, PAO2, PAO4, and PAO5). We observed that PAO1 expression increased continuously, whereas PAO4 and PAO5 were stably expressed at low levels and PAO2 expression tended to decrease (Fig. S4). On the basis of the subcellular localization and functional analysis of PAO5, we demonstrated that both PAO1 and PAO4 are located in the nucleus, whereas PAO2 is localized more in the nucleus rather than in the cytoplasm (Fig. S5). The downregulated expression of PAO1, PAO2, and PAO4 did not result in observable phenotypic changes, implying these genes may not contribute to fruit ripening. Accordingly, among the strawberry PAO genes, PAO5 appears to be critical for fruit ripening.

| D ISCUSS I ON
Although polyamines have relatively lower levels in the mature tomato fruits, indeed, PAs play a role in the ripening (Tsaniklidis et al., 2016). Actually, the roles of PAs show both short-term and long-term effects on peach fruit ripening (Torrigiani et al., 2012). In apple, Spd was a predominant form of polyamines during fruit development and ripening (Zhang et al., 2003). In eggplant fruit, dominant polyamines are Put and Spd, especially more Put, while no Spm found during ripening (Rodriguez et al., 1999). On the whole, the roles of PAs vary with their compositions and contents, fruit types, and developmental stages, to some extents, the ratios of (Spd + Spm)/Put control fruit ripening and quality .
It is previously reported that Arabidopsis AtPAOs mediates polyamine (PA) back-conversion from Spm into Spd and then Put (Zarza et al., 2017). In the present study, we find a special role of FaPAO5 in degradation of Spm/Spd by the way of terminal metabolism rather that the back-conversion metabolism, namely the purified FaPAO5 can specially oxidizes Spm or Spd into 1.3-diaminopropane (Dap) and H 2 O 2 , and cannot play a role in the degradation of Put. Notably, the downregulation of FaPAO5 expression decreased H 2 O 2 levels while promoted ABA accumulation, finally accelerated ripening, thus we consider that ABA rather than H 2 O 2 plays a vital role in the FaPAO5mediated ripening as a negative regulator. Given that ABA increases rapidly during strawberry fruit ripening , the higher ABA inhibited FaPAO5 expression, as a result, promoting both PA and ABA accumulation, thus a coordination mechanism between PA and ABA interaction, uncovers a new insight for the FaPAO5 in ripening, namely, FaPAO5, not only serves as enzyme function, but also acts as a signal role.

| FaPAO5 plays a central role during strawberry fruit ripening
The rice PAO homologs are classified into four subfamilies (I, IIa, IIb, and III;Chen et al., 2016). The Arabidopsis genome encodes at least five putative PAOs, of which AtPAO1 converts Spm to Spd in a back-conversion pathway and AtPAO3 catalyzes the sequential conversion of Spm to Spd and then to Put Moschou et al., 2008). In the present study, we identified four putative PAO genes in the strawberry genome, among which FaPAO1 and FaPAO4 were classified into a large group that included AtPAO2, AtPAO3, and AtPAO4. Additionally, FaPAO2 and FaPAO5 were classified into another large group along with OsPAO2 and OsPAO5 (Fig.   S6). Moreover, FaPAO5 and OsPAO5 were classified into a subgroup, whereas AtPAO5 was included in another subgroup. Consistent with the subcellular localization of OsPAO5, both FaPAO2 and FaPAO5 were located in the nucleus and cytoplasm (Fig. S5), suggesting that FaPAO2 and FaPAO5 may influence gene expression and regulation.
The fact that FaPAO1, 2, and 4 do not affect fruit ripening (data not shown) and only PAO5 is related to fruit ripening among the PAOs screened from the proteomic data imply that PAO5 is vital for regulating strawberry fruit ripening.

| Polyamines play essential roles in strawberry fruit ripening
Over the past several decades, substantial progress has been made in characterizing the PA roles related to fruit development and ripening in tomato (Mutschler, 1984;Dibble et al., 1988;Saftner and Baldiet, 1990;Mehta et al., 2002;Tassoni et al., 2006;Mattoo et al., 2006Mattoo et al., , 2007Gapper et al., 2013;Tsaniklidis et al., 2016;Sharma et al., 2017;Hao et al., 2018;Liu et al., 2018). Notably, a recent report described a positive role for Spm and Spd, especially Spm, in the regulation of strawberry fruit ripening because it manipulates the expression of FaSAMDC, which is important for the biosynthesis of Spd and Spm . This report provides an insight into the role of PAs in non-climacteric fruit ripening via FaSAMDC. However, in contrast to the FaSAMDC-mediated biosynthesis of Spd and Spm during ripening, a role for the PAO-mediated degradation of Spd and Spm in ripening fruits remains unknown.
In the present study, proteomic data combined with the results of pharmacological, physiological, biochemical, and molecular analyses revealed that (a) FaPAO5 abundance and FaPAO5 expression F I G U R E 8 Model for FaPAO5-mediated regulation of strawberry fruit ripening via ABA. Strawberry fruit ripening from large green (LG) to white (Wt), initially red (IR), partially red (PR), and fully red (FR) fruit stages. NCED3 promotes ABA accumulation, which inhibits FaPAO5 expression and Spm/Spd accumulation, thereby promoting SAMDC, SPDS, and SPMS expression, Spm/Spd accumulation, and fruit ripening. The graphic symbols ( ) indicate induced and inhibited, respectively continuously decrease during fruit development and ripening ( Figure   S1); (b) downregulated and upregulated FaPAO5 expression induces and inhibits ripening, respectively ( Figure 3); and (c) the K d values of the reactions between FaPAO5 and Spm/Spd are 0.21/0.29 µM, respectively ( Figure 6). These findings provide substantial evidence that FaPAO5 negatively influences fruit ripening, whereas Spm/Spd are positive regulators. Indeed, the results of a previous study on the FaSAMDC-mediated biosynthesis of Spd and Spm to promote strawberry fruit ripening     (Table S1).

| Characterizing the role of FaPAO5 in strawberry fruit ripening via ABA and Spm/Spd
The results of previous investigations and the current study indicate that: (a) ABA is important for strawberry fruit ripening  and NCED3 was screened based on the RNA-seq analysis of RNAi fruits (Table S2) (Figures 6 and 7). We developed a model for the PA regulation of strawberry fruit ripening as part of a signaling event. With the onset of ripening, NCED3 promotes the rapid accumulation of ABA, which may inhibit FaPAO5 expression, leading to Spm/Spd accumulation. The increased Spm/Spd may trigger SAMDC, SPDS, and SPMS expression, further promoting Spm/Spd accumulation, which ultimately accelerates strawberry fruit ripening (Figure 8). Future investigations should aim to uncover the molecular mechanism underlying the ABA-mediated inhibition of FaPAO5 expression.

ACK N OWLED G M ENTS
The authors offer particular thanks to Dr. Qing Zhang for ITC analysis.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest associated with the work described in this manuscript.