Pineapple (A. comosus cv. ‘Comte de Paris’) at stage 3 of maturation as proposed by Selvarajah et al. (2001) was obtained from a commercial orchard in Xuwen County, Guangdong Province in winter. Fruits of uniform size, color, and maturity, as well as being free of physical injury or disease symptoms, were selected, with the crown intact. Subsequently, fruits were immersed for 2 min in 0.05% (w/v) benomyl aqueous solution to control decay, air-dried at 25°C for 2 h and treated as follows. The selected pineapple fruits were immersed into 0.1% (v/v) hexaldehyde containing 0.01% Tween 80 for 5 min. The control fruits were immersed with distilled water, which also contained 0.01% Tween 80. Each treatment was applied to three replications, and each replication contained 100 fruits. Following treatment, all fruits were placed into unsealed plastic bags (0.04 mm thick), and stored at 25°C in the dark with relative humidity at 95% for 12 days. Samples of pulp tissues were collected at days 0, 2, 4, 6, 8, 10, and 12 frozen in liquid nitrogen; and stored at -80°C for future use.

For IB assessment, at least six fruits for each treatment were halved longitudinally to observe the incidence and severity of IB injury. IB intensity was visually scored from 0 (no IB) to 5 (maximum IB) according to Teisson (1979).

A putative pineapple PLD cDNA was identified by searching the pineapple genome database. The homology of predicted PLD protein sequences was checked using Clustal W, with the default parameters set by the DNA Data Bank of Japan¹. Domain and motif searches were conducted on the protein sequences in SMART². Blast searches were also performed against nucleic acid sequence data repositories at the National Center for Biotechnology Information (NCBI³). A phylogenetic tree was generated using Clustal Omega online tools⁴. The logos of the conserved domains were constructed using MEME program⁵ with the default parameters. The theoretical isoelectric point (pI) and molecular weight of the pineapple PLD proteins were calculated using PeptideMass program⁶. Transmembrane helix number was predicted by HMMTOP online tools⁷.

Total RNA was extracted from pineapple pulp using the hot borate method of Wan and Wilkin (1994). RNA purification and first-strand cDNA synthesis were performed as described in our previous paper (Hong et al., 2016). qRT-PCR analyses were conducted to detect gene expression patterns of the PLD family during the development of IB in fruits treated with PLD inhibitor (hexaldehyde). The oligonucleotide primers for qRT-PCR analysis were designed using Primer 5.0 software and Acactin (accession number HQ148720) was used as a reference gene. Reactions were carried out using SYBR^® premix Ex Taq^TMII (Takara, Shiga, Japan) on a LightCycler^®480 II (Roche, Switzerland). The relative expression levels of AcPLDs were calculated using the method of Livak and Schmittgen (2001). Values represented the average of three biological replicate. The sequences of all primers are listed in Supplemental File 1.

The coding sequences of AcPLD2 and AcPLD9 without the stop codon were amplified by PCR (primers are listed in Supplemental File 2) and subcloned into the pEAQ-GFP-HT vector to form fusion constructs AcPLD2::GFP and AcPLD9::GFP, under the control of the CaMV 35S promoter. The fusion constructs and a control vector were maintained in Agrobacterium tumefaciens strain EHA105 by electroporation. Transient transformation of tobacco leaves was carried out according to the method described by Sainsbury et al. (2009). Localizations of AcPLD2::GFP and AcPLD9::GFP were investigated with a fluorescence microscope (Zeiss Axioskop 2 Plus). All transient expression assays were repeated three times.

The pineapple pulp samples (1.0 g) were homogenized in 9.0 ml of phosphate buffered saline (PBS) (pH 7.4) and then centrifuged at 4000 ×g for 15 min. The supernatant solution was collected for analyses of SOD, CAT, NADPH, APX, PLD, and H₂O₂ contents, which were quantified by enzyme-linked immunosorbent assay (ELISA). ELISA was carried out as follows: (1) The 96-well microplates were coated with 100 μL of the captured specific antibody at 1:2000 dilution in coating buffer and incubated overnight at 4°C. (2) Plates were washed three times with PBS to remove unbound antibody, and each well was filled with 300 μL of blocking buffer and incubated for 30 min at 37°C. (3) Followed by washing, 100 μL sample solution mentioned above was added to each well and incubated for 30 min at 37°C. (4) After washing three times, 100 μL horseradish peroxidase-conjugated the specific antibody was added to each well and incubated at 37°C for 30 min. (5) The plate was washed again, and 100 μL of solution 3,3/5,5/-tetramethylbenzidine was added to each well. (6) After developing the color for 30 min at 37°C, the reaction was terminated after the addition of 50 μL of 2 M H₂SO₄ per well. (7) The absorbance values at 450 nm were determined, and the values of ELISA were given in relative units (ng.mg^-1protein). Protein content was measured by the method of Bradford (1976) using bovine serum albumin as the standard protein.

The experiment was performed in triplicate (n = 3). Data were presented as the mean ± standard error. Least significant difference at the 5% level was analyzed by SPSS software (version 13.0, SPSS Inc., Chicago, IL, United States).

Control fruits showed signs of IB after 6 days of storage, whereas no sign of IB was detected in fruits treated with hexaldehyde. When the treated fruits began to exhibit the IB at day 8 of storage, the IB intensity reached 1.7 in control samples. IB progressively increased as storage duration increased. The IB intensity was higher for control samples than for treated fruits, and a noticeable difference was observed between them after 6 days of storage (Figure 1).

Ten putative PLD genes were identified in pineapples by searching against the pineapple genome database. These genes were located on chromosomes 1, 2, 3, 5, 7, 8, 11, 18, 19, and 25, and their temporary names were designated as AcPLD1–AcPLD10 based on their order on the pineapple chromosome from 1 to 25 (Table 1). AcPLD1–AcPLD10 were predicted to encode proteins of 852, 813, 905, 814, 977, 1032, 850, 1126, 516, and 874 amino acids, respectively, with calculated molecular weights of 96.6, 91.2, 101.7, 92.4, 111.0, 114.5, 96.3, 127.8, 57.3, and 98.7 kDa, respectively, and pI values of 8.00, 5.88, 6.57, 5.80, 6.73, 6.91, 6.78, 6.50, 7.15, and 6.79, respectively. Alignment of the full-length deduced proteins of the AcPLDs clearly showed that the proteins contained two conserved HKD domains (HKD1 and HKD2) in their sequences. HKD1 regions displayed more diversity than HKD2 regions (Figures 2, 3). The two HKD domains were separated by 320 amino acids, except for PLD3, PLD5, PLD8, and PLD9, which demonstrated longer or shorter spacing sequences. The “IYIENQYF” motif, which was found among the 10 pineapple PLDs, with the exception of PLD9, was the most conserved domain with the least diversity. The C2 domain was present in pineapple PLDs, except PLD8 and PLD9. The PH domain was present in pineapple PLD8 but not in other pineapple PLDs (Figures 2, 3 and Tables 1, 2). Alignment analysis indicated that the similarities of the deduced amino acids of AcPLDs varied from 10% (AcPLD4 and AcPLD9; AcPLD8, and AcPLD9) to 73% (AcPLD5 and AcPLD10; Table 3).

To examine the phylogenetic relationship between the AcPLD proteins and other plant species, a phylogenetic tree was constructed on the basis of their translated amino acid sequences. The whole plant PLD proteins could be grouped into six types: α, β, γ, δ, 𝜀, and ζ (Qin and Wang, 2002). Based on the tree, the 10 PLD genes in pineapple were clustered into five types. AcPLD2, AcPLD4, and AcPLD7 were clustered to α type, and AcPLD6 was clustered to β or γ type. AcPLD1, AcPLD3, AcPLD5, and AcPLD10 were δ type, whereas AcPLD8 and AcPLD9 were ζ type (Figure 4).

To investigate the potential function of the AcPLD proteins, the transmembrane helix number was analyzed via topology prediction using HMMTOP online tools. The resulting data showed that zero, one and two transmembranes existed in the associated AcPLDs (Table 1).

These data suggested that AcPLD1–AcPLD10 exhibited great structural diversity and possibly exerted different physiological functions.

To understand the possible role of AcPLD1–AcPLD10 in pineapple during IB, the expression patterns of AcPLD1–AcPLD10 in pulp of fruits treated with hexaldehyde were investigated by qRT-PCR. As shown in Table 4, among the 10 AcPLD genes, AcPLD2 (1.6- to 5.7-fold) was strongly induced after hexaldehyde treatment. By contrast, the transcript level of AcPLD9 (1.9- to 4.0-fold) decreased following hexaldehyde treatment. AcPLD7 and AcPLD8 showed similar expression patterns, and they decreased by 1.3- to 2.5-fold and 1.1- to 2.3-fold in treated fruits, respectively, and did not substantially change between the treated and non-treated fruits. In addition, AcPLD1 and AcPLD3 transcripts remained nearly unchanged, and AcPLD4-6 fluctuated for both control and treated fruits during storage.

AcPLD2 and AcPLD9 were selected for subcellular localization analysis due to their gene expression patterns with obvious characteristics during IB. To detect the potential role of the two proteins (AcPLD2 and AcPLD9), we assayed their localizations in the cell. The results indicated that the fluorescent signals from the AcPLD2-GFP and AcPLD9-GFP constructs were observed in the plasma membrane of tobacco leaf epidermal cells, whereas the fluorescence from the GFP control was distributed within the cells (Figure 5).

To further investigate the potential regulatory mechanism of hexaldehyde treatment alleviating IB, we analyzed the contents of H₂O₂ and enzymes related to H₂O₂. As shown in Table 5, the SOD content was decreased before the eighth day of storage and then increased in the treated fruits compared with the control samples. The contents of CAT, NADPH, and APX abundances showed similar patterns; they increased and peaked after 6 days of storage and then decreased in treated and non-treated fruits. However, the CAT, NADPH, and APX levels were suppressed at day 2, and their contents were induced in treated fruits than in non-treated samples. The H₂O₂ content clearly fluctuated after hexaldehyde treatment during storage, increased at days 4 and 6, decreased at day 8, and increased again at day 12 in treated fruits. A significant difference was observed between the treated and non-treated fruits.

Hexaldehyde-treated fruits showed slightly higher PLD contents than control samples on day 2. However, on day 4, the PLD content was significantly higher in the former than in the latter. Hexaldehyde-treated fruits had lower PLD content than control samples, and treated fruits had relatively higher PLD content than non-treated fruits. Significant differences were noted between the groups (Table 5).