Actinidia arguta (A. arguta) is a kind of climacteric fruit that quickly softens and limits fruit shelf-life and commercial value. Therefore, it is of great significance to develop kiwifruit genotypes with an extended shelf-life of fruit. However, the ripening and softening mechanisms remain unclear in A. arguta. The mining of key gene involving in fruit ripening process is the basic of revealing ripening mechanism. With the help of transcriptome technology, combined with the results of gene differential expression profile and transient expression, we have completed the identification of the ripening-related gene AaPG18, which will help to understand the mechanism of fruit ripening in the future.
1. Introduction
Kiwifruit possesses rich germplasm resources, including 54 species and 21 varieties
[1]. For decades, the international market was mainly occupied by
Actinidia chinensis Planchon (
A. chinensis) and
Actinidia chinensis var. deliciosa A. Chevalier (
A. chinensis var.
deliciosa).
Actinidia arguta (Siebold and Zuccarini) Planchon ex Miquel
(A. arguta), a miniature kiwifruit, gradually developed into the second largest species to be cultivated worldwide
[2]. However,
A. arguta fruit at maturity is highly prone to softening, resulting in a shortened shelf-life. Therefore, it is necessary to explore the molecular mechanism underlying fruit ripening to provide a basis for the breeding of new cultivars with an extended shelf-life.
Fruit ripening is a continuous and dynamic process, accompanied by a series of physiological and biochemical changes including fruit color, fruit texture, and fruit hardness
[3][4]. Cell wall metabolism is the key stimulus causing the change in fruit texture. Enzymes encoding genes for cell wall degradation such as polygalacturonases (
PGs) and pectate lyases (
PELs) play vital roles in the softening process through the disassembly and dissolution of cell wall components
[5][6]. Previous studies have confirmed that the biotechnological silencing or knockout of these ripening-specific expressed genes would delay or reduce fruit softening, suggesting that these genes play an indispensable role in fruit ripening
[7][8][9][10]. The apple
MdPG1 was proved to be a softening-related gene that is induced by ethylene and cold treatment during fruit ripening and is specifically activated by both ethylene-insensitive3-like (EIL) and cold binding factor (CBF) transcription factors (TFs)
[11]. Likewise,
FaPG1 was demonstrated to be a key
PG gene associated with fruit softening in strawberries by the antisense-mediated reduced expression of
FaPG1, which attenuates fruit softening and extends the shelf life
[12][13]. Although a few studies about
PGs involved in fruit ripening have been investigated in kiwifruit, most of them focused on the expression of
PGs and physiological changes, such as the solubilization of pectic backbone in the cell wall and the appearance of loose cytoskeleton
[14][15]. So far, the knowledge about molecular mechanism underlying kiwifruit ripening has been scarce and demands extensive exploration.
A. arguta is a kind of newly developing kiwifruit species that is typically climacteric with rapid softening after fruit maturity, which greatly limits its commercial value. Previous studies revealed various TFs involved in the ripening processes of fruits using the comparative transcriptome analysis of
A. arguta at the commercial harvest stage and six days after harvest
[16]. However, the key candidate genes, especially the cell wall-degrading genes, have not been studied well in kiwifruit for ripening mechanism.
2. Physiological Changes in Fruit at Different Preharvest Stages
To investigate the ripening status of “TY” fruits, samples from six preharvest stages were assessed for fruit firmness, soluble solid content, and respiration intensity. The firmness of the fruits gradually decreased with ripening processes (Figure 1A), while the soluble solid content presented an upward trend (Figure 1B). In contrast, the respiration intensity showed a zigzag trend that initially decreased and then increased (Figure 1C). The changes in the above-mentioned indexes might be implicated in the dramatic changes of fruit during ripening processes.
Figure 1. Physiological indexes investigation. (A) Firmness values of fresh fruit flesh at six preharvest stages. (B) Soluble solid contents of fresh fruit flesh at six preharvest stages. (C) Respiration intensities of fresh fruit flesh at six preharvest stages. At least three readings were used to plot one point. The data represent the error bars ± SE (Standard Error) of three biological replicates.
3. Screening of Candidate Genes Based on Transcriptome Analysis
The transcriptome analysis of fruit flesh samples at three critical stages showed that large numbers of genes were strongly expressed during fruit ripening process. Three comparisons, S2 vs. S1, S3 vs. S1, and S3 vs. S2, were used for finding candidate genes. A total of 4619 differentially expressed genes (DEGs) were co-expressed among these three comparisons (Figure 2A), of which 659 DEGs were upregulated at each consecutive time point (Figure 2B). The most significant DEGs during fruit ripening were screened by setting cutoff values for transcripts at a five-fold change as the screening threshold. Finally, two transcripts, c109562_g1 and c111961_g1, were identified as potential candidate genes (Figure 2C). The expression levels of these two transcripts were significantly higher in S3 than those in S1 (Figure 2D). Both transcripts were annotated as PG on different online databases (KO, Swissprota, and MF) (Figure 2E).
Figure 2. The Venn diagrams of differentially expressed genes (DEGs) and the expression levels of transcripts. (A) Venn diagrams of all DEGs among three comparisons including S2 vs. S1, S3 vs. S1, and S3 vs. S2. (B) Venn diagrams of upregulated DEGs among three comparisons including S2 vs. S1, S3 vs. S1, and S3 vs. S2. (C) Venn diagrams of upregulated DEGs with a 5-fold change among three comparisons including S2 vs. S1, S3 vs. S1, and S3 vs. S2. (D) Expression levels of two key transcripts including c109562_g1 and c111961_g1. (E) Annotations of c109562_g1 and c111961_g1. The green font indicates c109562_g1 and c111961_g1 were both annotated as polygalacturonase in three known online databases including KO, Swissprot, and MF.
4. Identification and Classification of the PG Family in Kiwifruit
A genome-wide identification of the PG family was carried out to find the close relation of c109562_g1 and c111961_g1 with other family members. A total of 52 PG family members were excavated from kiwifruit “R5” genome. The chromosomal distribution showed these 52 PGs were irregularly located on 20 different chromosomes including chromosomes 02, 03, 05, 06, 07, 08, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 22, 23, 28, and 29 and five members were distributed in chromosomes 12, 17, and 29, while the rest of the members were found on chromosomes 11, 13, 14, 16, 20, 22, and 23. We conducted structural analysis, phylogenetic evolution, domain organization, and conserved motif analysis to find out the closely related PG family members of c109562_g1 and c111961_g1. Fifty-two PGs were further divided into six sub-families, namely A, B, C, D, E, and F. Through the sequencing and homology comparisons, we found that c109562_g1 and c111961_g1 corresponded to AaPG4 and AaPG18, respectively. Additionally, AaPG4 and AaPG18 were clustered into sub-family A. The expression levels of AaPG4 and AaPG18 in samples at six preharvest stages were assessed by qRT-PCR. The results showed that both genes were expressed during fruit development. Except for the low expression of AaPG18 at 104 d, that should be a stage at which the next maturity was prepared, following which the transcription level of AaPG18 presented an enhancing trend at 111, 118, and 125d. In general, it was noteworthy that the transcription level of AaPG18 was significantly higher than that of AaPG4 during the whole process of fruit maturity, indicating that AaPG18 might play an important role during fruit ripening (Figure 3).
Figure 3. The expression levels of AaPG4 and AaPG18 in fresh fruit flesh at six preharvest stages. Kiwifruit β-actin was selected as an internal control gene during qPCR. Values are presented as means ± SD for three replicates. The statistical significance of mean values is indicted by different letters.
5. A Membrane-Localizated AaPG18 Protein
The subcellular localization of AaPG18 protein in the plant was predicted by an online tool PSORT (
https://www.psort.org/psortb/index.html, accessed on 25 November 2021). The localization scores (cytoplasmic: 0.32; cytoplasmic membrane: 9.55; cell wall: 0.12; extracellular: 0.01) predicted its presence in the cytoplasmic membrane. To further determine this predicted result, we have analyzed the deduced amino acid sequence of PG18 in two different ways to find out the membrane-spanning domain. As expected, there were two N-terminal membrane-spanning domains in the deduced AaPG18 protein (
Figure 4A–D). To gain the direct evidence of its localization in plant cells, the constructs for subcellular localization were introduced into
N. benthamiana leaves. The GFP fluorescence signal was detected only in the cytoplasmic membrane of leaves injected with 35S:AaPG18-GFP, while it was observed in both the nuclear and membrane of leaves injected with the control (35S:GFP), which confirmed our initial prediction that AaPG18 might be a membrane-localized protein and worked in the outer layer of the cell (
Figure 4E).
Figure 4. Protein structure analysis and subcellular localization of AaPG18. (
A) Deduced protein sequence of AaPG18 with 395 amino acids. (
B) Schematic representation of AaPG18 protein structure. Two possible N-terminus transmembrane domains (orange boxes) were included. (
C) Hydropathic profile of AaPG18 protein using the predicting tool ProtScale (
https://web.expasy.org/protscale/, accessed on 25 November 2021). (
D) Prediction of transmembrane structure using TMHMM-2.0 (
http://www.cbs.dtu.dk/services/TMHMM/, accessed on 25 November 2021). (
E) Subcellular localization of AaPG18 in Nicotiana benthamiana (N. benthamiana) leaves. Cells expressing AaPG18–GFP fusion gene exhibited GFP fluorescence signals which were distributed in the cytoplasmic membrane. Cells expressing an empty vector only with a GFP tag showed GFP fluorescence signals at the nucleus and membrane. Three observed fields including brightness, GFP, and mergence were captured. The magnified field showed GFP signals in different parts of cells. White and orange arrows indicate the presence of GFP signals in the cell membrane and nucleus, respectively.
6. Transient Overexpression of AaPG18 in Strawberry
To explore the gene function of AaPG18, the transient overexpression was carried out in strawberry cultivar “snow” fruits using the method of the single fruit control (Figure 5A). The side injected with pBI121–AaPG18 showed an obvious grayish brown color in the infiltrated areas, while nothing happened on the other side of fruit injected with only pBI121 or no injection at all (Figure 5B). In order to find the reason behind this differential phenomenon at the cellular level, the infiltrated tissue sections were stained with toluidine blue O. The cross-section of the fruit side overexpression of AaPG18 presented the regular and flat status of cells (Figure 5C), while the cells from the fruit section infiltrated with the control or not infiltrated at all were irregular and loose in shape (Figure 5D,E), which indicated that the AaPG18 played a key role in cell wall stability and it helped the cell to maintain its morphology.
Figure 5. Transient overexpression in fruits of strawberry “snow”. (A) Diagrammatic sketch of the transient injection on the left and right sides of the same fruit of strawberry. (B) Phenotype observation of strawberry three days after injection. (C) Morphological observation of the cell section of strawberry fruit infiltrated with pBI121–AaPG18. (D) Morphological observation of the cell section of strawberry fruit without any injection. (E) Morphological observation of the cell section of strawberry fruit infiltrated with pBI121 (empty vector) as the control. The scale bar: 200 μm.
The homologous transformation is considered among the best ways to investigate gene function. To understand the roles of AaPG18 in kiwifruit, the transient overexpression was conducted in A. arguta “RB-4” by using the signal fruit control method. The fruit side injected with pBI121–AaPG18 turned red (Figure 6A), while nothing appeared on the side infiltrated with the empty vector (Figure 6B). The appearance of red color in “RB-4” overexpressing AaPG18 indicated its involvement in fruit maturity. Therefore, these obvious phenotypic results confirmed the key role of AaPG18 in fruit ripening and softening. The qPCR resulted showed a significantly higher expression level of AaPG18 on the fruit side infiltrated with pBI121–AaPG18 than on the opposite side infiltrated with pBI121 (Figure 6C).
Figure 6. Homologous overexpression in fruits of A. arguta “RB-4”. (A) Phenotype of the side infiltrated with pBI121–AaPG18 presenting an obvious red color. (B) Phenotype of the side infiltrated with pBI121 presenting no change. (C) Expression level of AaPG18 in “RB-4” fruits infiltrated with pBI121–AaPG18 and pBI121. Similar to strawberry fruit, the left and right sides of one fruit were injected with pBI121–AaPG18 and pBI121, respectively. Values are presented as means ± SD for three replicates. Statistical significance: *** p < 0.001.