Pollen vitality of the main pollination tree
The pollen vigor showed that the pollen germination rate of the pollination tree, was 15.56% (Figure.1 b). According to pollen’s morphological characteristics, it can be divided into two types, in which the pollen is oval, the plump pollen was called type I (Figure.1 c), and the pollen is irregular and withered, we called type II (Figure.1 d). The statistical data showed that the type I pollen accounted for 11.19% in the pollination tree (Figure.1 b). This result suggests that pollen vigor can be correlated with pollen morphology. And this lower pollen vigor of pollination trees will contribute to embryo abortion, which is a cause for abscission from sweet cherry fruitlet.
Morphology and anatomical structure of abscission fruit
According to fruitlet morphological character observations, the morphological structure of shedding fruits and carpopodiums was red and the embryos were dried out. In contrast, the retention of fruits and carpopodiums was green and the embryos were full (Figure.2a). Additionally, the weight of abscising embryos was also significantly lower than that of non-abscising embryos (Figure.2b). These results suggest that the embryo of the shedding fruit has aborted, which can trigger hormone imbalance and lead to the fruitlet being prematurely shedding. Such findings indicate that there be an excellent correlation between the shedding of sweet cherry fruitlet and embryo abortion.
Endogenous hormone analysis of abscising carpopodium
To further study the hormonal regulation of physiological fruitlet abscission in sweet cherry, endogenous levels of Auxin: Indole-3-Acetic Acid (IAA) and Indole-3-butyric acid (IBA), gibberellins (GAs): GA3, GA4, and GA7, cytokinins: trans-zeatin (TZ), abscisic acid (ABA), 1-aminocyclopropane-1-carboxylatewere (ACC), jasmonic acid (JA), and Methyl jasmonic acid (MeJA) analyzed between the abscising carpopodium and non-abscising ones in the fruitlet development stage, which precede the physiological abscission. The auxins (IAA, IBA) levels in the abscising carpopodium were reported to be low than the non-abscising carpopodium. Out of the three GAs analyzed, GA4 was detected at similar levels in both abscising carpopodium and non-abscising carpopodium, while the GA3 and GA7 levels significantly decreased in abscising carpopodium. Cytokinins (TZ) levels were found to reduce significantly in abscising carpopodium. However, the abscisic acid level was markedly increased in abscising carpopodium. Incredibly, the levels of 1-aminocyclopropane-1-carboxylatewere (ACC), jasmonic acid (JA), and Methyl jasmonic acid (MeJA) decreased significantly in abscising carpopodium. To observe the balance of plant hormones comprehensively, the ratio of (TZ + IAA + GA3) / ABA was calculated, and the results showed that the rate in abscising carpopodium was significantly lower than that of non-abscising carpopodium (Figure. 3). These results suggest that the reduced auxin may increase the sensitivity of abscission zone response ethylene, and the increased ABA content will also accelerate the abscission of sweet cherry fruitlet. However, in the abscising carpopodium, the content of GA3 and CTK related to inhibition of sweet cherry fruitlet abscission in lower. All these hormone changes may be the reasons for sweet cherry fruit shedding, while the auxin decline decrease may be the main key reason.
Transcriptome profiling and identifying differentially expressed genes (DEGs)
To obtain comprehensive and efficient transcriptome information for the sweet cherry fruitlet abscission, the transcriptomes of abscising carpopodium and non-abscising carpopodium were analyzed using the RNA-seq. Before RNA-seq analysis, six cDNA libraries were constructed and generated paired-end sequence reads using the Illumina Hiseq 4000 platform. The raw data have been deposited in NCBI Sequence Read Archive (SRA) through Gene Expression Omnibus (GEO) (access number: PRJNA636209). A sum of 3.07 billion raw reads being generated and each sample provided an average production of 51.33 million. A mean of 50.54 million clean reads was obtained from each library with an adequate read ratio of 98.48 percent after eliminating adaptor sequences, low-quality, and N-containing reads. (Additional file 1: Table S1). The reference genome exactly matched the average mapping ratio of 89.87 percent to roughly 45.41 million clean reads across each library. The Pearson correlation coefficient with all gene expression levels between every three samples was determined to investigate the gene expression correlation between samples. (Additional file 2: Figure. S1).
A total of 43,673 genes were mapped from all the samples. Abscising and non-abscising carpopodium gene expression levels were analyzed, and 6,462 differentially expressed genes (DEGs) were recognized. (Additional file 3: Table S2). Among such DEGs, the abscising carpopodium had 2,456 DEGs up-regulated and 4,006 DEGs down-regulated. (Figure.4 a). The number of related down-regulated DEGs was higher than that of up-regulated ones; It is worth pointing out that the plant hormone signal transduction pathway and the galactose metabolic pathway were significantly enriched, and these two metabolic pathways may play a regulatory role in plant organ shedding(Figure.4 b).
Enrichment analysis of DEGs duringcarpopodium abscission
Study of the KEGG pathways was also performed; according to the enrichment results, the top 20 pathways between abscising and non-abscising carpopodium were shown in Figure. 4 b and Additional file 4 Table S3. In these pathways, the plant hormone signal transduction and galactose metabolism were involved, which may regulate sweet cherry fruitlet abscission. Besides, the auxin and ethylene biosynthesis relate pathways were also enriched (Additional file 4 Table S3). Additionally, some pathways associated with cell wall modification were also augmented. These results suggested that plant hormone biosynthesis, plant hormone signal transduction, and cell wall modification play crucial roles in fruitlet abscission regulation. There were other pathways including biosynthesis of amino acids, carbon metabolism, and phenylpropanoid biosynthesis was found to be enriched.
Gene Ontology (GO) classification showed that 2,596 DEGs between the abscising and non-abscising carpopodium were graded into three categories: biological process, cellular component, and molecular function (Additional file 5 Table S4). In the biological process category, the metabolic process and the single-organism process were the most abundant terms. For the cellular component category, membrane, cell, and cell parts were the main terms. The top three terms of molecular function were bind, catalytic, and transporter activity (Figure.5).
Plant hormone biosynthesis and signal transduction
The KEGG enrichment analyses showed “plant hormone signal transduction” was a significant pathway. In the auxin signal transduction pathway, the down-regulated significantly genes include auxin influx carrier (AUX1), Auxin/Indole-3-Acetic Acid (AUX/IAA), auxin response factor (ARF), and small auxin up RNA (SAUR). In the cytokinin signal transduction pathway, the cytokinin receptor CRE1 was down-regulated. However, in the ABA signal transduction pathway, the PP2C and SnRK2 were up-regulated in abscising carpopodium. Simultaneously, in the ethylene signal transduction pathway, the ethylene insensitive 3 (EIN3) and ethylene response factor (ERF) were up-regulated which may improve the fruit abscission. Also, some essential genes are differentially expressed in the plant hormone synthesis pathways. These genes include tryptophan synthase alpha chain (α-Trp), which is related to auxin biosynthesis. Additionally, the 1-aminocyclopropane-1-carboxylatewere synthase and 1-aminocyclopropane-1-carboxylatewere oxidase were up-regulated in abscising carpopodium.
Cell wall remodeling related genes
Cell wall remolding was one of the reasons regulated abscission. Our RNA-Seq analyses of abscising carpopodium and non-abscising ones showed bidirectional changes in the expression of the cell wall remodeling genes. This phenomenon may be associated not only with the ongoing process of abscission but also with the progressive development of the organs that are not dropped (Additional file 6 Table S5). Among these cell wall remodeling-related DEGs, there were 4 cellulases (CELs), 7 polygalacturonases (PGs), 5 pectinases (PEs), 7 peroxidases (PODs), 2 beta-galactosidase (BGALs), 5 expansins (EXPs), and 5 xyloglucan endotransglucosylase/hydrolase (XTHs) was up-regulated, which may be regulated the cell wall remodeling. These results indicate that cell wall remodeling-related genes play a vital role during the fruitlet abscission. Due to the differential expression of these cell wall remodeling enzyme genes, the cell wall was remodeled, leading to the degradation of the cell wall or middle lamella, which leads to cell separation and fruitlet shedding.
Transcription Factor
Transcriptional regulation plays a pivotal role in the complex series of events leading to plant organ abscission. Therefore, transcription factors also enact an imperative role in the process. According to our data, there were 8 types of transcription factors that deserve attention, namely NAC, ERF, MYB, bZIP, WRKY, bHLH, MADS, HD-ZIP, which may regulate the shedding of sweet cherry fruitlet. Among these transcription factors, the most significant number of differential expressions was MYB, followed by WRKY and HD-ZIP (Figure.6). It is noteworthy that the genes of these three gene families are likely to regulate the shedding of plant organ abscission. Moreover, the HD-ZIP family has been shown to regulate shedding in litchi by regulating plant hormone biosynthesis genes and cell wall modification linked enzyme genes.
Cross-talk between carpopodium transcriptome and proteome
To find out more precisely the primary genes of sweet cherry fruit shedding, this study conducted a joint proteomics and transcriptomics analysis. The differentially accumulated proteins (DAPs) with 1.5-fold and differentially expressed genes with 2 folds are being used for cooperative investigation. There were 337 genes/proteins common differential expression between the transcription level and the protein level. Among these genes, there were 166 genes frequent up-regulation and 133 genes common down-regulation (Fig. 7 a). It is evident that in the plant hormone biosynthesis and signal transduction pathways, three genes a linked related to ethylene biosynthesis, 1-aminocyclopropane-1-carboxylate oxidase (ACO), were identified to be up-regulated in the abscising carpopodium; while one auxin efflux carrier (AUX1) gene was down-regulated which might involved in auxin transport. More significantly, enzymes relevant to the plant cell walls remodeling, such as cellulose, pectin acetylesterase, and polygalacturonase have been up-regulated. Likewise, peroxidase associated with lignin biosynthesis in plant cell walls is also being up-regulated. However, some tubulins related to cell wall synthesis including tubulin alpha chain (α-TUB), tubulin beta chain (β-TUB), and microtubule-associated proteins (MAPs) showed a downward trend. Excitingly, a homeobox leucine zipper transcription factor (PavHB16) was up-regulated at both transcription and protein levels (Figure.7 b).
Verification of differential expression gene by quantitative real-time PCR
To confirm findings of gene expression obtained from transcriptome data, 15 DEGs concerned to plant organ abscission were chosen for qRT-PCR. These DEGs are mainly involved in phytohormone biosynthesis, plant hormone signal transduction, plant cell wall remodeling. As shown in, the 15 DEGs had very similar expression patterns based on the transcriptome data and qRT-PCR results, which indicates the trustworthiness of the transcriptomic analysis (Additional file 7 Table S6 Figure. 8).
Identification and expression profiling of HD-ZIP gene family
According to previous studies, the HD-ZIP gene family plays an influential, role in the shedding of plant organs. To explore the relationship between HD-ZIP and sweet cherry fruitlet shedding, this study identified the HD-ZIP gene family of sweet cherry. The phylogenetic tree was constructed with 27 HD-ZIP TFs in sweet cherry plants and 55 ZmHD-ZIP TFs in Zea mays (Figure.9 a). The 27 PavHBs proteins were divided into 4 classes, namely HD-Zip I, HD-Zip II, HD-Zip III, as well as HD-Zip IV, and the proportion of each subgroup of HD-ZIP was calculated in the sweet cherry plant. HD-Zip I (ten members), which accounted for 37%, was the largest goup of PavHB TFs, followed by HD-ZIP IV and HD-ZIP II, with 26% (seven members) and 22%(six members), respectively; the smallest was HD-Zip III (four members), with just 15%. These sweet cherry HD-Zip TFs were divided into four subgroups based on the Zea mays classification (I, II, III, and IV). (Mao et al., 2016). According to the gene structure analysis, HD-Zip III and HD-Zip IV had more motifs than the other two groups, and most proteins in HD-Zip III had 15 or 16 motifs. Additionally, the same subfamily the similar motifs (Figure. 10). Among these HD-ZIPs, it is noticeable that PavHB16, which was up-regulated in the transcriptome and the proteome, belongs to the HD-ZIP I subgroup.