Effect of Pollination Methods on Fruit Development in Greenhouse Watermelon: Physiological and Molecular Perspectives

Abstract

Different pollination methods can affect the development and quality of watermelon fruit. The physiological changes in the early development of watermelon after using different pollination methods are unclear. In this study, we focused on the effects of hand pollination (H), honeybee pollination (HB), and bumblebee pollination (BB) at 1 day after pollination (1DAP) on the fruit setting rate, size, and endogenous hormone, gene, and protein expression levels using the transcriptome and proteome in greenhouse watermelon. Thus, we studied the physiological indicators of the final fruit at 40 DAP. At 1 DAP, the fruit setting rate and size of watermelon embryos showed no significant differences between the three groups. The indole-3-acetic acid (IAA) and isopentenyl adenosine (iPA) contents in the H group were highest, followed by the BB group and HB group. The abscisic acid (ABA) and gibberellin (GA3) contents were significantly higher in the BB group than in the H and HB groups. The zeatin (ZT) and carotenoid contents were lowest in the H group. The DEGs in H vs. HB and H vs. BB were mainly involved in plant hormone signal transduction, as well as amino acid and lipid metabolism. Moreover, phenylpropanoid biosynthesis and carotenoid biosynthesis were involved in H vs. HB, and carbohydrate metabolism was involved in H vs. BB. The DEGs in HB vs. BB were mainly involved in pathways including zeatin biosynthesis and photosynthesis. The DEPs in H vs. HB and HB vs. BB were involved in flavonoid biosynthesis, whereas the DEPs in H vs. BB were involved in ribosomes and oxidative phosphorylation. At 40 DAP, bee pollination can promote sugar content and transportation. Functional and pathway changes among key genes and proteins and pheromones may co-regulate plant development. This study provides data support for exploring the effects of pollination techniques on watermelon fruit development under greenhouse conditions.

1. Introduction

Watermelon (Citrullus lanatus), with 22 chromosomes and a genome size of 353.5 Mb, is a cucurbitaceous plant native to Africa [1]. Because it bears sweet fruit, has a high water content, and contains vitamins, lycopene, mineral substances, and other nutrients [2], it is very popular and ranks among the top five fruits consumed globally [3,4,5]. To meet the requirements of annual food production, greenhouse cultivation has developed rapidly [6]. Watermelons are a dioecious and allogamous crop and require insect pollination under natural conditions [7,8]. Because no pollination insects exist, fruit production and quality are very low under greenhouse conditions [6,9,10,11]. At present, bee pollination, hand pollination, and hormone pollination are commonly used for watermelon and can meet the needs of fruit setting and yield in greenhouses [10,11]. With the shortage of labor resources and increase in wages, hand pollination cannot meet the development requirements of intensive cultivation in at-scale watermelon greenhouses [7,12]. There have been previous reports of bees pollinating watermelons to improve setting and quality [13,14,15]. The working hours, visitation frequency, and stigmatic pollen deposition of honeybees and bumblebees were compared, indicating that bumblebees visit more flowers in the morning and have a higher visitation frequency and more visits, as well as higher fruit setting than honeybees [14]. Honeybees are more used for greenhouse watermelon pollination [8,16], and bumblebees are less used [8]. With the expansion of bumblebee commercialization, honeybees and bumblebees have been commonly used to pollinate greenhouse watermelons to enhance their fruit setting and quality [7,17,18,19].

Pollination and fertilization are precursors of fruit initiation and development [20,21,22,23,24,25]. These developmental processes are influenced by the environment and genetics, with pollination being a key external factor. After pollination, watermelon fruit development undergoes significant physiological and biochemical changes, including several changes in endogenous hormones, related genes and proteins, pigment accumulation, flavor, aroma volatile substance content, and maturation time [26,27]. Phytohormones, auxin, gibberellins, cytokinin, abscisic acid, and ethylene play an important role in fruit development and maturation, particularly auxin, gibberellins, and cytokinin in the early stage of fruit development [28,29,30,31,32]. Auxin, gibberellins, and cytokinin can regulate cell division and cell enlargement, resulting in improved plant fruit set and increased growth [29,30,31,32]. Abscisic acid and ethylene play an important role in ripening [33]. As the main pigments in fruit crops, carotenoids provide colors that range from yellow to red; the regulation of carotenoids in fruits can be affected by its content, the fruit tissue, and the developmental stage [34]. High-throughput sequencing technologies comprise an efficient strategy for characterizing different pollination events at the molecular level. Using high-throughput sequencing, differentially expressed genes (DEGs) involved in fruit development can be identified using different pollination methods [35]. The rapid development of bioinformatics and analytical instruments makes proteomics an important tool for examining plant development [36]. The transcriptome and proteome during fruit development (starting from 10 days after pollination) and ripening in watermelon by hand pollination can also be observed [37].

Different pollination methods also influence fruit development and thus fruit setting and quality [22,25,38,39]. At present, most studies on different pollination methods on watermelon have focused on pollination technology [40], pollination effects [41], and yield and quality [7,18,42]. Watermelon flowers bloom in the morning and can last for about 9 h, meaning that pollination time is important for fruit development [19]. However, there are few studies on the early development of watermelon. To date, there has been no systematic comparative evaluation of the effects of hand pollination and bee pollination on the early development of watermelon. The physiological changes in the early development of watermelon after using different pollination methods are unclear. In this study, we comparatively analyzed the fruit setting rate, the longitudinal and transverse diameter, and transcriptome and proteomics data at 1 day after pollination (1 DAP) by honeybees, by bumblebees, and by hand, and the weight and quality of the final fruit at 40 DAP was determined with the aim of elucidating the differences in the early development of watermelon between three pollination methods. The findings of this study will provide theoretical data for revealing the pollination mode on the mechanism of early fruit development and for determining the optimal pollination method under greenhouse conditions.

2. Materials and Methods

2.1. Plant Materials and Bee Colonies

Field work was conducted at a greenhouse watermelon base from April to June in 2020 in Dabai Village, Taigu County, Shanxi, China (east longitude 112°69″ E; north latitude 37°46″ N; altitude 827.81 m). The watermelon variety, Zhengkang No.1 (National Audit Vegetable 2002034, variety source M-4×CP-6, selected by the Zhengzhou Fruit Research Institute of the Chinese Academy of Agricultural Science), had characteristics of high yield, disease resistance, and tolerance to multiple cropping. The watermelon flowering period is typically from the middle to the end of April. Watermelon was cultivated in creeping mode, and one fruit was retained from each watermelon plant.

Honeybees (Apis mellifera ligustrica) were obtained from the JinGu beekeeping cooperative in Taigu County, Shanxi, China. Bumblebees (Bombus terrestris) were purchased from Woofun Biotechnology Co., Ltd. (Hengshui, Hebei, China).

2.2. Experimental Design

Three identical greenhouses were used for the three treatment groups, namely the hand pollination (H), honeybee pollination (HB), and bumblebee pollination (BB) groups. The size, conditions, and management of the three greenhouses arranged in parallel are basically the same. Each greenhouse was about 115 m × 8.6 m × 3.5 m, carrying 1200 watermelon plants. A honeybee colony with three combs (6000–7500 honeybees/colony) and a bumblebee colony with 80–100 bumblebees were, respectively, placed in two greenhouses in the evening 3 d before pollination. Hand pollination was performed by picking male flowers and touching female flowers from 8 to 10 a.m. daily according to watermelon flower biology.

After accounting for differences between the greenhouses and the marginal effect, three blocks with 300 plants per block were divided in each greenhouse as three replicates. During 8–10 a.m. on the same day, when honeybees and bumblebees visited a newly open flower, the flowers were pollinated. Hand pollination was performed simultaneously. The pollinated flowers were tagged and bagged. Forty identical intact flowers from each biological replicate were selected as samples. At 1 day after pollination (1DAP), 5 watermelon embryos from the 40 marked flowers from each replicate were randomly picked. Then, the transverse diameter and longitudinal diameter were determined, and the embryos were smashed into pulp, put into liquid nitrogen, and stored at −80 °C to determine the endogenous compounds (indole-3-acetic acid (IAA), abscisic acid (ABA), gibberellin (GA), zeatin (ZT), isopentenyl adenosine (iPA), and carotenoids) and to perform transcriptome and proteome sequencing analysis.

2.3. Yield-Related and Physiological Property Measurement

The fruit setting rate in each block of 30 plants was investigated 10 days after pollination. Fruit setting rate = number of fruit sets/number of flowers investigated × 100.

At 40DAP, five ripe watermelon fruits were randomly selected from each replicate to first investigate the single fruit weight, transverse diameter, and longitudinal diameter. Then, each fruit was cut to evaluate the thickness, sugar content, and number of full seeds.

The transverse and longitudinal diameters and thickness of the watermelon rinds were measured using Vernier calipers. The center and edge sugar contents of the watermelon were measured using a refractometer (PAL-1, ATAGO). The proportion of full seeds was calculated by separating full seeds from the total number of seeds. The weight per fruit was measured using an electronic balance.

Endogenous compounds, including IAA, ABA, GA3, ZT, iPA, and carotenoids, were determined using an ELISA kit according to the manufacturer’s instructions (Jiangsu Kete Biotechnology Co., Ltd, Yancheng, China).

These data are presented as mean ± SE. We analyzed these data with a one-way ANOVA and Tukey’s multiple comparison test and plotted the graphs in Graphpad Prism 5. Different lowercase letters (p < 0.05) and capital letters (p < 0.01) represent statistically significant differences between the two groups, and no letters indicate no significant difference between the two groups (p > 0.05).

2.4. RNA Extraction and Sequencing

Watermelon pulp samples preserved in liquid nitrogen were used to extract the total RNA. TRIzol reagent (Invitrogen Life Technologies, Waltham, MA, USA) was used to extract RNA according to the manufacturer’s instructions. RNA integrity was assessed using the Agilent 2100 Bioanalyzer. mRNA was broken into 200–300 bp fragments. The first cDNA strand was synthesized using reverse transcriptase, and this strand was used as the template to synthesize the second cDNA strand. An Agilent 2100 Bioanalyzer was used to determine the library quality. Illumina HiSeq was used as the sequencing platform.

2.5. Quality Control and Read Mapping

Sequences containing adapters at the 3′ end of the raw data were removed using Cutadapt. Reads with an average quality score lower than Q20 were also removed. TopHat2-upgraded HISAT2 software (http://ccb.jhu.edu/software/hisat2/index.shtml) was used to map the reads to the reference genome. The reference genome used in this study was watermelon_97103_v2.genome.fa.

2.6. Protein Digestion, iTRAQ Labeling, and LC-MS/MS Analysis

The frozen watermelon samples were ground into a powder, and SDT (4% (w/v) SDS; 100 mm Tris/HCl; pH 7.6; 0.1 mm DTT) lysis was performed for protein extraction. Protein quantification was performed by BCA. Proteins extracted from each sample were trypsinized using the filter-aided proteome preparation method, and the digested peptides were desalted using a C18 cartridge. Dissolution buffer (40 μL) was then added to the lyophilized and reconstituted peptides, and the peptides were quantified (OD280).

The 100 μg peptide samples were labeled using the AB SCIEX iTRAQ Labeling Kit according to the manufacturer’s instructions.

The labeled peptides of each group were mixed and graded by Akta Purifier 100. Each fraction was separated by HPLC system Easy NLC and then analyzed by a Q-Exactive mass spectrometer (MS/MS).

2.7. Identification of Differentially Expressed Genes (DEGs) and Differentially Expressed Proteins (DEPs) Between Groups

The DESeq2 R 3.6.0 software package was used to identify DEGs [43]. Genes with a p-value < 0.05, |log2FoldChange| > 1, and FDR < 0.05 were considered to be DEGs.

Protein identification and quantitative analysis were carried out using Mascot 2.2 and Proteome Discoverer 1.4 software. The definition of upregulated differentially expressed proteins (DEPs) in this study is FDR ≤ 0.01, p value < 0.05, and |log2FoldChange| > 1.2, whereas those with |log2FoldChange| < 0.83 were considered as downregulated proteins.

2.8. Enrichment Analysis of DEGs and DEPs

A Gene Ontology (GO) enrichment analysis of DEGs was implemented using the GOseq R package. GO terms with a p-value < 0.05 were considered significantly enriched by differentially expressed genes. GO enrichment results were evaluated in terms of molecular function (MF), biological process (BP), and cell component (CC). KEGG Automatic Annotation Server (https://www.genome.jp/tools/kaas/) was used to test the statistical enrichment of differentially expressed genes in Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways.

The enrichment analysis of the GO and KEGG pathways of a target protein set was carried out using Omicsbean (http://www.Omicsbean.cn/). GO and KEGG terms with a p-value < 0.05 were considered significantly enriched by DEPs. The bubble diagram of KEGG enrichment results was generated by software R version 3.5.1.GO, and KEGG enrichment analyses were also performed for DEPs in each comparison group using Fisher’s exact test.

3. Results

3.1. Yield-Related and Physiological Characteristics

The fruit setting rates of the H, HB, and BB groups are shown in Figure 1A, being 75.36% ± 5.29, 81.07% ± 3.78, and 78.57% ± 3.93, respectively. There was no significant difference between the three groups (p = 0.6954). There were no significant differences between the three groups in the transverse diameter (p = 0.3400) and longitudinal diameter (p = 0.2007) at 1 DAP (Figure 1B).

At 1 DAP, six endogenous compounds (IAA, ABA, GA, ZT, iPA, and carotenoids) of watermelon were detected (Figure 2). The IAA content of watermelon in the H group was 18.68 ± 0.09 μg/g, the highest out of the three groups, followed by the BB group at 17.48 ± 0.12 μg/g and the HB group at 15.42 ± 0.07 μg/g. There were significant differences between the three groups (p < 0.0001) (Figure 2A). In terms of ABA content, the values were 50.99 ± 0.26 μg/g, 39.24 ± 0.47 μg/g, and 54.15 ± 0.47 μg/g in the H, HB, and BB groups, respectively. There was also a significant difference between H and HB, HB and BB (p < 0.0001), and H and BB (p < 0.01) (Figure 2B). The GA content of watermelon was significantly higher in the BB group (81.62 ± 0.45 ng/g), lower in the H group (47.95 ± 0.30 ng/g), and lowest in the HB group (45.13 ± 0.82 ng/g) (p < 0.0001) (Figure 2C). The ZT content in the HB group (61.41 ± 0.48 ng/g) was the highest, followed by that in the BB group (53.54 ± 0.20 ng/g) and H group (37.35 ± 0.44 ng/g), with significant differences observed among the three groups (p < 0.0001) (Figure 2D).The iPA content of watermelon was highest in the H group (9.03 ± 0.03 μg/g), lower in the BB group (8.32 ± 0.04 μg/g), and lowest in the HB group (7.08 ± 0.04 μg/g), and a significant difference was found among the three groups (p < 0.0001) (Figure 2E). The carotenoid content in the BB group (40.50 ± 0.27 ng/g) was the highest, followed by that in the HB group (36.71 ± 0.12 ng/g) and H group (32.84 ± 0.21 ng/g), with significant differences observed among the three groups (p < 0.0001) (Figure 2F).

At 40DAP, the single fruit weight, the transverse and longitudinal diameter of the whole watermelon and the flesh, rind thickness, center and edge sugar content, and the proportion of full seeds were measured in the H, HB, and BB groups (Figure 3). The transverse diameter of the whole watermelon in the HB (185.8 ± 1.33 mm) and BB (188.6 ± 2.16 mm) groups was significantly higher than those in the H group (171.9 ± 2.95 mm) (p = 0.0004) (Figure 3B). The rind thickness of watermelons pollinated by honeybees (15.28 ± 1.16 mm) and bumblebees (15.76 ± 1.47 mm) was 1.68 times and 1.74 times that of watermelons pollinated by hand (9.08 ± 0.34 mm) (p = 0.0016) (Figure 3C). There was no significant difference in HB vs. BB (p = 0.0512). The center sugar content of watermelons pollinated by honeybees (13.28 ± 0.15%) and bumblebees (13.46 ± 0.23%) increased by 14.68% and 16.23%, respectively, compared with those pollinated by hand (11.58 ± 0.57%) (p = 0.0058) (Figure 3F). The other six parameters of single fruit weight, the transverse and longitudinal diameter of the whole watermelon and the flesh, the edge sugar content, and the proportion of full seeds showed no significant differences between the three groups (p > 0.05) (Figure 3A,B,D–F).

3.2. Differentially Expressed Genes Enrichment

Nine cDNA libraries were constructed and analyzed (Supplementary File Table S1). The differentially expressed genes (DEGs) identified between the groups are shown in Figure 4 (Supplementary File Table S2). In total, 280 DEGs (143 upregulated DEGs and 137 downregulated DEGs), 324 DEGs (184 upregulated DEGs and 140 downregulated DEGs), and 136 DEGs (76 upregulated and 60 downregulated) were identified in H vs. HB, H vs. BB, and HB vs. BB, respectively. Gene annotation identified the differential expression between the three groups, including gene-encoding transcription factors (TFs), zinc finger proteins (ZFPs), cytochrome P450, heat shock proteins (HSPs), and so on (Supplementary File Table S3).

GO items with the smallest p-values (top 10) in each GO category are displayed (Figure 5A–C). The DEGs between the H and HB groups were involved in functions such as transcriptional regulation and macromolecular biosynthesis regulation (Figure 5A). The DEGs between groups H and BB were involved in plastid structure, oxidoreductase activity, and carbohydrate metabolism processes (Figure 5B). The DEGs between the HB and BB groups were mainly involved in chloroplast, plastid, and oxidoreductase processes, as well as ion binding, enzymatic activity, and photosynthesis (Figure 5C).

For all groups, KEGG results showed that most DEGs were enriched in metabolic pathways, including amino acid metabolism, lipid metabolism, carbohydrate metabolism, energy metabolism, terpenoid and polyketide metabolism, etc. (Figure 5D–F, Supplementary File Table S4). These DEGs in H vs. HB and H vs. BB were mainly enriched in five pathways (environmental information processing, metabolism, genetic information processing, organismal systems, cellular processes), yet were only enriched in three pathways in HB vs. BB (environmental information processing, metabolism, genetic information processing) (Figure 5D–F). The KEGG pathways involved in the DEGs between the H and HB groups included phenylpropanoid biosynthesis, plant hormone signal transduction, amino acid and lipid metabolism, isoflavonoid biosynthesis, and carotenoid biosynthesis (Figure 5D). The pathways involved in the DEGs between the H and BB groups included carbohydrate metabolism, plant hormone signal transduction, amino acid and lipid metabolism, protein processing in the endoplasmic reticulum, and caffeine metabolism (Figure 5E). The DEGs between the HB and BB groups were mainly involved in pathways including zeatin biosynthesis, glucosinolate biosynthesis, photosynthesis, amino acid metabolism, and monoterpenoid biosynthesis (Figure 5F).

3.3. Differentially Expressed Proteins Enrichment

A total of 10,110 proteins were identified using ITRAQ-labeled quantitative proteomics (Supplementary File Table S5). The results of the DEPs between the different groups are shown in Figure 6. In total, 380 differentially expressed proteins (DEPs) (275 upregulated and 105 downregulated), 260 DEPs (239 upregulated and 21 downregulated), and 444 DEPs (134 upregulated and 310 downregulated) were screened in H vs. HB, H vs. BB, and HB vs. BB, respectively.

The GO terms associated with the identified DEPs between groups were analyzed (Figure 7A–C). In H vs. BB, most DEPs were enriched in biological process (BP), followed by cell component (CC) and molecular function (MF). Among all the groups, in terms of the number and percentage of DEPs, more proteins were distributed in CC. The DEPs between the H and HB groups were mainly concentrated in cell, intracellular part, cellular component organization or biogenesis, and protein and RNA binding (Figure 7A). The DEPs between the H and BB groups were mostly involved in macromolecular complex assembly and binding (Figure 7B). The DEPs screened between the HB and BB groups were involved in cell, intracellular part, and cellular component organization or biogenesis, as well as macromolecular complex binding (Figure 7C).

KEGG analysis revealed that the DEPs between the H and HB groups were significantly enriched in flavonoid biosynthesis, RNA transport, and linoleic acid metabolism (Figure 7D). The DEPs between the H and BB groups were mainly involved in ribosomes, oxidative phosphorylation, and photosynthesis (Figure 7E). Flavonoid biosynthesis, RNA transport, and ribosomes were the main pathways involved in DEPs between the HB and BB groups (Figure 7F).

3.4. Association Analysis of DEGs and DEPs

Based on the analysis of quantitative differences between the proteome and transcript, the Pearson correlation coefficients were −0.0135, 0.1967, and −0.0385 in H vs. HB, H vs. BB, and HB vs. BB, suggesting that the correlations of mRNA and protein are low (Supplementary File Figures S1–S3).

Based on the enrichment analysis results for the proteome and transcripts, the top 10 pathways in the set analysis were plotted using a histogram (Figure 8). In H vs. HB, there was one shared upregulated gene (histone H1, Cla97C01G019560) on mRNA and protein, which was involved in flavonoid biosynthesis and the spliceosome (Supplementary File Figure S4, Figure 8A). In H vs. BB, heat shock 70 kDa protein 4 (Cla97C09G181360), chaperone protein (Cla97C04G078780), and 18.5 kDa class I heat shock protein-like (Cla97C03G051930) were downregulated at the mRNA level but upregulated at the protein level, and they were involved in linoleic acid metabolism; protein processing in the endoplasmic reticulum; and valine, leucine, and isoleucine biosynthesis (Supplementary File Figure S5, Figure 8B). In HB vs. BB, there was one shared downregulated gene (Cla97C11G214070), which was involved in flavonoid biosynthesis regulation, oxidative phosphorylation, and ribosomes at both the mRNA and protein levels (Supplementary File Figure S6, Figure 8C).

4. Discussion

Adequate pollination services positively affect watermelon fruit development [44]. In this study, we compared physiological indicators, endogenous hormones, and the polyomics of watermelon 1 day after pollination by honeybees, by bumblebees, and by hand (1 DAP). We further studied the effects of three pollination methods on the weight and quality of mature fruits (40 DAP). At 1 DAP, the transverse diameter and longitudinal diameter of watermelon embryos showed no significant difference between the three groups. Bee pollination may accelerate flower fertilization, subsequent ovary development, and the early expansion of passion fruit [39]. The length, width, and thickness of the ovary demonstrated a minimal increase between 0 and 2 days after anthesis in Gleditsia sinensis, and some pollen tubes reached the ovules within the ovary by 24 h post-pollination [45]. In Arabidopsis thaliana and Brassica napus, cell division and cell expansion occur within the first 24 h post-pollination [20,21] and 18–24 h [25]. The melon fruit after artificial pollination and dipping flower treatment became larger than the unpollinated fruit, and the difference was very obvious 3 days after flowering [38]. The ovary size, namely cell division and cell expansion, was increased, but there was no significant change in fruit size by 24 h post-pollination, which is consistent with our results.

Plant hormones regulate plant development at all stages. The hormone levels at 1 DAP showed that the abscisic acid (ABA) and gibberellin (GA3) content of the watermelon was significantly higher in the BB group than in the H and HB groups. The content of IAA and GA was higher in the early stage of fruit development [28]. ABA has a negative effect on the initial developmental stages [46,47,48]. The fruit setting rate after bumblebee pollination was lower than that after honeybee pollination. As members of the auxin and cytokinin family, IAA and iPA levels regulate cell division and cell enlargement, resulting in increased growth [30,31]. IAA and iPA content was significantly higher in the H group than in the BB and HB groups. Indole-3-acetic acid-amido synthetase (Cla97C07G138770 and Cla97C10G190160) associated with IAA was downregulated both in H vs. BB, H vs. HB, and HB vs. BB. This is consistent with an upward adjustment. These results corroborate each other. IAA plays an important role in enhancing cell division and enlargement in the early phases of grape development [49]. GA and ABA content was significantly higher in the BB group than in the H and HB groups. The zeatin content in HB at 1DAP was the highest among all the groups, followed by the BB group and H group. Zeatin and iPA belong to the cytokinin family and can promote cell division. DEGs between the BB and HB groups may be involved in zeatin biosynthesis. In the DEPs, CKX5, IPT9, and UGT85A1 were related to zeatin biosynthesis. Carotenoids can affect the color of watermelon flesh, ranging from white to yellow and red [50,51], and they affect abscisic acid and strigolactone biosynthesis as precursors [34,52,53,54]. The carotenoid content of watermelon at 1 DAP was the highest in the BB group, followed by the HB and H groups. Bee pollination increased the carotenoid content during early development. DEGs between the H and HB groups may be involved in carotenoid biosynthesis. GO terms showed that the DEGs were significantly enriched in plastids, apoplasts, and chloroplasts in H vs. BB and HB vs. BB, were enriched in apoplasts in H vs. HB, and may be involved in carotenoid biosynthesis in H vs. HB. In plants, carotenoids were synthesized and accumulated within plastids and produced via the 2-C-methyl-D-erythritol 4-phosphate pathway [34,55]. These data indicated the pollination process was completed at 1 DAP; there were complex changes and balances in the endogenous hormones, genes, and proteins in the watermelon fruit, but there were no significant difference in fruit size, indicating that the fruit was still in the process of nutrient and energy accumulation during this period.

Pathway enrichment analysis showed that DEGs between groups H and HB, as well as BB, were both involved in amino acid and lipid metabolism, and the carbohydrate metabolism pathway was upregulated. Both the pollen-grain wall and stigma secretion contained proteins, carbohydrates, acidic polysaccharides, lipids, and phenolics in early fruit development. The ovule becomes a significant sink for resources, as sucrose and starch are accumulated by the endosperm, and the embryo begins to develop [56,57]. The enrichment pathways of the DEPs in the different groups were also investigated. These results revealed many potential pathways that may affect watermelon fruit development. The DEPs in H vs. HB and HB vs. BB were all involved in flavonoid biosynthesis. As factors affecting secondary metabolism, flavonoids originate from the general phenylpropanoid pathway and regulate multiple plant developmental processes [58,59]. The DEPs between the H and BB groups are involved in oxidative phosphorylation, photosynthesis, and brassinolide synthesis. Photosynthesis is a process by which plants use light energy to convert carbon dioxide and water into organic matter, such as sugar and starch, while releasing oxygen, resulting in sugar accumulation [60]. Brassinolides synergistically regulate fruit ripening via auxins. Brassinolide can also synergize with abscisic acid to increase soluble solid content and form anthocyanins in grapefruit [49]. The Pearson correlation analysis of DEGs and DEPs suggests that the correlations of mRNA and protein are low. The same phenomenon occurs during watermelon development [30]. It was seen that both of them regulate the expression of whole fruit development-related genes. These DEGs, DEPs, and pheromones co-regulate watermelon development [51].

Pollination treatments affect fruit fertilization and development, thereby affecting the final fruit size and quality [22,25,39]. There was no significant difference among the three groups in terms of the final fruit weight and size. It has been verified that bee pollination and hand pollination can meet the needs of watermelon production in greenhouses, and bee pollination can replace hand pollination to save labor costs [7,8,22]. The number of bee visits and the number of stigmatic pollen depositions can affect fruit set and pollination efficacy [14]. In this study, all the samples were fruits taken after the flowers were visited once by bees; one male flower was used to pollinate one female flower by hand. The amount of pollen was in the stigma in the H group, and the stigmatic pollen deposition was not measured, which may result in no significant difference in the fruit size-related indicators and fruit set between hand- and bee-pollinated fruits. In production, hand pollination mostly uses one male flower to pollinate multiple female flowers, with poor results. Bee pollination affects the fruit set and yield and also enhances fruit quality [61,62,63]. At 40 DAP, the sugar content in watermelons that received bee pollination was higher than that in the H group; in particular, the center sugar content was significantly higher. This may be related to the fact that bee pollination can carry more allogeneic pollen [39]. This result is consistent with that of a previous study [64]. Sucrose synthase, a sucrose-metabolizing enzyme, plays an important role in determining the sugar composition of cucurbit fruit [65]. Sugar content was not significantly different between the HB and BB groups, which may be related to the similarity of the two groups in the sugar metabolic pathway. In addition to the internal sugar content, the rind thickness of the watermelons was significantly higher in the HB and BB groups than in the H group. The rind undergoes intense meristem activity during thickening [66]. A previous study reported that watermelon rind thickness was significantly negatively correlated with rind toughness [67]. Another study reported that thicker rinds could reduce damage during transportation to ensure fruit integrity [68]. Related studies have reported that rind thickness is largely determined by genes; however, studies on watermelon grafting have shown that grafting affects rind thickness [69].

5. Conclusions

This study delineated the physiological and phytohormone changes during watermelon development under three pollination methods. Integrating transcriptomic and proteomic analyses, we identified distinct molecular pathways and regulatory networks associated with each pollination type. DEGs in H vs. HB mainly involved phenylpropanoid biosynthesis, plant hormone signal transduction, amino acid and lipid metabolism, and carotenoid biosynthesis. The DEGs between the H and BB groups mainly involved carbohydrate metabolism, plant hormone signal transduction, and amino acid and lipid metabolism. The DEPs in H vs. HB and HB vs. BB were all involved in flavonoid biosynthesis. The DEPs between the H and BB groups were involved in oxidative phosphorylation and photosynthesis. DEGs and DEPs between the BB and HB groups were involved in zeatin biosynthesis. The zeatin content in HB and BB at 1DAP was higher than that in the H group and was the highest in the HB group. These DEGs and DEPs in the three groups mainly participate in regulating hormone synthesis, carbohydrate metabolism, and nutrient metabolism. Genes, proteins, and hormones co-regulate the development of the fruit. Notably, the differences in endogenous hormones and genes and proteins of the early fruit under three pollination methods affected the quality of the final watermelon fruit, resulting in fruits that were sweeter and more transportable after bee pollination than after hand pollination. These findings elucidate the active complex regulatory networks during early fruit development and provide a valuable resource for understanding how pollination patterns influence watermelon fruit set, growth, and quality.