Abstract
This study aims to isolate biocontrol microorganisms that inhibit Penicillium crustosum and explore their effects on microbial control and microbial diversity changes in P. crustosum causing postharvest decay in Orah.
The biocontrol effect was verified by confrontation experiments. The microbial diversity was analyzed using high-throughput sequencing technology. Metabolomic analysis was performed using liquid chromatography-tandem mass spectrometry (LC-MS/MS).
A strain of Meyerozyma guilliermondii PJ15 that strongly inhibited P. crustosum was isolated from the grape surface, which could effectively inhibit the mycelial growth and green spore production of P. crustosum. PJ15 could reduce the sourness and nutrient loss of Orah caused by P. crustosum. Compared with the control, PJ15 treatment increased the pH by 11.32%, the soluble protein content by 163.64%, the ascorbic acid content by 160.84%, and the total soluble sugar content by 203.53%. Microbial diversity analysis showed that PJ15 has a relatively small effect on the bacterial composition and diversity on the surface of Orah but has a significant impact on the fungal composition and diversity. It can increase the fungal diversity of Orah invaded by P. crustosum, increase the density of fungal interaction networks, and form a stronger coexisting survival interaction network. The vaccination of PJ15 downregulated the differential metabolite 5,6-epoxytetraene, which showed a positive correlation with P. crustosum and a negative correlation with PJ15.
Meyerozyma guilliermondii PJ15 has a strong inhibitory effect on the growth of P. crustosum and has a significant impact on the fungal composition and diversity on the surface of Orah. It has potential for biocontrol of P. crustosum causing postharvest decay in Orah.
Introduction
Orah is a late-maturing citrus variety obtained by cross-breeding ‘Temple’ orange and ‘Dancy’ red orange. The variety is widely cultivated in Southwest China, with an annual output value of USD 1.4 billion, bringing considerable income to farmers and high-quality fruit to consumers (Huang et al., 2023). Orah is vulnerable to mechanical damage in the process of harvesting and sales and is vulnerable to various pathogenic microorganisms, which cause decay, especially fungal infection. Postharvest decay in the process of storage and sales has caused serious economic losses. The Food and Agricultural Organization (FAO) estimates that 21.6% of the world’s fruits and vegetables are lost during storage and distribution (FAO, 2023).
Postharvest decay of citrus is mainly caused by Penicillium infection, including Penicillium digitatum, Penicillium italiana, and Penicillium extensus (Moraes et al., 2019; Wang et al., 2022). Pathogenic fungi colonize fruit and produce toxic metabolites (such as mycotoxins), which can harm human health (Sanzani et al., 2016; Leyva et al., 2017; Ráduly et al., 2019). In view of these pathogenic fungi, a lot of prevention and control measures have been carried out and have achieved certain results, with the common main methods being physical control, chemical control, and biocontrol. However, the equipment for physical control is expensive, has high energy consumption and high cost, and it is difficult to be widely applied. Chemical control is easy to cause residual chemical substances to threaten human health, and the long-term use of chemical reagents can also cause microbial resistance (Fisher et al., 2018; Deng et al., 2020). In addition, the impact of synthetic chemicals on food safety and environmental protection has caused public concern, and more eco-friendly control methods are the trend of future development. Biocontrol is mainly to prolong the preservation of fruit by antagonizing pathogenic fungi with the advantages of being safe, highly efficient, nontoxic, and easy to degrade. The main mechanism is to inhibit or kill pathogenic fungi that cause fruit decay. Biocontrol is an effective way to realize the pollution-free storage of fruit, and it is also one of the development directions to develop new postharvest biocontrol technologies for fruit, so it has broad prospects in the field of fruit storage. Many studies have shown that probiotics, such as lactic acid bacteria and yeast, can effectively inhibit the growth of mold and the production of mycotoxins and have potential as biocontrol agents (Mewa et al., 2019; Abouloifa et al., 2020; Leneveu et al., 2020; Seshadri et al., 2020; De Simone et al., 2021; Tejero et al., 2021).
When biocontrol microorganisms are used to control postharvest decay of fruit, they are essentially introduced into a complex microbial community, and previous studies have mainly focused on the interactions between biocontrol microorganisms and pathogenic microorganisms while ignoring other microorganisms associated with fruit (Spadaro and Droby, 2016; Zhao et al., 2023). The structure and composition of postharvest microbiota are considered key factors in the prevention of postharvest decay and have been studied in a variety of fruits, including strawberry (Cruz et al., 2018; Zou et al., 2021), cherry tomato (Liu et al., 2020), apple (Abdelfattah et al., 2016; Angeli et al., 2019), and kiwifruit (Zhao et al., 2023). A better understanding of the interactions between microbial communities could provide additional opportunities for the development of innovative biocontrol methods to inhibit postharvest decay in fruit (Sébastien et al., 2015). Previous work has isolated and identified P. crustosum from decayed Orah and demonstrated that it causes postharvest decay in Orah. In this study, a strain of Meyerozyma guilliermondii PJ15 antagonistic to P. crustosum was isolated from the grape surface, and the biocontrol of P. crustosum was studied on the plate and Orah. The effects of M. guilliermondii PJ15 on microbial diversity and metabolites in Orah were analyzed using multiomics techniques to understand the mechanism by which biocontrol microorganisms regulate the microbial community and control decay in Orah. This study provides a new idea for the further development of effective methods to control postharvest decay in Orah.
Materials and Methods
Isolation and identification of M. guilliermondii PJ15
Fruit samples such as grapes, Orah, orange, mango, and peach from the farmers’ market near the laboratory were collected, the surface was cut into a mortar, and an appropriate amount of sterile water was poured for grinding so that the microorganisms on the surface of the fruit dissolved in the sterile water. Then, the corresponding gradients 10−1, 10−2, and 10−3 were diluted and 100 μL of each on a Bengal red plate was collected. The mixture was incubated at 30 °C until single colonies grew. Spore suspension of 50 μL P. crustosum (1×106 spores/mL) was inoculated in the center of the potato dextrose agar (PDA) plate. The single bacteria on the primary screen plate were inoculated with sterile toothpicks in four directions perpendicular to each other around the decayed bacteria, approximately 3 cm away from P. crustosum. They were incubated at 30 °C. Yeast with antagonistic effects on P. crustosum was screened by plate confrontation experiments. The 18S sequence was amplified and sequenced by polymerase chain reaction (PCR) and compared with the National Center for Biotechnology Information (NCBI) database to identify biocontrol microorganisms. The phylogenetic tree of PJ15 was analyzed and drawn using MEGA 11 software (MegaSoft Computers, Dindigul, India). The 18S sequence obtained from sequencing was uploaded to NCBI for homologous sequence retrieval, obtaining relevant information on the known yeast with similar genetic relationships. ClusterW alignment analysis was performed between the 18S sequence of PJ15 and the known sequence obtained from the NCBI. The neighbor-joining method was used to analyze the genetic relationship of PJ15 and obtain a phylogenetic tree.
Inhibition of P. crustosum by biocontrol microorganism PJ15 on Orah
The hole punch was used to make 4-mm holes around the equator of the Orah fruit, and 50 μL P. crustosum spore suspension (1×106 spores/mL) was added to each hole, with four holes per fruit. In one of the holes, 50 μL sterile water was added as a control (CK), and in the other three holes, 50 μL yeast suspension (1×106 cells/mL) was added each as a biological control group (T). Each experiment was repeated ten times. The samples were placed in an aseptic plastic bag, tied, sealed, and incubated at 30 °C. The rot diameter of the Orah fruit at 0, 7, 14, and 21 d was measured by vernier calipers. After 21 d, the peels from the wound of the control and biocontrol microorganism PJ15 were collected in 2-mL sterile EP tubes for the analysis of microbial diversity and metabolomics. Then, the peel was removed and the juice was extracted. pH, soluble protein, ascorbic acid, and soluble total sugar were determined. pH was determined by the electrode method. Soluble protein was determined by the Coomassie brilliant blue method. Ascorbic acid was determined by the 2,6-dichloro-indophenol method. Total soluble sugar was determined by the anthrone colorimetric method.
Analysis of microbial diversity in Orah
For the detection of microbial diversity, the method of Miya et al. (2015) was used. Total DNA extraction was carried out by the cetyltrimethylammonium bromide (CTAB) method. Amplicon sequencing was performed on an Illumina NovaSeq 6000 (Illumina, San Diego, CA, USA) to obtain raw reads. Non-chimeric reads were obtained by filtering, identification and removal of primer sequences, noise reduction, and removal of chimeric sequences. Using SILVA (Leibniz Institute DSMZ-German Collection of Microorganisms and Cell Cultures GmbH, Braunschweig, Germany) as the reference database, a naive Bayes classifier was used to make taxonomic annotations for the feature sequences, and the community composition of each sample was determined at the genus level. QIIME2 software (University of California San Diego, San Diego, CA, USA) was used to generate species abundance tables at different taxonomic levels, and the R language tool was used to draw community structure maps of the samples at different taxonomic levels. QIIME2 software was used to evaluate and analyze the alpha diversity index of the samples. The microbial correlation network analysis was performed according to the abundance and variation of each species in each sample, Spearman rank correlation analysis was performed, and data with correlations greater than 0.1 and P-value less than 0.05 were screened to construct correlation network. The samples used for microbial diversity analysis were all constructed and sequenced on the Illumina sequencing platform of Biomarker Technologies (Beijing, China). Data analysis of microbial diversity was performed on the BMKCloud platform of Biomarker Technologies.
Analysis of metabolomics in Orah
Metabolomics detection was carried out according to Zhu et al. (2022). The metabolites were detected by Waters ACQUITY UPLC I-Class PLUS with a Xevo G2-XS QTOF high-resolution mass spectrometer system (Waters Corporation, Milford, CT, USA). The column used was the Waters ACQUITY UPLC HSS T3 Column (1.8 μm, 2.1 mm×100 mm; Waters Corporation). The mobile phases of the ultra performance liquid chromatography (UPLC) were aqueous solution with 0.1% (volume fraction) formic acid (A) and 0.1% (volume fraction) formic acid in acetonitrile (B). The injection volume of the UPLC was 1 μL. The primary and secondary mass spectrometry (MS) data were collected in MSE (E represents collision energy) mode under the control of MassLynx 4.2 (Waters Corporation). In each data acquisition cycle, dual-channel data acquisition can be carried out for both low- and high-collision energy. The low-impact energy was 2 V, the high-impact energy ranged from 10 to 40 V, and the mass spectrum scan frequency was 0.2 s. Electrospray ionization (ESI) ion source parameters were as follows: capillary voltage, 2000 V (positive ion mode) or −1500 V (negative ion mode); cone voltage, 30 V; ion source temperature, 150 °C; desolvent gas temperature, 500 °C; backwash gas flow, 50 L/h; and dissolving gas flow, 800 L/h. The original data collected by MassLynx 4.2 were processed by Progenesis QI software (Waters Corporation) for peak extraction and peak-to-peak data processing operations. The identification was carried out based on the METLIN database of Progenesis QI software and the self-built biomarker database. At the same time, the theoretical fragment identification and quality deviation were within 100×10–6. After the original peak area information was normalized with the total peak area information, the identified compounds were searched in the Kyoto Encyclopedia of Genes and Genomes (KEGG), Human Metabolome Database (HMDB), and lipid map databases to obtain classification and pathway information. Differential metabolites were screened using the combination of difference multiple, P-value, and variable importance in projection (VIP) value of the orthogonal partial least squares discriminant analysis (OPLS-DA). The screening criteria were fold change (FC)>1, P-value<0.05, and VIP>1. A hypergeometric distribution test was used to calculate different metabolites of enrichment significance by the KEGG pathway. The metabolites were detected via the Waters platform of Biomarker Technologies. Metabolomics analysis was performed on the BMKCloud platform of Biomarker Technologies.
Statistical analysis
All the parameters were analyzed using a one-way analysis of variance (ANOVA), and Duncan multiple range tests were performed for multiple comparisons using SPSS Statistics 26.0 (IBM, Amherst, MA, USA).
Results
Screening of biocontrol microorganism of P. crustosum causing postharvest decay in Orah
A total of 1000 single colonies were obtained from various fruit samples, and thirty strains of yeast and nineteen strains of Bacillus that may have an inhibitory effect on P. crustosum were obtained by primary screening. After further screening by a flat panel confrontation experiment, six strains of yeasts with significant inhibitory effects on P. crustosum were isolated from the grape surface and identified by the 18S DNA sequence as M. guilliermondii. A yeast with the strongest inhibitory effect on P. crustosum was selected as the biocontrol microorganism, named PJ15. On the yeast extract peptone dextrose (YPD) plate, the colony morphology of PJ15 is cellular, oval, or elliptical, with uniform texture and a creamy white cheese-like appearance (Figure 1A). The surface is smooth or wrinkled, and the edges are etched or tassel like, with false hyphae (Figure 1A). Under the microscope, the cells of PJ15 are spherical, elliptical, and elongated, occasionally conical but not forming a pointed tip, with some having conidia (Figure 1B). The growth curve of PJ15 is similar to that of other microorganisms and can be roughly divided into four stages (Figure 1C). During the lag phase (0–4 h), PJ15 is newly inoculated into a fresh YPD medium and cultured in a new growth environment, with slow division and growth, and in a state of adaptation to the environment. During the logarithmic growth phase (4–24 h), PJ15 begins to grow rapidly, and the number of cells increases exponentially. During the stationary phase (24–32 h), the growth of PJ15 tends to be balanced, and after a short period of growth after 28 h, it enters stability again. During the decay period (32 h later), the number of PJ15 cells begins to decrease (Figure 1C). It can be intuitively seen from the phylogenetic tree that PJ15 has the closest genetic relationship with M. guilliermondii (Figure 1D).
Colony morphology and growth curve of PJ15. (A) colony morphology of PJ15 on the YPD, (B) morphology of PJ15 under the microscope, (C) growth curve of PJ15 on the YPD, and (D) phylogenetic tree of PJ15 was constructed based on the 18S gene sequence.
Inhibition of P. crustosum by biocontrol microorganism PJ15 on the plate
PJ15 and P. crustosum were inoculated at the same time and cultured under the same conditions. After 7 d, a significant inhibitory effect appears. PJ15 has a strong antagonistic effect on P. crustosum on the plate, which could effectively inhibit the growth of P. crustosum mycelium and the production of green spores. The mycelial growth of P. crustosum cultured against PJ15 was inhibited, and the mycelia growth near PJ15 was almost abnormal. The mycelial growth far from PJ15 was sparse and irregular. From a distance, it seems that the colony of P. crustosum was bitten by PJ15, and the overall appearance of the colony was fluffy white hyphae with fewer green spores (Figure 2A). In the control, the mycelium of P. crustosum grew vigorously and luxuriantly, expanding and growing around the nearly circular circumference, forming a layer of thick green spores (Figure 2B). Under an optical microscope, it was found that the mycelium of PJ15 appeared to have protoplast agglutination at the edge of the colony after culture treatment. The mycelium grew sparsely, tangled, deformed, irregularly twisted, and weak, similar to the chaotic state of upward-growing rice seedlings in a paddy field after being hit by a hurricane (Figure 2C). However, the mycelium of the control P. crustosum grew densely and was thicker, with stronger vitality, and showed uniform and regular outward growth overall (Figure 2D). Furthermore, plate pairing tests were conducted, and the results showed that there was no overlap between the colonies of PJ15 and P. crustosum. The growth of pathogenic hyphae closer to PJ15 was significantly inhibited, with the strongest antibacterial efficacy in the middle and weaker inhibitions at both ends. The growth of pathogenic hyphae further away from PJ15 was basically not inhibited (Figures 2E and 2F). After 10 d of cultivation on the plate, PJ15 showed an antibacterial effect of 23.59% against pathogens compared with the diameter of pathogen diffusion in the control group (Table 1). PJ15 has strong antibacterial efficacy against P. crustosum.
Antibacterial effect of the plate pairing test
| Group | Diameter of pathogen diffusion (mm) | Antibacterial efficacy (%) |
|---|---|---|
| CK | 42.10±0.83a | − |
| T | 32.17±0.67b | 23.59% |
| Group | Diameter of pathogen diffusion (mm) | Antibacterial efficacy (%) |
|---|---|---|
| CK | 42.10±0.83a | − |
| T | 32.17±0.67b | 23.59% |
The results are expressed as mean±standard deviation. Different lowercase letters in the same column indicate significant difference between the two groups.
Antibacterial effect of the plate pairing test
| Group | Diameter of pathogen diffusion (mm) | Antibacterial efficacy (%) |
|---|---|---|
| CK | 42.10±0.83a | − |
| T | 32.17±0.67b | 23.59% |
| Group | Diameter of pathogen diffusion (mm) | Antibacterial efficacy (%) |
|---|---|---|
| CK | 42.10±0.83a | − |
| T | 32.17±0.67b | 23.59% |
The results are expressed as mean±standard deviation. Different lowercase letters in the same column indicate significant difference between the two groups.
Inhibition of Penicillium crustosum by PJ15. (A) Inhibition of P. crustosum by PJ15 on the plate, (B) P. crustosum on the plate, (C) growth of mycelium at the confrontation between P. crustosum and PJ15 under the microscope, (D) mycelial growth of P. crustosum under the microscope, (E) plate pairing test for PJ15 and P. crustosum, and (F) blank control of plate pairing test.
Inhibition of P. crustosum by biocontrol microorganism PJ15 on Orah
PJ15 can inhibit the decay of Orah caused by P. crustosum. Only Orah inoculated with P. crustosum rotted, whereas Orah inoculated with PJ15 did not cause decay. Over time, the rotted area of the Orah inoculated with P. crustosum alone continued to spread, whereas the Orah treated with PJ15 showed no decay. After 21 d of inoculation, the decay diameter of Orah inoculated only with P. crustosum reached 47.87 mm, whereas PJ15 still prevented and treated no decay (Figure 3). PJ15 can inhibit sourness and nutrient loss of Orah caused by P. crustosum. Compared with the control, the PJ15 treatment increased the pH by 11.32% (P<0.05), the soluble protein content by 163.64% (P<0.05), the ascorbic acid content by 160.84% (P<0.05), and total soluble sugar content by 203.53% (P<0.05). The incidence rate of fruit in the PJ15 treatment group was significantly lower than that in the control group (P<0.05; Table 2).
Comparison of Orah quality and incidence rate
| Index | CK | T |
|---|---|---|
| Rot diameter (mm) | 47.87±1.17a | 0.00±0.00b |
| Fruit acidity (pH) | 4.33±0.02b | 4.82±0.01a |
| Soluble protein content (mg/kg) | 0.11±0.00b | 0.29±0.00a |
| Ascorbic acid content (mg/kg) | 33.63±5.06b | 87.72±4.37a |
| Total soluble sugar content (g/kg) | 31.15±0.03b | 94.55±0.08a |
| Incidence rate (%) | 100.00±0.00b | 10.00±8.17a |
| Index | CK | T |
|---|---|---|
| Rot diameter (mm) | 47.87±1.17a | 0.00±0.00b |
| Fruit acidity (pH) | 4.33±0.02b | 4.82±0.01a |
| Soluble protein content (mg/kg) | 0.11±0.00b | 0.29±0.00a |
| Ascorbic acid content (mg/kg) | 33.63±5.06b | 87.72±4.37a |
| Total soluble sugar content (g/kg) | 31.15±0.03b | 94.55±0.08a |
| Incidence rate (%) | 100.00±0.00b | 10.00±8.17a |
The results are expressed as mean±standard deviation. Different lowercase letters in the same row indicate significant difference between the two groups.
Comparison of Orah quality and incidence rate
| Index | CK | T |
|---|---|---|
| Rot diameter (mm) | 47.87±1.17a | 0.00±0.00b |
| Fruit acidity (pH) | 4.33±0.02b | 4.82±0.01a |
| Soluble protein content (mg/kg) | 0.11±0.00b | 0.29±0.00a |
| Ascorbic acid content (mg/kg) | 33.63±5.06b | 87.72±4.37a |
| Total soluble sugar content (g/kg) | 31.15±0.03b | 94.55±0.08a |
| Incidence rate (%) | 100.00±0.00b | 10.00±8.17a |
| Index | CK | T |
|---|---|---|
| Rot diameter (mm) | 47.87±1.17a | 0.00±0.00b |
| Fruit acidity (pH) | 4.33±0.02b | 4.82±0.01a |
| Soluble protein content (mg/kg) | 0.11±0.00b | 0.29±0.00a |
| Ascorbic acid content (mg/kg) | 33.63±5.06b | 87.72±4.37a |
| Total soluble sugar content (g/kg) | 31.15±0.03b | 94.55±0.08a |
| Incidence rate (%) | 100.00±0.00b | 10.00±8.17a |
The results are expressed as mean±standard deviation. Different lowercase letters in the same row indicate significant difference between the two groups.
Inhibition of Penicillium crustosum by PJ15 on Orah. CK: Orah only inoculated with P. crustosum; T: Orah wounds inoculated with P. crustosum and PJ15.
Effects of biocontrol microorganism PJ15 on composition of microbial community in Orah
By amplicon sequencing, 79 872 and 80 152 bacterial raw reads and 509 235 and 436 281 fungal raw reads were obtained in the control and PJ15 treatment group, respectively. Non-chimeric reads obtained after quality control included 63 508 and 60 585 bacterial reads and 439 087 and 380 905 fungal reads, respectively. By cluster analysis, 741 and 786 bacterial genera and 32 and 151 fungal genera were obtained, respectively. The bacterial composition of Orah was more abundant than that of fungi. There was no significant difference in the bacterial community composition in Orah by PJ15 treatment (P<0.05), but there was a significant difference in the fungal community composition (P<0.05). The dominant bacterial genera were unclassified Lachnospiraceae, Escherichia Shigella, and unclassified Muribaculaceae. The species of bacterial genera were relatively large, but the relative abundance of a single bacterial genus was low (Figure 4A). The dominant fungal genus in the control was Penicillium (99.65%), which had a substantial advantage in Orah, occupying almost all the ecological niches in Orah, and had strong ability to invade Orah. The dominant fungus in the PJ15 treatment was Meyerozyma (87.35%), whereas the percentage of Penicillium was much lower than that of the control (9.36%; P<0.05; Figure 4B).
Composition of (A) bacterial community and (B) fungi community in Orah.
Effects of biocontrol microorganism PJ15 on microbial community diversity in Orah
Abundance-based coverage estimator (ACE), Simpson, and Shannon indices of microorganisms were significantly higher than those of fungi, indicating that bacterial diversity was higher than fungal diversity in Orah (P<0.05). There was little difference in the alpha diversity index between the control and PJ15 treatment, indicating that PJ15 had little effect on bacterial diversity (P>0.05). The alpha diversity index among fungi was compared, and that of the biocontrol group was higher than that of the control group, indicating that PJ15 had a greater impact on fungal diversity and could improve the diversity of fungi (P<0.05; Table 3).
Alpha diversity index statistics
| Species | Group | ACE | Simpson | Shannon |
|---|---|---|---|---|
| Bacteria | CK | 3039±112a | 0.9985±0.00a | 10.93±0.08a |
| T | 3072±39a | 0.9984±0.00a | 10.96±0.03a | |
| Fungus | CK | 45±17c | 0.0080±0.00c | 0.05±0.00c |
| T | 300±58b | 0.2360±0.08b | 0.89±0.27b |
| Species | Group | ACE | Simpson | Shannon |
|---|---|---|---|---|
| Bacteria | CK | 3039±112a | 0.9985±0.00a | 10.93±0.08a |
| T | 3072±39a | 0.9984±0.00a | 10.96±0.03a | |
| Fungus | CK | 45±17c | 0.0080±0.00c | 0.05±0.00c |
| T | 300±58b | 0.2360±0.08b | 0.89±0.27b |
The results are expressed as mean±standard deviation. The lowercase letters in the columns indicate that the difference between the two groups is not significant if they are the same and significant if they are not the same.
Alpha diversity index statistics
| Species | Group | ACE | Simpson | Shannon |
|---|---|---|---|---|
| Bacteria | CK | 3039±112a | 0.9985±0.00a | 10.93±0.08a |
| T | 3072±39a | 0.9984±0.00a | 10.96±0.03a | |
| Fungus | CK | 45±17c | 0.0080±0.00c | 0.05±0.00c |
| T | 300±58b | 0.2360±0.08b | 0.89±0.27b |
| Species | Group | ACE | Simpson | Shannon |
|---|---|---|---|---|
| Bacteria | CK | 3039±112a | 0.9985±0.00a | 10.93±0.08a |
| T | 3072±39a | 0.9984±0.00a | 10.96±0.03a | |
| Fungus | CK | 45±17c | 0.0080±0.00c | 0.05±0.00c |
| T | 300±58b | 0.2360±0.08b | 0.89±0.27b |
The results are expressed as mean±standard deviation. The lowercase letters in the columns indicate that the difference between the two groups is not significant if they are the same and significant if they are not the same.
Effects of biocontrol microorganism PJ15 on microbial correlation networks in Orah
The microbial correlation network is complicated, and many kinds of positive and negative correlations are interwoven. The number of bacterial interactions in the PJ15 treatment was slightly higher than that in the control, and the PJ15 treatment had little effect on the bacterial interactions. The number of fungal interactions in the PJ15 treatment was much higher than that in the control, and the network interaction between fungi was more complex than that of the control. PJ15 had a greater effect on the interaction between fungi, increasing the density of the fungal interaction network and leading to a closer interaction network relationship. Meyerozyma and Penicillium showed a strong negative correlation. In the control, there were six positive correlations with Penicillium and Cladosporium, unclassified Fungi, Vishniacozyma, Aspergillus, unclassified Saccharomycetales, Plectosphaerella, unclassified Cystobasidiomycetes, and one negative correlation with Tetracladium. The relationship between Meyerozyma and these seven fungi was the opposite of that of Penicillium. A similar inverse relationship was also found in the PJ15 treatment. Penicillium showed three groups of positive correlations with Pichia, unclassified Basidiomycota, and Uwebraunia, and the relationship between Meyerozyma and these three fungi was opposite to that of Penicillium. The number of Penicillium interactions decreased (Figure 5).
Microbial correlation network in Orah. (A) Bacteria correlation network in the control, (B) bacteria correlation network in the PJ15 treatment, (C) fungi correlation network in the control, and (D) fungi correlation network in the PJ15 treatment.
Effects of biocontrol microorganism PJ15 on the metabolite composition of Orah
A total of 15 187 peaks were detected in the pericarp of Orah, of which 4757 metabolites were noted. The most abundant metabolites were carboxylic acids and derivatives (7.04%), organooxygen compounds (5.51%), and fatty acyls (4.39%). According to the KEGG annotation of all identified metabolites, the biosynthesis of other secondary metabolite pathways in Orah accounted for the largest proportion (29.22%), and the KEGG Orthology (KO) pathway level 2 pathway accounted for the most species. This was followed by the amino acid metabolism pathway (7.35%) and the metabolism of terpenoids and polyketides pathway (5.93%; Figure 6). A total of 2372 metabolites (49.86%) were found to be different between the control and PJ15 treatment, of which 997 were upregulated metabolites and 1375 downregulated metabolites. The main upregulated metabolites were N-monomethyl-2-aminoethylphosphonate, 4-oxo-9Z,11Z,13E,15E-octadecatetraenoic acid, (3R)-6-hydroxy-3-isopropenyl-heptanoate, ParishinE, Bayogenin, 6-keto-prostaglandin F1alpha, N4-acetylcytidine, 3-hydroxy-2-methylpyridine-4,5-dicarboxylate, pelargonidin 3-O-rutinoside, and 5-methoxysalicylic acid. The main downregulated metabolites were 3-(2,5-dimethoxyphenyl) propanoic acid, ʟ-arginine phosphate, isopteropodine, garbanzol, methylorsellinic acid, ethyl ester, dacarbazine, 5,6-epoxytetraene, 2-benzylmalic acid, ʟ-homocysteine, and 4-ethylsyringol (Figure 7).
Functional annotation of metabolites in Orah. The horizontal axis represents the proportion of metabolites in the pathway to the total metabolites, whereas the vertical axis represents the name of the metabolic pathway.
Analysis of differential metabolites in Orah. Red represents the upregulation of metabolites, green represents the downregulation of metabolites, and the column length represents the log2FC value.
Correlation analysis of differential metabolites and microorganisms in Orah
The relationship between differential metabolites and bacteria was mostly negative. 5,6-Epoxytetraene was positively correlated with unclassified Vicinamibacteraceae and Acetobacterium. It was negatively correlated with unclassified Oxalobacteraceae, Methylobacterium, Methylorubrum, and Coxiella. The bacterial Saccharofermentans, Prevotella, and dacarbazine, methylorsellinic acid, ethyl ester, garbanzol, isopteropodine, ʟ-arginine phosphate, and Procyanidin A1 showed negative correlation with each other. The bacterial unclassified Bacillales showed negative correlation with the major differential metabolites except N-monomethyl-2-aminoethylphosphonate (Figure 8A). The relationship between major differential metabolites and fungi was also mostly negative. Except for N-monomethyl-2-aminoethylphosphonate metabolites, the major differential metabolites were positively correlated with Penicillium and negatively correlated with Meyerozyma. The metabolites neoandrographolide, 5,6-epoxytetraene, lumichrome, atenolol, and chenodeoxycholate showed a significant positive correlation with Penicillium. However, there was a negative correlation with Meyerozyma (Figure 8B).
Correlation analysis between differential metabolites and microorganisms in Orah. (A) Correlation between differential metabolites and bacteria and (B) correlation between differential metabolites and fungi. Red represents a positive correlation and green represents a negative correlation. ‘*’ indicates P≤0.05, with a significant correlation; ‘**’ indicates P≤0.01, with a highly significant correlation.
Discussion
Fungi are the main cause of microbial deterioration that causes significant economic losses in fresh produce. Yeast, which is present on plant surfaces and in fermented products, has the advantages of not producing allergenic spores or mycotoxins, having simple nutritional requirements, and easy colonization. It is considered a promising biocontrol agent against notorious plant pathogenic fungi, and some yeast-based biocontrol products have been commercialized (Mukherjee et al., 2020; Agirman et al., 2023; Rueda et al., 2023). In this study, a strain of M. guilliermondii (PJ15) was isolated from the surface of grapes and proved to be effective in inhibiting the growth of P. crustosum. Previous studies have also shown that M. guilliermondii is effective against Penicillium expansum (blue mold of citrus), Botryosphaeria dothidea (ring rot of apple), Diaporthe actinidiae (soft rot of kiwifruit), Botrytis cinerea (gray mold of grape), and Penicillium digitatum (green mold of citrus; Cordero et al., 2017; Wang et al., 2018; Agirman and Erten, 2020; Huang et al., 2021; Pan et al., 2022; Sepúlveda et al., 2023; Zhang et al., 2023). It can be seen that M. guilliermondii can control a variety of pathogenic fungi associated with fruit postharvest decay and has great potential to be developed as a biocontrol agent for fruit postharvest decay.
Competition for nutrients and space is the main mode of action of biocontrol yeasts against postharvest pathogenic fungi (Spadaro and Droby, 2016; Agirman and Erten, 2020; Chen et al., 2023), and this study reported similar results. We studied the inhibition of P. crustosum by biocontrol microorganism PJ15 on plates. The mycelial growth of P. crustosum cultured against PJ15 was inhibited, the growth of the colony was dominated by the growth of air mycelia, and the production of green spores was inhibited. The morphology of the edge of the colony was observed under a microscope. Protoplast agglutination occurred in the mycelia of P. crustosum, and the mycelia grew sparsely, entangled into clusters, deformed, and were irregularly twisted and thin. The results of Dikmetas et al. (2023) and Huang et al. (2021) also showed that M. guilliermondii could inhibit the mycelial growth and spore production of pathogenic fungi, which was similar to the findings of this study. In addition, the results of the plate pairing test showed that the antibacterial efficacy against the pathogen was strongest in the middle of the PJ15 colony and weaker at both ends. The growth of the pathogen hyphae far from the PJ15 colony was basically not inhibited. This may be due to the main antibacterial efficacy of metabolite secretion on pathogenic bacteria, as there is resistance to the diffusion of secreted metabolites in solid culture medium, and the range of action is short in a short period of time. The metabolites on both sides diffuse toward the middle, resulting in the highest concentration of antibacterial substances in the middle and exhibiting the strongest antibacterial efficacy in the middle. If the volatile gas plays a major role in inhibiting pathogens, the inhibited fungal hyphae should be parallel to the PJ15 colony rather than curved, and the growth of fungal hyphae farther away from the PJ15 colony should also be inhibited.
We further studied the inhibition of P. crustosum by biocontrol microorganism PJ15 on Orah. The results showed that PJ15 can effectively inhibit the postharvest decay in Orah caused by P. crustosum but did not cause the decay reaction of Orah or the loss of nutrients in Orah. Previous experimental studies in orange have also proved that yeast has potential to control postharvest decay in citrus, blue mold, and green mold (Ferraz et al., 2016; da Cunha et al., 2018; Chen et al., 2020). Studies have confirmed that P. expansum, P. digitatum, and Penicillium italicum can cause acidity of fruit pH, and pH can be used as a factor to increase pathogenicity and adjust suggestion (Zhang et al., 2013). This study also found that the invasion of Orah by P. crustosum also led to acid in wounds. Previous studies using biocontrol yeast after harvest showed that citrus quality was better maintained; weight reduction, total soluble solid content, and ascorbic acid loss were less than those in the untreated group; and sensory perception of fruit aroma was protected (Habiba et al., 2019; Díaz et al., 2020). This study is similar to previous research results.
The microbial diversity of healthy fruit is higher than that of decayed fruit (Wu et al., 2019). This study found that PJ15 had no significant effect on bacterial diversity in Orah but had a significant effect on the composition of fungal community, which could improve the diversity of fungi. In the control, P. crustosum was the dominant fungus, occupying almost all the ecological niches in Orah. PJ15 could effectively colonize and grow on Orah and become a dominant fungus, which could significantly inhibit the growth of P. crustosum. In addition, when inoculated with PJ15 at the same time, the diversity of fungi was higher than that of the control, which may be due to PJ15’s inhibition on P. crustosum, freeing up more ecological space for the growth of other fungi. Jing et al. (2023) found that the top ten genera with the highest abundance of bacteria and fungi on the surface of different citrus varieties were distributed in all four citrus varieties, but their relative abundance varied greatly. Zhao et al. (2023) also studied the fruit microbial diversity of kiwifruit treated by yeast and found that there were eight fungi genera with relative abundance greater than 1%, the proportion of pathogenic microorganisms was the highest in the control, and the relative abundance of biocontrol yeast was the highest in the biocontrol, whereas the relative abundance of pathogenic microorganism was decreased. In addition, previous studies have shown that there was no difference in the effect of postharvest fruit preservative treatment on bacterial diversity. However, the effects on fungal diversity were different (Gomba et al., 2017; Zhang et al., 2020; Shi et al., 2022; Zhao et al., 2023). This study is similar to previous studies.
For the network relationships among microorganisms, PJ15 had little effect on the interaction relationships between bacteria but had a greater effect on the interaction relationships between fungi, increasing the density of the fungal interaction network and forming a more strongly coexisting survival interaction network. This may be because PJ15 inhibited P. crustosum and its companion microbiota upon entry, freeing up more ecological niches for the growth of other microorganisms. This study also found that the genus positively correlated with P. crustosum was precisely negatively correlated with PJ15. Research by Shi et al. (2020) indicates that the role of antagonists is to regulate benign interactions between microbe–microorganism and microbe–host, limit harmful interactions between pathogens and diseased supporting bacteria, and promote fruit health. The genera with large nodes and many edges in a network diagram are considered key genera in microbial communities. The study of Zhao et al. (2023) also found that the density of microbial networks treated with yeast was higher than that of the control, which could increase the correlation between dominant genera of kiwifruit and make the connection between microbial communities closer. Microbial communities have been widely reported to play important roles in plant protection, exploring population dynamics, constructing and synthesizing microbial communities, improving their stability and effectiveness, and accelerating the industrialization process of biocontrol agents (Niu et al., 2017; Li et al., 2021). Tetracladium showed a positive correlation with PJ15 and a negative correlation with P. crustosum, indicating that Tetracladium may be a potential biocontrol microorganism for P. crustosum disease. Studies have shown that Tetracladium is a multilineage group of phylogenetically diverse fungi that grow on decaying leaves and plant litter in streams and has beneficial effects on host health and growth (Lazar et al., 2022).
The biosynthesis of other secondary metabolites had the highest abundance of metabolites in Orah, which might be the signal molecules sent by the microorganisms to adapt to and compete with the environmental pressure. The correlation between differential metabolites and microorganisms was further analyzed, and it was found that differential metabolites were mainly negatively correlated with bacteria, fungi, and PJ15. On the contrary, P. crustosum showed a positive correlation, which also revealed that PJ15, most bacteria and fungi, formed a team to resist P. crustosum. The differential metabolites 5,6-epoxytetraene, lumichrome, neoandrographolide, atenolol, chenodeoxycholate, and P. crustosum showed positive correlation, but the relationship with PJ15 was also the opposite. Lumichrome is a secondary metabolite produced by Penicillium (Ge et al., 2008; Yang et al., 2013). Atenolol proved to be a global secondary metabolic activator, and characterization of other hidden metabolites led to the discovery of a novel epoxy-naphthoquinone. Research of Moon et al. (2019) has shown that atenolol induces occultic antibiotics and has selective growth-inhibiting activity against Gram-negative bacteria. Chenodeoxycholate has antibacterial properties that inhibit Clostridium growth and mediate colonization resistance in vivo (Buffie et al. 2015; Yang et al., 2023). Hao et al. (2012) isolated four analogues of 5,6-epoxytetraene and 4,5-epoxycyclohexenoid from a crude extract of Penicillium and demonstrated moderate cytotoxicity to HeLa cells. PJ15 downregulates differential metabolism 5,6-epoxytetraene, which may be due to PJ15 inhibiting the production of metabolic product 5,6-epoxytetraene by P. crustosum.
Conclusions
This study isolated a strain of M. guilliermondii PJ15 from the surface of grapes that can inhibit the growth of P. crustosum. PJ15 can effectively inhibit the growth of P. crustosum hyphae and the production of green spores and has good biological control potential for postharvest decay caused by P. crustosum in Orah. PJ15 has a relatively small effect on the bacterial composition and diversity on the surface of Orah, but it has a significant effect on the fungal composition and diversity. It can increase the fungal diversity of Orah invaded by P. crustosum, increase the density of fungal interaction networks, and form a stronger coexisting survival interaction network. The vaccination of PJ15 downregulates differential metabolite 5,6-epoxytetraene, which shows a positive correlation with P. crustosum and a negative correlation with PJ15. These results provide a reference for a better understanding of the mechanism of microbial control and provide ideas for the biological control of postharvest decay in Orah.
Acknowledgements
We are grateful to Beijing Biomarker Technologies Co., Ltd. for technical support with microbial diversity sequencing and metabolome detection.
Author Contributions
Jianzong Meng and Liqin Zhou conceived and designed this project. All the authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Liujian Ye, Jialin Han, Xiaohu Wang, Shuang He, Shengbo Wei, and Qixia Zhu. The first draft of the manuscript was written by Liujian Ye, Jianzong Meng, and Liqin Zhou. All the authors commented on previous versions of the manuscript. All the authors read and approved the final manuscript.
Funding
This research was funded by the Guangxi Key Research and Development Program (No. 2021AB26001) and The Central Government Guides Local Science and Technology Development Funds (No. 2020ZYZX3027).
Conflict of Interest
The authors declare no conflict of interest.
