Novel isolate of Cladosporium subuliforme and its potential to control Asian citrus psyllid, Diaphorina citri Kuwayama (Hemiptera: Liviidae)

Abstract

Background

Citrus fruits are economically and nutritionally important but have been severely affected by Huanglongbing disease (HLB), its natural spread is mainly by the Asian Citrus Psyllid, Diaphorina citri Kuwayama (Hemiptera: Liviidae). Chemicals are often used to control this pest, but this is not sustainable. Meanwhile, there are few environmentally friendly bioinsecticides to control D. citri in China.

Results

In this study, an entomofungal pathogen wz-1 was isolated from a D. citri carcass in the field, which resulted in a cumulative mortality rate of 75.27% in adult D. citri 7 days after inoculation of the spore suspension. It was identified as Cladosporium subuliforme based on morphological analysis as well as sequence analysis of several molecular markers (Internal Transcribed Spacers, Translation Elongation Factor 1-α and Actin). Remarkably, the lethality rate of adult D. citri reached 53.13%, 48 h after treatment with the aqueous phase extracts of wz-1. Hydroxyquinoline and phytosphingosine in the extracts were identified as potentially active metabolites using LC–MS.

Conclusions

The entomopathogenicity and bioinsecticidal potential of C. subuliforme were previously unknown. Obtained results showed that both spores and extracts of wz-1 can effectively kill adult D. citri, providing an available fungal resource and a theoretical basis for biocontrol of the HLB insect vector D. citri.

Background

Huanglongbing (HLB) is currently a devastating bacterial disease that seriously affects citrus production worldwide, causing incalculable economic losses. The causal agent of HLB is “Candidatus Liberibacter asiaticus” (CLas), a phloem-limited while in vitro unculturable alpha-proteobacterium. CLas is carried by the Asian citrus psyllid Diaphorina citri Kuwayama (Hemiptera: Psyllidae) for horizontal transmission between plant hosts, and HLB spreads rapidly when ambient temperatures are favorable for the insects. Management of HLB is still challenging despite the fact that numerous advances have been made in improving CLas detection (Gottwald et al. 2020), targeting CLas (Wang et al. 2017), treating HLB-diseased trees (Li et al. 2019), controlling D. citri (Boina and Bloomquist 2015), and breeding tolerant or resistant citrus varieties (Alquézar et al. 2021). To date, the use of pathogen-free citrus seedlings, removal of infected trees, and chemical control of D. citri are among the best strategies to prevent HLB outbreaks, and these strategies are extensively used in China (Zhou 2020).

The control of D. citri through the frequent use of insecticides in citrus orchards is not sustainable as this has led to serious environmental problems and the development of insecticide resistance (Tudi et al. 2021). Therefore, it is of utmost urgency to develop other strategies to achieve sustainable control of D. citri. In terms of environmental safety, the use of new biocontrol agents that occur in the same habitats of D. citri or new environmentally friendly bioinsecticides may be better choices. The search for new microbial or molecular bioinsecticides among microorganisms or their natural bioactive metabolites is feasible (Chandler et al. 2011). Among the various endophytic microbes, fungi in particular have been proven to be a largely untapped reservoir of potential bioinsecticides (Srivastava et al. 2009), and we will focus on those entomofungal pathogens that produce insecticidal metabolites playing important roles in the pathogenesis.

In the present study, a new fungus closely related to Cladosporium subuliforme was isolated from a cadaver of D. citri on citrus grown in orchard in Guilin City of Guangxi Province, China. To explore the potential of this fungus in biological control of D. citri, it was an attempt to (i) identify the isolate in the respects of morphological and molecular characteristics, (ii) investigate its virulence to D. citri adults under laboratory conditions, and (iii) extract its metabolites to be tested for insecticidal activity.

Methods

Sampling, insects and plants

A field survey was first initiated in October 2021 in citrus orchards located in Guilin City (110°19′31" E, 25°16′15" N) for collecting the bodies of D. citri individuals presumed to be dead due to diseases. D. citri population used for subsequent tests was collected from a citrus orchard in Guilin City more than three years ago, and has been raised in the laboratory of Guangxi Academy of Specialty Crops (110°18′51" E, 25°5′18" N). To maintain a population, a total of 300 healthy Murraya paniculata plants were cultivated in plastic pots inside a net within a greenhouse (27℃, photoperiod of 16 h light: 8 h dark).

Isolation and identification of the entomopathogenic fungus

The cadavers were surface-sterilized by immersion in 75% ethanol, and rinsed three times with sterile distilled water to isolate any potential entomopathogenic fungus (Ishii et al. 2015). Then, the cadavers were incubated on PDA medium at 26 ± 1 °C, 65 ± 5% RH with a 14:10 h light:dark (L:D) photoperiod. Subculturing of microbes that grew on the cadavers resulted in the isolation of a morphologically uniform fungal colony, which was designated as wz-1. After seven days of incubation, the mycelium and conidia mixtures were scraped off from the plates, and the morphological characteristics were observed under an Olympus fluorescence microscope (Olympus, Tokyo, Japan). Later, the D. citri adults were inoculated with the conidia (1 × 10⁸ spores/ml) by spray method, and raised in cups with hydroponic fresh citrus leaves. Once the inoculated D. citri died prematurely, any mycelium observed on the cadavers after a certain period of time would be isolated for further analysis. This would help determine whether the mycelium is identical to the wz-1 isolate, thereby fulfilling the Koch’s postulates.

After incubation of the wz-1 isolate on PDA medium at 26 ± 1 °C for one week, the mycelia were harvested and genomic DNA (gDNA) was extracted using a fungal DNA kit following the instructions (Sangon Biotech, Shanghai, China). To amplify the specific DNA fragments of the fungus, three pairs of primers were used: ITS1/ITS4 for a region spanning the internal transcribed spacers (ITS) (Zhang et al. 2005), EF1-F/EF1-R for the translation elongation factor 1-α (EF-1α) gene, and ACT-512F/ACT-783R (Carbone and Kohn 1999) for the actin gene (ACT). The total reaction volume of the PCR is 25 µl, with 0.2 µl TaKaRa Ex Tap HS (5 U/µl), 2.5 µl 10 × Ex Tap Buffer (Mg²⁺Plus), 2 µl dNTP Mixture (2.5 mM) (Takara, Shiga, Japan), 1 µl 5 µM forward primer, 1 µl 5 µM reverse primer, 1 µl the DNA template, and 17.3 µl ddH2O. The PCR products were separated by 1% agarose gel electrophoresis and the fragments of expected size were purified according to the instructions of QIAquick Nucleotide Removal Kit (QIAGEN, Valencia, USA). The purified PCR products were then cloned into pGEM®-T Easy vectors (Promega, Madison, WI, USA) and transformed into JM109 High Efficiency Competent Cells (Promega, Madison, WI, USA). Finally, at least five positive clones of each the amplicon were randomly selected for sequencing in both directions. Sequencing reads were assembled using DNAMAN 6.0 (Lynnon Biosoft, San Ramon, USA). The identity of the assembled sequences was confirmed by BLASTn searches against ITS, EF-1α, and Act sequences in the GenBank databases (http://www.ncbi.nlm.nih.gov/). The obtained sequences were submitted to the NCBI GenBank database.

Multiple sequence alignments of each locus amplified by the specific primers with its related sequences retrieved from databases were performed using MAFFT v7.490 (Katoh and Standley 2013). Phylogenetic analyses were carried out using IQ-TREE v.1.6.8 (Minh et al. 2020) and MrBayes 3.2.5 (Ronquist et al. 2012) with the maximum-likelihood and Bayesian inferences, respectively. The best nucleotide substitution model was estimated using Partition Finder v.2.1.1 (Lanfear et al. 2017). In the ML inference, bootstrap support (BS) value of each split node was estimated using 1,000 bootstrap replicates, with a branch of BS ≥ 70% considered to be strongly supported. In Bayesian inference, we ran two independent sets of Markov chains, each with one cold and three heated chains. Each chain was run for 10⁷ steps, with samples drawn every 1,000 steps. The convergence was assessed with effective sample size (ESS) values ≥ 200 in Tracer v1.7.0 (Rambaut et al. 2018). Finally, the fungal isolate was sent to and stored at Guangdong Microbial Culture Collection Center (GDMCC) with the conservation number GDMCC 62135 in Guangdong Province, China.

Bioassays

The pathogenicity of the isolate wz-1 to D. citri adults (3–5 days old) was evaluated under laboratory conditions. Four M. paniculata plants with flushes were placed into 120 mesh screen cages (60 × 60 × 90 cm). D. citri adults with mixed gender and similar growth were introduced in each cage for oviposition for 24 h. After that, the adults were removed and the newly emerged adults (3–5 days old) were collected and prepared for spray inoculation with the conidia. Conidia were suspended in 0.05% Tween-80 (v/v) and then diluted with distilled water. After gradient dilution, the concentration was calculated by hemocytometer. Each spray assay consists of five corresponding conidia concentration treatments (1 × 10⁶, 5 × 10⁶, 1 × 10⁷, 5 × 10⁷, 1 × 10⁸ spores/ml) plus a control treatment with 0.05% Tween-80. Thirty anesthetized D. citri adults were placed on a spray tower round table, sprayed with 1 ml of spore suspension and transferred to a cup with M. paniculata leaves for rearing. There were three independent replicates in each treatment. All the treated D. citri were maintained in the incubator (26 ± 1 °C, 80 ± 5% RH with a 14:10 h L:D photoperiod) and checked daily for seven days. Dead D. citri from all treatments were transferred to moist filter paper for further fungal incubation. To ensure a good fit for probit models and accurate calculation of the slopes of the log-dosage (dose) probit lines (Ld-P), median lethal concentration (LC₅₀), and 95% confidence interval values, it is necessary to use a minimum of five concentrations of conidia for bioassays and ensure that the mortality rate caused by them is between 10 and 90%.

Sequential extraction and UPLC-Q/TOF–MS analysis of insecticidal metabolite (s)

The fungi were cultured on PDA for five days, after which they were harvested from the plates, frozen in liquid nitrogen, and ground. Sequential extractions were performed with aqueous and five organic solvents (Methanol, Ethanol, Dichloromethane, Ethyl acetate, and Petroleum ether) to identify the most suitable solvent for extraction of high quantity of biologically active metabolites. For each extraction, 10 ml of each solvent was added to 10 mg of the fungus, and the mixture was subjected to ultrasound for 30 min. The resulting aqueous and five organic phases were collected and concentrated to dryness by rotary evaporator under vacuum, keeping the temperature below 45 °C. All extracts were then dissolved in 1 ml of water and sprayed onto 30 anesthetized D. citri using a spray tower. The treated D. citri were transferred into cups containing M. paniculata leaves for rearing. Three independent replicates were performed for each extract, with the solvents (water or organics) and the extracts (aqueous or organic) of PDA culture medium used as controls.

The chemical constituents of the aqueous extracts from C. subuliforme and the aqueous PDA extracts as negative control were separately analyzed twice using an Agilent 1290 LC coupled with an Agilent 6530 Quadrupole/Time of flight mass (Q/TOF) spectrometer (Agilent, Santa Clara, CA, USA) equipped with a Waters BEH C18 (2.1 × 50 mm, 1.7 µm; Waters, USA). The Q/TOF–MS (mass spectrometry) instrument was equipped with a Dual-AJS ESI source in the positive ionization mode, with the mobile phase consisting of mobile phase A (0.1% formic acid in ultrapure water) and mobile phase B (acetonitrile). The injection volume was 3 µl, and the flow rate was 0.3 ml/min. The gradient elution profile was as follows: 0–3 min, 5–5% B; 3–8 min, 5–15% B; 8–18 min, 15–20% B; 18–22 min, 20–45% B; 22–35 min, 45–50% B; 35–40 min, 50–95% B. The MS was performed under the following conditions: Mass Hunter workstation (Agilent) was used to obtain profile data with a mass range of m/z 100–1000 at a rate of 1 spectra/s. The temperature and the gas flow rate of drying gas were set at 325℃ and 8 l/min, respectively. The sheath gas temperature was set at 350℃, sheath flow rate was set at 11 l/min. The nebulizer pressure was set at 45 psi. The capillary voltage and fragment voltage were set at 4000 V and 140 V. The fixed collision energies were 10, 20, and 40 V, and reference masses of m/z 121.0509 and 922.0098 were continuously introduced for accurate mass calibration. Compounds in the extracts of C. subuliforme were tentatively identified by comparing the parent ion, retention time, and mass fragments with the correlated references and database. Agilent MassHunter Qualitative Workflows B.08.00 was used for processing MS1 and Auto MS/MS data acquired using LC/Q-TOF. The Find by Auto MS/MS function was used to search for compounds and generate precursor and spectra of ion fragment. The obtained precursor ion data of the metabolites were searched against the METLIN (https://metlin.scripps.edu/), MoNA (http://mona.fiehnlab.ucdavis.edu/), and PubChem (https://pubchem.ncbi.nlm.nih.gov) online databases to identify the inferred compounds. The fragment spectra were compared with the experimental MS and MS/MS results using the commercial METLIN database (MassHunter PCDL Manager version B.06.00), which contains 64,092 compounds. The compounds were tentatively identified with a database match score > 90. Only the compounds that consistently occurred in the fungal aqueous extracts, while not found in the aqueous PDA extracts were considered to be specific to the fungus.

Statistical analyses

Statistical analysis was performed using SPSS Statistics version 22.0 software (IBM Corporation, Armonk, USA). Corrected mortality was used based on Abbott’s formula (Abbott 1925). The median lethal concentration values of C. subuliforme for adult D. citri and the 95% confidence interval of LC₅₀ were calculated using the probit regression graphing. For determining significant differences of insect mortality rate between the control and fungal treatment in the bioassay, one-way ANOVA and Tukey’s honest significant difference (HSD) tests were carried out. Differences were considered significant if p-value < 0.05.

Results

Identification and characterization of the isolate

The investigation in the orchards showed that some D. citri cadavers coated with brown mold layer. To determine whether this mold layer was formed by the pathogen causing the death of D. citri, the cadavers removed of the mold layer and other microbial contaminants on the surface were cultured on PDA medium to produce new microbes from which the fungal colony wz-1 was isolated through subculturing. Cultured on PDA medium for seven days, the wz-1 colony reached 24–30 mm in diameter. In the front side, the colony was grayish olive, flocculent, fluffy, without exudate but with a narrow white margin that was regular or slightly undulating; aerial mycelia abundant, sparse, fluffy; spores abundant in the center (Fig. 1A). In the back side, the colony was dark olive gray, deep into the agar, with grooves and folds in the center, and with the white margin (Fig. 1B). Hyphae sparingly branched, 1−4 µm wide, septate; conidiophores branched, branches straight; ramoconidia (conidiogenous cells) cylindrical, chain like; conidia numerous, catenate, extended from the terminal part of intercalary or apical ramoconidia; small terminal conidia limoniform (Fig. 1C and D).

Fig. 1

Cladosporium subuliforme isolate wz-1 on PDA medium and its infection on the adult of D. citri. A The fungal colonies on PDA (Petri dishes with 90 × 15 mm); B The back of fungal colonies on PDA; C Trifurcate conidiophore with ramoconidia and conidia; D Subulate conidiophore, scale bars = 10 µm; E D. citri adult 7 days post infection (dpi) by C. subuliforme with 1 × 10⁸ spores/ml; F D. citri adult 15 dpi by C. subuliforme with 1 × 10⁸ spores/ml

To fulfill Koch’s postulates, adult D. citri were spray-infected with a spore suspension. Compared to the control group, in which the insects dispersed on the leaves, the spore-treated insects began to disperse on the bottom of the apparatus. The treated insects began to die after 72 h, and the dead insects were transferred to wet filter paper for incubation. After four days, bright green mycelia appeared on the legs, mouthparts, and intersegmental areas of the insects (Fig. 1E). By the fifth day, the entire body was covered with light green mycelia. Over time, the mycelia gradually turned brown (Fig. 1F), which resembled the hue observed on the field cadavers. Thus, these results suggest that the mold layer on the original cadaver is produced by a fungus that can kill adult D. citri under laboratory conditions.

To clarify the taxonomic status of wz-1, partial sequences of the ITS (GenBank accession no. OQ351785), EF-1α (OQ357404), and Act (OQ357405) genes were obtained and analyzed. These sequences were homologous to those of different members of the genus Cladosporium, as the online BLASTn analyses showed 99.58–100% identities with the closest relatives. Phylogenetic analysis was performed based on the combined ITS, EF-1α, and Act sequences. Representative sequences of 21 in-group taxa were also included in the analysis (Table 1), as well as that of Cercospora beticola as an out group. The best nucleotide substitution models applied to construct a ML tree was GTR + G + I for the EF-1α, TIMEF + I + G for the ITS, and HKY + G for the Act. The BI consensus tree confirmed the topology and bootstrap support value of the ML tree. The wz-1 isolate and members of the genus Cladosporium were phylogenetically clustered together, and distinct from the out group. It was located in the branch for Cladosporium subuliforme, which was supported with a high bootstrap value of 74%, supporting the inclusion of wz-1 as a new isolate in this species (Fig. 2). In this branch, the clade of wz-1 was closest to that of the strain CBS:126500.

Table 1 GenBank accession numbers of DNA sequences of Cladosporium species, used in the phylogenetic analysis

Fig. 2

Maximum likelihood tree inferred from combined ITS, EF-1α, and Act sequences showing the phylogenetic relationships between the isolate wz-1 and representative members of Cladosporium spp. Branch lengths are proportional to phylogenetic distance. In the ML inference, bootstrap support (BS) value of each split node was estimated using 1,000 bootstrap replicates. In Bayesian inference, two independent sets of Markov chains were run, each with four chains. Each chain was run for 10⁷ steps, with samples drawn every 1,000 steps. ML BS values of ≥ 70% and posterior probability (PP) values of ≥ 0.95 are shown above the branches. Tickened branches indicate BS = 100% and PP = 1.00. Cercospora beticola (CBS 116456) was used to root the tree

Lethality of the fungal spores and extracts

Effect of wz-1 spore on D. citri adults was assessed by spay inoculation with spore suspension. Results demonstrated that the mortality of the inoculated D. citri adults in a certain period of time increased with the spore concentration. The D. citri began to die on the second day, and the mortality gradually increased with time, with the mass mortality beginning to occur on the fifth day. After treating the D. citri with different concentrations of spores for 7 days, the median lethal concentration (LC₅₀) was determined to be 1.55 × 10⁷ spores/ml (Table 2), and the mortality of the D. citri was more than 70% after 7 days when the treatment was 1 × 10⁸ spores/ml (Fig. 3). These results indicated that the wz-1 strain was highly pathogenic to D. citri. This is the first report of the pathogenicity of C. subuliforme to an insect.

Table 2 Regression of Ld-P line of the Cladosporium subuliforme wz-1 strain against Diaphorina citri after 7 days of infection

Fig. 3

Observation of adult Diaphorina citri mortality caused by Cladosporium subuliforme wz-1 isolates with varying conidial concentrations over time. Different colored lines showed different conidia concentrations. The black-dashed line represented the control (0.05% Tween-80, v/v)

To understand the pathogenesis of wz-1 infecting D. citri adults, the organic and aqueous phases obtained after extraction of the fungus were tested for insecticidal activity. It was shown that only the aqueous extracts caused premature death of D. citri adults, as compared to the solvents (water or organics), the PDA extracts (aqueous or organic), and the fungal organic extracts. The mortality rate at 48 h after treatment with the aqueous phase extracts reached 53.13% ± 2.74%, and were significantly higher than the control (p < 0.001) (Fig. 4). The results indicated that C. subuliforme wz-1 can produce certain extracellular substances against D. citri, and these substances can be extracted with water.

Fig. 4

Toxicity of the wz-1 metabolites against Diaphorina citri. Trend over time in the percentage mortality of D. citri adults with hours

Characterization of chemical components in aqueous extracts by UPLC-Q/TOF–MS

To assess the exploitation potentiality of the aqueous extracts of C. subuliforme wz-1, the chemical constituents of the extracts were analyzed using UPLC-Q/TOF–MS. A total of 11 fungus-specific compounds were preliminary identified from the aqueous extracts by comparing the precursor ion m/z and MS/MS fragments with data in the references and online database. The retention times, molecular formula and weight, and MS/MS fragments of all identified compounds are listed in Table 3. The 11 compounds identified were divided into six classes, namely Nucleosides, Organic acids, Organoheterocyclic, Organic nitrogen, Lipids, Benzenoids. Quinolines and the derivatives among organoheterocyclic compounds are generally good candidates for insecticides and acaricides.

Table 3 Compounds preliminary identified from aqueous extracts of Cladosporium subuliforme by UPLC-Q/TOF-MS

Discussion

To develop new bioinsecticides for sustainable control of D. citri, a series of efforts that included field surveys, microbial subcultures, and preliminary morphology and sequence analyses led to the identification of a new fungus wz-1. The wz-1 belongs to the genus Cladosporium, which is one of the largest genera of dematiaceous hyphomycetes (Costa et al. 2022), and morphological identification of its species is difficult due to the high degree of morphological similarity between closely related species (Sandoval-Denis et al. 2015). Therefore, ITS has been widely used for molecular identification of these fungi. However, analyses using this single locus in genomes can still result in poor resolution of the different species. Several studies have shown that combined utilization of ITS, EF-1α and Act loci can provide an advantage in species delimitation in Cladosporium (Bensch et al. 2010). In the present study, a multi-loci-based DNA sequence typing approach with phylogenetic analysis that complemented the morphological approach to accurately classify the Cladosporium-related isolate wz-1 from D. citri (Figs. 1 and 2) (Costa et al. 2022) was used. Overall, wz-1 was found to be a new isolate of the species C. subuliforme.

The importance of Cladosporium spp. as one of the effective biological control agents against whiteflies, aphids and scale insects has been highlighted, although the killing is relatively slow compared to other entomopathogenic fungi (Bahar et al. 2011). To date, three species of Cladosporium spp. including: C. uredinicola, C. cladosporioides and C. chlorocephalum have been collected from infected Bemisia argentifolii, and were pathogenic to their nymphs and adults (Abdel-Baky et al. 1998). Thirteen isolates of C. cladosporioides were tested against Tetranychus urticae in a single concentration (1 × 10⁸ spores/ml), and the highest total mortality caused by one isolate was 74.76% (Eken and Hayat 2009). Moreover, morbidity caused by C. aphidis in laboratory tests was 83% in whiteflies and 37.5% in aphids (Abdel-Baky and Abdel-Salam 2003). However, only a few entomopathogenic fungi have been found to inhibit D. citri. These included Cordyceps fumosorosea, which resulted in a cumulative mortality rate of less than 70% in adult D. citri (Luo et al. 2022). In contrast, the C. subuliforme strain wz-1, which was isolated from the field carcass, showed a high cumulative mortality rate of 75.27% under the same inoculation conditions (1 × 10⁸ spores/ml on adult D. citri for 7 days), indicating a good potential of wz-1 in controlling D. citri. In addition, C. subuliforme has been documented to colonize on oranges (Citrus spp.) without causing disease. The strong virulence against D. citri and harmlessness to citrus plants make C. subuliforme an excellent bioinsecticide candidate for use in the field to control D. citri, as well as a synergist candidate to enhance the action of other insecticides while reducing consumption.

The potential killing mechanism of entomoinsecticide spores could involve the mechanical pressure that the spores create on the host surface during germination, and this process could potentially weaken and destroy the immune system of D. citri, ultimately leading to the death (Wang et al. 2016). Since spores of wz-1 caused slow death of infected adult D. citri, it is possibly that wz-1 generated appressorium and penetration structures during invasion into the insects. In addition to the direct use of spores and vegetative mycelia, which multiply inside the host body in haemocoel for pest control (Lu 2000), a complex of toxic metabolites and peptides produced by the pathogenic fungi can also be used (Esparza Mora et al. 2018). Several studies have highlighted the toxic metabolites produced by Cladosporium species in the rapid control of various insect pests. For example, secondary metabolites such as azaphilones, xanthonols and sterols (Salvatore et al. 2021) and their derivatives have been shown to have insecticidal activity against the aphid, Acyrthosiphon pisum (Lacatena et al. 2019), Aedes aegypti (Ondeyka et al. 2006) and Trypanosoma cruzi (Alexandre et al. 2017). As the aqueous extracts of wz-1 exhibited toxicity toward adult D. citri, it can be concluded that wz-1 produced bioactive metabolites against the pests during its infection.

Conclusion

A new entomopathogenic fungus was isolated from D. citri, and determined to be a strain of C. subuliforme—wz-1. WZ-1 caused premature death of adult D. citri treated with spores and the aqueous fungal extracts, suggesting that the pathogenicity mechanism is complex. A total of 11 compounds were identified in the aqueous extracts, among them hydroxyquinoline and phytosphingosine were emphasized in the potentiality of bio-insecticidal activity. The information is important for developing bioinsecticides for sustainable D. citri control.