Identification of a new Cordyceps javanica fungus isolate and its toxicity evaluation against Asian citrus psyllid

Abstract The Asian citrus psyllid, Diaphorina citri Kuwayama, is the most serious pest of citrus worldwide. It acts as a vector for a group of phloem‐limited bacteria (Candidatus Liberibacter spp.) that causes Huanglongbing (HLB) disease. Thus, D. citri management is an important strategy against HLB, and biological control is currently considered as the most effective method because of the unsustainable and negative side effects of chemical control. Here, we isolated a new strain of entomopathogenic fungus, Cordyceps javanica (GZQ‐1), from one cadaver of D. citri adult based on its morphological and phylogenetic data. Five conidial concentrations of the C. javanica pathogen (1 × 103, 1 × 104, 1 × 105, 1 × 106, and 1 × 107 conidia/ml) were assessed against six life stages of D. citri (1st‐5th instar nymphs and adults). Results showed that C. javanica GZQ‐1 was highly pathogenic to D. citri nymphs (69.49%–90.87% mortality) and adults (69.98% mortality). The LC50 and LT50 values of C. javanica against 1st‐2nd instar (younger), 3rd‐4th instar (middle aged), 5th instar (older), and adults were 1.20 × 105, 1.10 × 106, 4.47 × 106, 8.12 × 106 conidia/ml and 4.25, 4.51, 5.17, 5.49 days, respectively. Moreover, glasshouse experiments indicated that this C. javanica GZQ‐1 caused higher infection rates of D. citri adults compared to two other fungal strains we previously isolated in the laboratory, Cordyceps fumosorosea (IF010) and Metarhizium anisopliae (CNGD7).


| INTRODUC TI ON
The Asian citrus psyllid, Diaphorina citri Kuwayama (Hemiptera: Liviidae), is one of the most economically important pests of citrus worldwide. It serves as a vector of phloem-limited bacteria (Candidatus Liberibacter spp.), which can cause Huanglongbing (HLB; citrus greening disease), a devastating disease of citrus (Bové, 2006;Lin, 1956). Thus, population management of D. citri is a basic and most important strategy in blocking the spread of HLB disease.
Intensive chemical control of D. citri is the primary strategy currently advocated for HLB management. However, this strategy is pernicious and unsustainable with several negative side effects such as chemical residues, resistance, and pest resurgence being noted (Tansey, Vanaclocha, Monzo, Jones, & Stansly, 2017).
In view of the undesirable side effects of pesticides, biological control has become an advantageous development direction for the prevention and control of agricultural pests. Within this area, entomopathogenic fungi play a particularly important role. In addition to efficacy and cost, there are several advantages of using entomopathogenic fungi: they have broad-spectrum insecticidal activity, diversified species range, complex metabolic types, and offer appropriate safety levels for humans and other non-target organisms (Lacey, Frutos, Kaya, & Vail, 2001). They are also easy to mass-produce and development of host resistance against them is unlikely to occur. They also infect their insect hosts in a unique way entering the hemocoel cavity through the external cuticle, where they absorb nutrients, produce immunosuppressive toxins (Vilcinskas & Götz, 1999), damage the host cells, and interact with the gut microbiota to promote host death. They finally kill the host, and the given fungus grows out the cuticle of the dead insect releasing conidiophores to infect other host individuals Wei et al., 2017;Xiong et al., 2013). All this promotes their good application potential within invertebrate pest control strategies.
Entomopathogenic fungi account for more than 60% of insect pathogenic microorganisms. The utilization of entomopathogenic fungi to control D. citri offers a great development opportunity on account of their significant epidemic potential and the convenience of production. Lezama-Gutiérrez et al. (2012) Kepler et al., 2017), isolated from an adult cadaver of D. citri was identified through morphological as well as phylogenetic data. The pathogenicity of this newly isolated strain to D. citri was analyzed through bioassays against different developmental stages.

| Psyllid cadaver collection and fungus isolation
A new fresh entomopathogenic fungal strain was isolated from an adult D. citri cadaver collected from a glasshouse at South China Agricultural University (SCAU) in Guangzhou, China. The carcass was infiltrated in water containing 5% Tween-80 and shaken vigorously to re-suspend the fungal spores or the mycelium fragments present on the cuticle surface. Then 10 μl of this suspension was plated on standard PDA medium. After 1-2 days of incubation at 27°C, individual germinations (or mycelium regeneration) were transferred to a new plate. Thereafter, this isolate was submitted to several rounds of purification in order to follow morphological stability after the successive transfers. This purified strain was named as "GZQ-1," and deposited in Guangdong Microbial Culture Collection Center (GDMCC) with the deposition number GDMCC 60437.
Before morphological observation, this GZQ-1 isolate was firstly cultivated on SDAY/4 medium (SDAY/4:10 g/L dextrose, 2.5 g/L peptone, 2.5 g/L yeast extract, and 15 g/L agar) in Petri dishes placed within a biochemical incubator (26 ± 1°C, 60% RH, L:D = 14:10). Fungal conidia were then harvested from plates after 5 days, suspended in 0.05% Tween 80 (v/v) and shaken on a vortex mixer for 10 min. The conidial suspension was then filtered, counted and adjusted to the target concentration for pathogenicity tests. The infected nymphs and adults of D. citri were observed and recorded under a binocular microscope (ZEISS, Discovery.V20, Germany), and preliminary morphology was identified based on phenotypic properties and pathogenic fungi morphology (Jaber, Mercier, Knio, Brun, & Kambris, 2016).

| Morphological observation
For the morphometric evaluation of the GZQ-1 isolate, its micro-cultures were first grown on SDAY/4 and incubated at 27°C for 10 days.
Slides were then prepared with lactophenol/blue cotton (10:1) and examined with phase contrast optics under an optical microscope Olympus BX51 (Microscopy GmbH, Gottingen, Germany). Images of the conidia were photographed digitally with an Axio Cam HRc camera (Carl Zeiss) using the Axion Vision SE64 Release 4.9.1 software.
To measure the growth rate and conidia yield of GZQ-1, the fungi were cultured on SDAY/4 at 27°C for 10 days, and then the conidia were scraped from the plates and suspended in 10 ml of sterile water. Following this, the suspension was filtered through Miracloth held in a funnel and quantified using a hemocytometer. The growth rate of GZQ-1 hypha was measured based on their morphology on SDAY/4 medium plate on day 10 of culturing. Both the examinations were repeated three times.

| DNA extraction and ITS sequencing of the GZQ-1 isolate
Total DNA of the GZQ-1 isolate was isolated from each sample of the test strains using a fungal DNA kit following the standard manufacturer's instructions (Fungal DNA Kit; Sangon Biotech, Shanghai, China). Purified DNA specimens were amplified with universal primers, ITS4: 5'-TCCTCCGCTTATTGATATGC-3', and ITS5: 5'-GGAAGTAAAAGTCGTAACAAGG-3'. Each PCR was carried out in 25 μl volumes, containing 12.5 μl 2 × HiFiTaq PCR StarMix buffer (GenStar, China), 1 μl of each primer, 1 μl of total DNA (53.1 ng/μl), and 9.5 μl of ddH 2 O. The reaction was then performed using the following reaction cycles: initial denaturation at 94°C for 3 min, followed by 35 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s and extension at 72°C for 45 s, then a final extension phase at 72°C for 10 min. PCR products were visualized by 1.0% agarose gel electrophoresis, stained with Gold View in 0.5 × TBE buffer (Sangon, Shanghai, China) and photographed under UV light. Following this the target PCR products were sent to The Beijing Genomics Institute (BGI; Shenzhen, China) for complete bidirectional sequencing with PCR primers.
The resulting sequences were checked and aligned using Lasergene v7.1 (DNASTAR, Inc., Madison, WI). Then the similarity of sequences compared with homologous sequences deposited in GenBank (Table 1) was calculated using "BLAST" tools on the National Center for Biotechnology Information (NCBI) website, and a resulting phylogenetic tree constructed using MEGA 6 software (Felsenstein, 1985;Saitou & Nei, 1987). C. javanica strain CBS 134.22 and C. fumosorosea strain CBS 107.10 were used as the Extype strains. Metarhizium anisopliae strains ZJ and YD2-1-8 were used as the out group in the phylogeny analysis.
Phylogenetic hypotheses were analyzed independently for ITS with Maximum Likelihood (ML) using MEGA 6 software based on the Tamura-Nei model for ITS, and a discrete gamma distribution was applied for each analysis with 1,000 bootstrap replicates (ML BS).
The resulting trees are visualized in Figure 4.

| Plants and insects in testing
Murraya paniculata (L) Jacks plants were used in this study. New seedling plants were cultured in 30-cm-diameter plastic pots containing a soil-sand mixture (10% sand, 5% clay, and 85% peat) in a glasshouse at ambient temperature and photoperiod.
The colony of D. citri was first collected from M. paniculata plants at SCAU and then continuously reared on M. paniculata seedlings in a glasshouse onsite at SCAU at 26-28°C and 60%-65% relative humidity under a 14:10 hr (L:D) photoperiod. New plants with fresh shoots were provided for the D. citri culture once a month. For bioassay analysis, six life stages of D. citri were used: 1st-5th nymphal instars and mature adults. Nymphs were collected into Petri dishes using a fine camel hairbrush, and following morphological examination under a stereomicroscope, were classified into their individual nymphal stages based on morphological features.

| Glasshouse bioassays
The glasshouse bioassay of GZQ-1 isolate to D. citri adults ( The mathematical calculation of lethal concentrations LC 50 and confidence limits were carried out by the method of probit analysis (Zhang et al., 2018).

| Morphological identification of GZQ-1 isolate
The morphological features of GZQ-1 isolate are illustrated in Figure 1. The fungal colonies had a concentric ring pattern.
Mycelium texture was velvet like. The center of the colony was pale pink while spores on the edge of the colony were white in color with radial growth (Figure 1a). The mature colony was brownish gray ( Figure 1b). The conidiophores were straight with a long ovoid conidial shape linked into a chain, and the phialides were characterized by a wide globose basal portion with a long distal neck (Figure 1c,d).
The morphological identification and infection observation revealed that GZQ-1 isolate was Cordyceps javanica.
When D. citri nymphs and adults were infected by the GZQ-1 isolate in the glasshouse, the insects were observed to move slowly and suffer twitching of legs and antennae after 48 hr infection. Infected adult D. citri scratched actively, clung tightly to branches and leaves until they were completely covered with hyphae and finally died. Microscopic observation showed that fungal hyphae emerging from the tarsi and intersegmental regions of the legs of infected D. citri after 48-72 hr had grown from the metamere and intersegmental membranes. Then covering the whole of the D. citri including the wings, it caused death after seven days or longer (Figures 2 and 3).

| Sequencing of the ITS genes and phylogenetic analysis
Polymerase chain reaction amplification and DNA sequencing results indicated that the rDNA-ITS gene of GZQ-1 isolate was 616 bp (data not shown), following this the DNA sequence was submitted to GenBank; where it gain the GenBank accession number of MG742216. BLAST results in GenBank revealed that the ITS gene of GZQ-1 was 100% homologous to the C. javanica strain (GenBank accession number MG837718). Moreover, phylogenetic analysis showed that GZQ-1 isolate was closely clustered in the C. javanica clade (Figure 4), which supported our morphological identification that the GZQ-1 isolate is a C. javanica strain. instar, 5th instar nymphs, and adults of ACP following the 7th day of infection, respectively ( Figure 5); there were significant differences among the mortality rates of 1st-2nd instar nymphs, 3rd-4th instar nymphs, 5th instar nymphs, and adult D. citri (p < 0.05).

| Glasshouse bioassays
The semi-field pathogenicity of GZQ-1 isolate to D. citri was compared with two other excellent entomopathogenic fungi, which have been screened under laboratory conditions: C. fumosorosea and M. anisopliae at 1 × 10 7 conidia/ml. Results showed that, when infected with C. javanica GZQ-1 isolate, the survival rates of D. citri adults was 82.6% on the 3rd day, reducing to 35.7% on the 9th day and finally 18.3% on the 11th day. When infected with C. fumosorosea and M. anisopliae, their survival rates were 90.0% and 100% on the 3rd day, reducing to 60.0% and 66.7% on the 9th day and finally 22.3% and 20.0% on the 11th day, respectively. In contrast, only 7.7% mortality was recorded in the non-infected controls on the 9th day ( Figure 6). Results revealed that the infection and pathogenicity of C. javanica isolate to D. citri in semi-field conditions was faster or higher than the two other fungal strains we previously isolated in the laboratory (C. fumosorosea (IF010) and M. anisopliae (CNGD7)) after 7 days of infection in the current study.  (Gallou et al., 2016); its conidial characteristic is also consistent to the findings of Gallou et al. (2016). Moreover, the growth speed and sporulation quantity of our C. javanica GZQ-1 colony was faster than that reported by Yin, Qiong-Bo, Guo-Hua, and Mei-Ying (2010) and Shimazu and Takatsuka (2010).

| D ISCUSS I ON
In terms of the pathogenicity bioassays, the LC 50 and LT 50 values of C. javanica GZQ-1 isolate showed good biological control potential for D. citri management. Pinto, Filho, Almeida, and Wenzel (2012)  In conclusion, the efficacious strategy against Huanglongbing disease by suppressing D. citri populations has been required to affiliate a multi-tactic integrated pest management (IPM) programme (Meyer et al., 2008;Weintraub & Beanland, 2006). In the current study, we identified a high pathogenic C. javanica GZQ-1 isolate based on its morphology and phylogeny and estimated its potential use in Asian citrus psyllid biological control. Our study is expected to enrich the available resource library of entomopathogenic fungi and provide an alternative preparation for inclusion within D. citri IPM strategies.
F I G U R E 5 The mortality of Diaphorina citri different developmental stages treated with various concentrations of Cordyceps javanica (1 × 10 3 , ×10 4 , ×10 5 , ×10 6 , and ×10 7 conidia/ml). Data are mean ± SEM of three tests   to BLQ. The funders had no role in the study design, data collection and interpretation, or the decision to submit the work for publication.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R S CO NTR I B UTI O N
DO and BLQ designed the study. DO, LHZ, and CFG performed the experiments and analyzed the data. XSC and SA participated in data analysis and photography. DO, BLQ, and SA wrote the paper. BLQ supported the grants.

E TH I C S S TATEM ENT
None required.

DATA ACCE SS I B I LIT Y
All DNA sequences are submitted to GenBank with Accession number MG742216. All data generated or analyzed during this study are included in this published article.