Phylogenetic Analysis and Development of Molecular Tool for Detection of Diaporthe citri Causing Melanose Disease of Citrus

Melanose disease caused by Diaporthe citri is considered as one of the most important and destructive diseases of citrus worldwide. In this study, isolates from melanose samples were obtained and analyzed. Firstly, the internal transcribed spacer (ITS) sequences were used to measure Diaporthe-like boundary species. Then, a subset of thirty-eight representatives were selected to perform the phylogenetic analysis with combined sequences of ITS, beta-tubulin gene (TUB), translation elongation factor 1-α gene (TEF), calmodulin gene (CAL), and histone-3 gene (HIS). As a result, these representative isolates were identified belonging to D. citri, D. citriasiana, D. discoidispora, D. eres, D. sojae, and D. unshiuensis. Among these species, the D. citri was the predominant species that could be isolated at highest rate from different melanose diseased tissues. The morphological characteristics of representative isolates of D. citri were investigated on different media. Finally, a molecular tool based on the novel species-specific primer pair TUBDcitri-F1/TUBD-R1, which was designed from TUB gene, was developed to detect D. citri efficiently. A polymerase chain reaction (PCR) amplicon of 217 bp could be specifically amplified with the developed molecular tool. The sensitivity of the novel species-specific detection was upon to 10 pg of D. citri genomic DNA in a reaction. Therefore, the D. citri could be unequivocally identified from closely related Diaporthe species by using this simple PCR approach.


Introduction
Citrus and their allied genera (including Eremocitrus, Fortunella, Microcitrus, and Poncirus) are widely distributed worldwide, among them, the most popular cultivars belong to the Aurantioideae subfamily of the Rutaceae family. Allegedly, the citrus was originally cultivated in Himalayas 4000 years ago [1]. Nowadays, Citrus is one of the most widely cultivated fruit crops with a planting area of 2.5 million ha and production of more than 38 million tons per year in China [2]. The popular citrus cultivars in China include Citrus reticulata (mandarin), Citrus sinensis (sweet orange), Citrus grandis or Citrus maxima (pumelo), and Citrus paradisi (grapefruit) [3]. 1 ITS = nuclear ribosomal internal transcribed spacer regions; TUB = beta-tubulin gene; TEF = translation elongation factor 1-α gene; HIS = histone-3 gene; and CAL = calmodulin gene.    Table 2. BI tree constructed with combined five-loci data was presented with annotations for isolate number, plant host, and locality. MP tree was similar to the BI tree, therefore only BI tree was shown. D. citri was dominant species and occurred on citrus hosts in countries including China, Japan, Korea, New Zealand, Portugal, and USA. D. citriasiana and D. discoidispora wasere found on citrus plants only. However, D. eres, D. sojae, and D. unshiuensis were found on host plants from multiple genera. Seven isolates obtained in this study clustered in the same group with three isolates from previously known as D. infertilis including ex-type strain (CBS 230.52) and several isolates known as D. citri, this group should be the D. infertilis (Figures 3, S1 and S2). Based on the similar phylogenetic analysis, all of the 140 isolates were identified ( Figure S3). Results showed that D. citri was the predominant species which accounted for 44.3%, following the species of D. eres, D. unshiuensis, D. sojae, D. discoidispora and D. citriasiana, which accounted for 11.4%, 10%, 9.3%, 6.4%, and 3.6%, respectively. There were still 15% isolates that could not be identified to the species level (Supplementary Figure  S3).

Specificity and Sensitivity of PCR Method for Detection of D. citri
As mentioned above, sequences of five loci were obtained for phylogenetic analysis (Table 4, Supplementary Figures S1 and S2), among them, TUB showed the best capability of D. citri distinguishing different from other Diaporthe species ( Figure 5). Therefore, TUB gene was chosen for designing the species-specific primers by matching the forward primer in the varied region and the reverse primer in the conserved region of TUB gene ( Figure 5). As the PCR reaction is performed with the commercial PCR amplification mixture, only the annealing temperature is optimized. Results showed that consistent amplification could be obtained at the annealing temperature from 50 to 60 °C for the species-specific primer pair as shown in Figure 6. Thus, 55 °C was considered as the optimized annealing temperature and used in the following experiments. For the specificity evaluation, the specific primer set TUBDcitri-F1/TUBD-R1 amplified a single product of 217 bp only from the D. citri isolates. The 217 bp amplicon was not observed in other five Diaporthe species (D. citriasiana, D. discoidispora, D. eres, D. sojae, and D. unshiuensis), indicating that the method has good specificity for D. citri ( Figures 7A and S4). The sensitivity was evaluated by using a serial dilution of genomic DNA (gDNA) as templates, results showed that it could amplified the 217 bp fragment from 10 pg of isolate NFHF-8-4 gDNA in 20 μL reaction mixture, indicating very high sensitivity ( Figure 7B).

Discussion
D. citri, a phytopathogenic fungus causing melanose disease has become one of the most devastating citrus pathogens. According to data recorded, the geographic distribution of D. citri has been documented in Asia (China, Japan, and Korea), New Zealand, Portugal (Azores Islands), and USA. Even without the DNA sequence database, D. citri has also been reported in many other countries, e.g., Brazil, Cambodia, Cuba, Cook Islands, Cote d'Ivoire, Dominican, Haiti, Panama, Puerto Rico, Trinidad and Tobago, Venezuela, Mexico, Fiji, Mauritius, Philippines, Thailand, Myanmar, Niue, Samoa, Tonga, Zimbabwe, and Cyprus. In China, D. citri has been documented in several citrus plantations, e.g., Chongqing, Guangxi, Hunan, Jiangxi, Zhejiang, Hong Kong, and Taiwan [21,[47][48][49][50].
For Diaporthe species identification, Santos, et al. [38] suggested the combined multi-locus sequences of ITS, TEF, TUB, CAL, and HIS, which were highly effective for resolving boundaries of Diaporthe species. Also, a single locus TEF gave better delimitation for Diaporthe species in phylogeny analysis [38]. Nevertheless, more accurate identification could be obtained based on the combined sequences from TUB, CAL, HIS, and ITS loci [38]. It has been reported that several Diaporthe species could be confusing, and conflicting results could be observed if only ITS region was used to construct phylogenetic tree [6,39,51]. The D. citri strains were isolated from citrus in China and USA, and pathogenicity test confirmed that D. citri was the causal agent of melanose and stem-end rot of citrus plant [21,32,33]. However, one cluster named as D. citri appeared conflict demonstration with the multi-gene phylogenetic analysis [6,21]. Guarnaccia and Crous [20] analyzed Diaporthe species emerging on citrus in European countries and reconsidered that three isolates which were previously recognized as D. citri, should be the D. infertilis because they were obviously different from other clusters of D. citri based on the phylogenetic analysis. In current study, strong evidence with concatenated multi-locus sequences also showed that D. infertilis was distinct with D. citri. To date, D. infertilis has been found on C. sinensis (Suriname), Glycine max (Brazil), unknown host (Italy), Citrus limon (India), and Mikania glomerate (Brazil), respectively.
In earlier studies, methods based on PCR were developed for detecting fungal pathogens on citrus. For instance, Bonants, et al. [52] designed species-specific primers from the ITS region to detect Phyllosticta citricarpa, a black spot pathogen of orange (Citrus sinensis), and lemon (C. limon). Wang, et al. [53] also designed species-specific primer pair from ITS to detect black spot disease of pumelo (C. maxima). Also, simple PCR was developed to distinguish Phyllosticta citricarpa from Phyllosticta mangiferae by directly using fungal mycelia on PDA or fruit lesions [54,55] with TaqMan probe was developed for routine quarantine of citrus black spot disease [56]. Similarly, real-time PCR based on ITS was used to distinguish Phyllosticta citricarpa from Phyllosticta citriasiana, both species could not be distinguished from each other based on morphological characterization [57].
SCAR-marker was developed to detect Pseudofabraes citricarpa, a fungus causing target spot on Satsuma mandarin (Citrus unshiu) and kumquat (Fortunella margarita) in China [58]. Similarly, SACR-marker derived from random amplified polymorphic DNA (RAPD) was used to simultaneously detect Phytophthora nicotianae and Candidatus Liberibacter asiaticus, the causal agents of citrus roots rot and greening [59]. Pereira, et al. [60] developed a multiplex real-time PCR assay to detect Colletotrichum abscissum and Colletotrichum gloeosporioides, the causal agents of citrus post-bloom fruit drop.
Latent infected D. citri may be the initial source of inoculum of melanose, and a rapid and sensitive diagnosis for detection of this pathogen is currently limited. In previously study, a conserved ITS region was used to design a molecular detection on D. longicolla, D. azadirachtae, and D. sclerotioides [42][43][44][45]. Several studies reported that molecular detection of Diaporthe species from conserved ITS region was weak and poor, thus could not distinguish the Diaporthe complex species [38]. A specific gene TEF was used to detect D. azadirachtae [46]. However, the molecular tool for D. citri detection has not been published. In present study, the novel species-specific PCR assay for detection of D. citri was established. This tool can be useful for routine diagnostic work and would be useful to monitor the prevalence of the D. citri.

Sample Collection and Fungal Isolation
Leaf, fruit, and twig tissues with melanose symptomatic sweet orange (Citrus sinensis) and nanfengmiju mandarin (C. reticulata cv. Nanfengmiju) were collected from Ganzhou city (Xinfeng, Nankang) and Fuzhou city (Nanfeng) Jiangxi Province, China. The samples were collected and took back to Key Lab of Horticultural Plant Biology, Ministry of Education, Huazhong Agricultural University, Wuhan, China. Photos of the diseased samples were captured by using Cannon 600D digital camera (Cannon Inc., Tokyo, Japan). Isolates of Diaporthe-like species were isolated from two citrus cultivars, sweet orange and nanfengmiju mandarin showing melanose symptoms in Jiangxi Province, China. Pure isolates were obtained by cutting off the hyphal tips growing from surface-sterilized diseased material. For fungal isolation, each sample of symptomatic tissues was cut into small pieces (5 × 5 mm) with the junction of diseased and healthy tissues. Small pieces of plant tissues were soaked in 75% ethanol solution for 1 min, surface disinfected in 1% sodium hypochlorite solution (NaClO) for 1 min, then rinsed three times with double sterilized water, and dried on sterile tissue paper. Dried small pieces of plant tissues were placed onto potato dextrose agar medium (PDA) amended with 100 μg/mL streptomycin and 100 μg/mL ampicillin (PDA-SA), then incubated for 2-5 days at 25 °C. After that, mycelium tips growing from small pieces of plant tissues were harvested and transferred to Petri dishes with fresh PDA medium for sporulation at 25 °C for 20-30 days. Monosporic isolation was performed according to the method by Goh [61] and Yin, et al. [62]. Pure fungal isolates were kept at 4 °C whenever they are used.

DNA Extraction from Fungal Mycelia
For genomic DNAs (gDNAs) extraction, fresh fungal mycelia were harvested from 7-day old culture on PDA [21]. A hyphal plug about 1.5 square centimeters was cut off and placed into a 2 mL micro-tube with 200 mg of sterile stainless-steel beads (1.6 mm in diameter). Next, 500 μL gDNAs  The micro-tube was vigorously homogenized at maximum speed for 10 min on the Bullet Blender ® Storm 24 (BBY24M; Next Advance, Inc., New York, NY, USA), then centrifuged at 12,500× g for 6 min. Three hundred microliters of gDNAs supernatant were transferred to a new 1.5 mL micro-tube and 300 μL isopropyl alcohol was added. Then, the mixture was gently mixed at room temperature. The solution was centrifuged at 12,500× g for 6 min. After discarded the supernatant, gDNAs pellets were rinsed twice with 300 μL of 70% ethanol, and air dried. At last, 30 μL of sterile water (ddH2O) was added to dissolve gDNAs pellets following Chi's protocol [64]. The gDNAs quality and quantity were measured via UV absorption at wavelength 260 and 280 nm by Thermo Scientific™ NanoDrop 2000 (Thermo Fisher Scientific Inc., Massachusetts, Waltham, MA, USA). The gDNAs was either used or stored at −20 °C until further processing.

Phylogenetic Analyses of Diaporthe Species
Phylogenetic analysis was carried out by using sequences obtained in current study and those downloaded from NCBI's GenBank (www.ncbi.nlm.nih.gov). Diaporthella corylina (CBS 121124) was selected as an outgroup (Table 4). All unique DNA sequences were consensus and edited with DNASTAR Lasergene Core Suite software programme (SaeqMan v.7.1.0; DNASTAR Inc., Madison, WI, Wisconsin, USA). Sequences combined different loci were aligned using Clustal W program with supplement software package in BioEdit v.7.2.5 [69]. Maximum parsimony (MP) analysis was done by using PAUP (Phylogenetic Analysis Using Parsimony, v.4.0b10) [70]. The goodness of fit values including tree length (TL), consistency index (CI), retention index (RI), rescaled consistency index (RC), and homoplasy index (HI) were calculated for parsimony and the bootstrap analyses [71]. The heuristic search function was used with 1000 random stepwise addition replicates, with tree bisection and MrModeltest v.2.3 [73] was used to perform statistical selection of the best-fit model of nucleotide substitution with corrected Akaike information criterion (AIC). BI analyses were launched with six simultaneous Markov chains which were run for 105 generations, and trees were sampled every 100 th generation (resulting in 10,000 total trees). The calculation of BI analyses was stopped when the average standard deviation of split frequencies fell below 0.01. The consensus trees and posterior probabilities (PP) values were calculated after discarding the first 2000 resulted trees of the analyses as burn-in phase. Finally, above 8000 trees were summarized to calculate the PP in the majority rule consensus tree. Phylogenetic trees were visualized and annotated in FigTree v.1.4.2 [74]. The concatenated alignments and phylogenetic trees were deposited in TreeBASE (study no. S25607), new sequences obtained in this study were submitted to NCBI's GenBank nucleotide database. [ Rayner [745]. The morphology imagines were taken using Canon 600D digital camera (Canon Inc., Tokyo, Japan) after 10 days of incubation. Conidiomata and conidia were observed under the OLYMPUS SZX16 stereomicroscope (Olympus Corporation, Tokyo, Japan), conidial length/wide ratio of 30 conidia was measured with a stage micrometer under a Motic BA200 light microscope (Motic China Group Co., Ltd., Nanjing, China). Alpha and beta conidia were measured for calculating means (x) and standard deviations (SD). The conidia ranges were shown as (min−)x − SD − x + SD (−max) μm (x ± SD). Conidia digital images were captured using Nikon Eclipse 80i compound light microscope imaging system (Nikon Corporation, Tokyo, Japan).

Primer Design and Development of the Molecular Tool to Detect D. citri
A highly varied region in TUB gene was selected as the target for developing molecular tool based on PCR to specifically detect D. citri from other Diaporthe species. Partial TUB gene of D. citri was retrieved from NCBI GenBank database (accession no. MN894459). The obtained sequences were aligned by using Clustal W algorithm in software package BioEdit v.7.2.5 [69]. The primers were designed by analyzing hairpin-dimer potential, length of the desired amplicon, %GC content, and melting temperatures (Ta) in Primer premier 6.0 software (Premier Biosoft International, Palo Alto,California, CA, USA). The primers were synthesized by Wuhan Tianyi Huiyuan Biotechnology Co., Ltd. (Wuhan, China). All the primer sequences used in this study are listed in Table 3.

Conclusions
In current study, it has been documented that Diaporthe species could cause devastating citrus diseases and D. citri was the causal agent of the citrus melanose disease. Based on the phylogenetic analysis with five multi-locus sequences, Diaporthe species boundaries could be clearly delimitated. We also designed species-specific primers from TUB gene to develop PCR method for detecting D. citri. The PCR-based method showed high specificity and sensitivity, that could be applied for detection of D. citri efficiently in practice. In the future, efficient PCR should be developed with citrus tissues infected by D. citri and multiple PCR which can distinguish different Diaporthe species should be developed for the phytosanitary assay in plant quarantine routine work.
Supplementary Materials: The following are available online at www.mdpi.com/xxx/s1, Table S1: Checklist of Diaporthe citri and D. infertilis associated with details citrus host and their allied genera, locality and their reference(s), Figure S1: Phylogenetic trees of Diaporthe spp. by Bayesian inference (BI) analysis based on combined data set and individual locus (ITS, TUB, TEF, CAL, and HIS, respectively). Ex-type, ex-isotype, and ex-epitype strains are indicated in bold. The species Diaporthella corylina (CBS 121124) was selected as an outgroup, Figure S2: Phylogenetic tree of Diaporthe spp. generated by Maximum Parsimony (MP) analysis based on combined data set and individual locus (ITS, TUB, TEF, CAL, and HIS, respectively). Ex-type, ex-isotype, and ex-epitype strains are indicated in bold. The species Diaporthella corylina (CBS 121124) was selected as an outgroup, Figure S3: The prevalence of Diaporthe species on citrus in Jiangxi Province, China based on phylogenetic identification. Number (%) indicate the number of obtained isolates of certain species and the percentage among the total 140 isolates. Figure S4: Species-specific 217 bp TUB gene amplified by the primer pair TUBDcitri-F1/TUBD-R1 was shown with 2% gel electrophoresis. Thirty-eight representatives that were identified based on phylogenetic analysis were used to confirm the specificity of PCR approach. The numbers of D. citri, D. citriasiana, D. discoidispora, D. eres, D. sojae, and D. unshiuensis isolates were 10, 3, 5, 10, 5, and 5, respectively. Lane CK is the double sterile water (ddH2O) as negative control and lane M, 100 bp ladder.