Fusarium spp. Associated with Dendrobium officinale Dieback Disease in China

A rare plant species of the Orchidaceae family, Dendrobium officinale is considered among the top ten Chinese medicinal herbs for its polysaccharide. Since 2021, when the dieback disease of D. officinale was first reported in Yueqing City, Zhejiang Province, China, Fusarium isolates (number = 152) were obtained from 70 plants in commercial greenhouses. The disease incidence ranged from 40% to 60% in the surveyed areas. Multilocus sequence analysis (MLSA) coupled with morphological characterization revealed that the collected isolates belonged to five species (sp.), viz., Fusarium concentricum, F. fujikuroi, F. nirenbergiae, F. curvatum, and F. stilboides, with isolation frequencies of 34.6%, 22.3%, 18.4%, 13.8%, and 10.5%, respectively. Notably, at least two Fusarium species were simultaneously isolated and identified from the infected plants. Finally, the pathogenicity test results demonstrated that such species were responsible for the dieback disease of D. officinale. However, F. concentricum and F. fujikuroi were more invasive compared to the other species in this study. To the best of the authors’ knowledge, this study was the first report of F. concentricum, F. curvatum, F. fujikuroi, F. nirenbergiae, and F. stilboides causing the dieback disease of D. officinale in China and worldwide. This work provides valuable data about the diversity and pathogenicity of Fusarium populations, which will help in formulating effective strategies and policies for better control of the dieback disease.

Tiepishihu is extensively cropped in several provinces of China including Zhejiang, Anhui, Fujian, Guizhou, Guangxi, Sichuan, and Yunnan, with a total cultivation area of nearly 4000 hectares [3][4][5], and it is valued between $450 and $3100 per kg. Recently, micropropagation and greenhouse farming technologies have been utilized to enhance the rate of low natural regeneration of this plant species [2,3,5]. The seedlings derived from tissue cultures are thus transplanted from March to May in greenhouses to produce tillers, and the flowers emerge about 15 months later, between May and July. The stems are further harvested from the fields from the end of October to the beginning of March, approximately 31 months after planting. For three years after the first harvest, the plant can also produce commercial yields (i.e., stems) annually.

Sample Collection, Fungal Isolation, and Morphological Characterization
The D. officinale plants (70) with dieback symptoms were initially sampled from commercial greenhouses in Yueqing (28.07 • N, 120.57 • E), Zhejiang Province, China. The incidence rate of the disease was assessed by visual observations, and then calculated for the presence or absence of symptomatic plants in the surveyed greenhouses. Afterward, the symptomatic stem tips were cut with a sterilized scalpel, superficially disinfected with a 2% solution of sodium hypochlorite (0.1% active ingredient of chlorine; [23]) for 1 min and 75% ethanol for 30 s, rinsed thrice with sterile distilled water, air dried on sterile filter papers under aseptic conditions, and finally placed onto potato dextrose agar (PDA) medium. The plates were subsequently incubated at 25 • C in the dark, and the colonies were purified by the hyphal tip method and then sub-cultured on the PDA and carnation leaf agar (CLA) media for morphological observation [13,24]. The conidial morphology and sporulation of the pure fungal colonies were finally examined under a Nikon Eclipse microscope (Japan).

DNA Sequencing and Molecular Phylogeny
DNA was extracted from the mycelia of 7-day-old cultures of the representative isolates using the Plant Genomic DNA kit (Tiangen, China) according to the manufacturer's instructions. The fragments of the translation elongation factor 1-alpha (tef1), second largest subunit of RNA polymerase II gene (rpb2), and β-tubulin (tub2) genes were then amplified by the primers EF-1/EF-2, RPB2-5f2/RPB2-7cr, and Tub2F/Tub2R, respectively [13,23]. The polymerase chain reaction (PCR) was also performed in 25 µL volumes, containing 1 µL of genomic DNA, 12.5 µL 2 × Phanta ® Flash Master Mix Dye Plus (Vazyme, Nanjing, China), 9.5 µL of DNase-free water, and 1 µL of each forward and reverse primer (10 µM). Notably, the cycling conditions included the initial denaturation of 30 s at 98 • C, followed by 30 cycles of denaturation at 98 • C for 10 s, the annealing at 52 (tef1), 59 (rpb2), and 55 (tub2) for 10 s, the extension of 10 s at 72 • C, as well as the final extension at 72 • C for 1 min. The PCR products were first visualized on a 1% (w/v) agarose gel, and then Sanger sequencing was conducted by Sangon Biotech Co., Ltd. (Shanghai, China) for both directions to ensure high accuracy. The accession number of all generated sequences in this study was further obtained from the GenBank, as listed in Table 1. The aligned sequences of the novel isolates were also subjected to the Basic Local Alignment Search Tool (BLAST) to collect related sequences for inclusion in phylogenetic analysis. The BLASTN searches were fulfilled using the rpb2, tef1, and tub2 sequences against the Nucleotide collection (nr/nt) database by restricting the material type. Multiple sequence alignments were correspondingly inferred in Molecular Evolutionary Genetics Analysis (MEGA) X software version 10.2.4 [25] using the MUSCLE (multiple sequence comparison by log-expectation) program [26] and refined manually if necessary. To generate concatenated datasets, singlegene sequences (tef1, rpb2, and tub2) were manually combined utilizing the BioEdit version 7.1 [27]. The phylogenetic trees were further constructed based on the individual and concatenated sequences (rpb2, tef1, and tub2) using the MEGA X software. The maximum likelihood (ML) and neighbor-joining (NJ) methods were also employed to approximate the distances and complete bootstrapping. As well as the general time reversible model assuming a discrete gamma distribution and invariant sites (GTR+I+G) for the combined aligned dataset, the Tamura-Nei model with gamma-distributed (TN93+G) for rpb2 and the Kimura two parameter model (K2+G) for tef1 were applied as the best evolutionary models for the phylogenetic analyses [25]. The topological support was then determined by 1000 bootstrap replicates. The sequences from the Fusarium spp. type strains, initially identified as closely related to the sequences here, were finally included by the preliminary BLAST searches (Table 1).

Pathogenicity Studies
To reproduce the dieback disease symptoms, the fungal isolates were tested for pathogenicity on the original host. A small, excised wound was accordingly made on the tip of each intact stem after being swabbed with ethanol 75% and washed with sterile water, then a mycelial agar disc (5 mm diameter) from each of the 7-day-old cultures of the fungal isolates was placed onto the surfaces of each stem tip and wrapped with Parafilm [8]. Afterward, the incubated plants were placed in a growth chamber at 25 • C and 75% relative humidity (RH) and maintained for 14 days. In contrast, the control plants received non-colonized agar plugs. Of note, the test was independently replicated thrice. All inoculated plants were visually assessed on a daily basis for up to two weeks. To fulfill Koch's postulates, the same fungal isolates were re-isolated and their identity was confirmed by the tef1 sequence data. To evaluate the disease severity, pear fruits (Pyrus pyrifolia) were also sterilized with 75% ethanol (3 min) and washed with sterile distilled water. Next, a mycelial agar disc (9 mm diameter) was placed on the fruit surface and covered with Parafilm to maintain high humidity [28]. All inoculated fruits were incubated under the same condition as mentioned above for one week. The fresh PDA agar plugs were further used as a negative control. The fruit rot diameter was finally measured by an electronic caliper seven days post inoculation (DPI). Each treatment included three replicates, and the experiment was independently repeated at least twice for both tests.

Field Survey, Disease Symptoms, and Pathogen Isolations
In September 2021, symptoms of the dieback disease on D. officinale emerged in Yueqing, Zhejiang Province, China. According to the field observations, high temperature, high humidity, and poor ventilation would accelerate the incidence rate of this condition, which was at about 40-60% based on the number of plants with dieback disease symptoms recorded in 30 rows randomly picked. The symptoms appeared as chlorotic, blighted, and wilted leaves of the apical meristem with the shoot tip showing dark brown necrosis, dieback, and eventually shoot death (Figure 1). A total of 152 Fusarium-like isolates were also recovered from 70 infected plants, and 20 representative isolates were selected for further analysis ( Table 2). Each isolate was recovered from different infected stems. Consistent with their morphological traits as well as molecular methods, the isolated fungi belonged to five genera, encompassing F. concentricum, F. curvatum, F. fujikuroi, F. nirenbergiae, and F. stilboides. Comparing the isolation frequency accordingly revealed that F. concentricum was the most abundant species, followed by F. curvatum, F. fujikuroi, and F. nirenbergiae, while F. stilboides was found the least (Table 2). Interestingly, two, and occasionally more than two, different Fusarium spp. were simultaneously isolated from some samples, and finally confirmed by the tef1 and rpb2 sequence analyses.
ture, high humidity, and poor ventilation would accelerate the incidence rate of this condition, which was at about 40-60% based on the number of plants with dieback disease symptoms recorded in 30 rows randomly picked. The symptoms appeared as chlorotic, blighted, and wilted leaves of the apical meristem with the shoot tip showing dark brown necrosis, dieback, and eventually shoot death (Figure 1). A total of 152 Fusarium-like isolates were also recovered from 70 infected plants, and 20 representative isolates were selected for further analysis (Table 2). Each isolate was recovered from different infected stems. Consistent with their morphological traits as well as molecular methods, the isolated fungi belonged to five genera, encompassing F. concentricum, F. curvatum, F. fujikuroi, F. nirenbergiae, and F. stilboides. Comparing the isolation frequency accordingly revealed that F. concentricum was the most abundant species, followed by F. curvatum, F. fujikuroi, and F. nirenbergiae, while F. stilboides was found the least (Table 2). Interestingly, two, and occasionally more than two, different Fusarium spp. were simultaneously isolated from some samples, and finally confirmed by the tef1 and rpb2 sequence analyses.

Morphological Identification
The phenotypic criteria of the representative isolates obtained from the symptomatic stem tips matched with the descriptions of F. concentricum, F. fujikuroi, F. nirenbergiae, F. curvatum, and F. stilboides morphology.
In this regard, F. concentricum showed yellow-white, abundant, densely lanose to velutinous aerial hyphae with concentric rings on the PDA. Also, the colonies produced mainly 3-5-septate, naviculate to fusiform, slender macroconidia (Sporodochial conidia) with beaked apical and foot-shaped basal cells. Microconidia (aerial conidia) were also obovoid to fusoid, predominantly with no septa, but occasionally with 1 septum, and borne on mono or poly-phialides found in the aerial mycelia. However, chlamydospores were not observed. On the CLA, orange sporodochia were found (Figure 2).

Morphological Identification
The phenotypic criteria of the representative isolates obtained from the symptomatic stem tips matched with the descriptions of F. concentricum, F. fujikuroi, F. nirenbergiae, F. curvatum, and F. stilboides morphology.
In this regard, F. concentricum showed yellow-white, abundant, densely lanose to velutinous aerial hyphae with concentric rings on the PDA. Also, the colonies produced mainly 3-5-septate, naviculate to fusiform, slender macroconidia (Sporodochial conidia) with beaked apical and foot-shaped basal cells. Microconidia (aerial conidia) were also obovoid to fusoid, predominantly with no septa, but occasionally with 1 septum, and borne on mono or poly-phialides found in the aerial mycelia. However, chlamydospores were not observed. On the CLA, orange sporodochia were found (Figure 2). On the PDA, F. curvatum formed abundant floccose aerial mycelium with pale rosy white hue. Microconidia were also hyaline, ellipsoidal to falcate, 0-1-septate, and bore forming small false heads (i.e., short unbranded conidiophores) on the tips of the phialides. Besides, macroconidia were hyaline, 2-4-septate, banana-shaped, with blunt to papillate apical and blunt basal cells. Also, chlamydospores were not observed, and orange sporodochia formed on the carnation leave ( Figure 2).
The F. fujikuroi colonies on the PDA consisted of floccose white aerial mycelia that became gray-violet or magenta with age, lacking chlamydospores. Notably, some swollen On the PDA, F. curvatum formed abundant floccose aerial mycelium with pale rosy white hue. Microconidia were also hyaline, ellipsoidal to falcate, 0-1-septate, and bore forming small false heads (i.e., short unbranded conidiophores) on the tips of the phialides. Besides, macroconidia were hyaline, 2-4-septate, banana-shaped, with blunt to papillate apical and blunt basal cells. Also, chlamydospores were not observed, and orange sporodochia formed on the carnation leave ( Figure 2).
The F. fujikuroi colonies on the PDA consisted of floccose white aerial mycelia that became gray-violet or magenta with age, lacking chlamydospores. Notably, some swollen cells could develop in the hyphae and superficially appear chlamydospores or pseudochlamydospores. Aerial conidia were also oval-shaped with a flattened base and 0-1-septate on the CLA. The long, slender, usually 3-6-septate macroconidia further proliferated on the monophialides of the branched conidiophores in the sporodochia. Moreover, the pale orange sporodochia was sparsely produced on the CLA (Figure 2).
The F. nirenbergiae colonies were pale vinaceous to burly-wood color, with abundant flocculent aerial hyphae on the PDA. Sporodochial conidia also formed small false heads on the tips of the phialides, lucid, oval to falcate with 0-1-septa. As well, macroconidia were hyaline, generally 3-septate, and in the shape of crescents or sickles with an attenuated to semi-papillate, curved apical and foot-shaped basal cells. The globose to spherical, aseptate chlamydospores were further produced terminally or intercalary. The aerial mycelium also formed abundantly bright orange sporodochia on the CLA (Figure 2).
The aerial mycelia of F. stilboides strains were cottony, velvety, reddish orange to maroon on the PDA. The aerial conidia were also long, cylindrical, smooth-walled, 3-5-septate, straight to almost slightly flexuous in the center and sharpened at the apices with marked foot-shaped cells. Moreover, formation of orange sporodochia was observed on CLA. The microconidia were typically obovoid to elliptical and 0-1-septate, and chlamydospores were present (Figure 2).

Molecular Identification and Phylogenetic Analyses
The PCR amplified partial sequences of the genes tef1, rpb2, and tub2, yielded 651, 718, and 481 bp fragments, respectively. The BLASTN searches against all Fusarium sequences in the GenBank additionally showed that the 20 representative isolates included in this study shared 99-100% similarity with type-strains of five Fusarium spp., namely, F. concentricum, F. fujikuroi, F. nirenbergiae, F. curvatum, and F. stilboides, which supported previous efforts for the identification of these pathogens based on the macro and micro-morphological characteristics. To clarify the phylogenetic relations, the phylogenetic trees were built here from the single genes tef1 and rpb2. These trees included several sequences from the new isolates, Fusarium type strains, plus a few non-type strains. The given trees further supported the identification of the D. officinale isolates (Supplementary Figure S1). The partial tub2 sequences also displayed close similarity with the Fusarium species but provided an insufficient resolution to identify them. For further molecular verification, multilocus phylogenetic analysis (MLSA) was further performed based on 1850 nucleotide positions among 91 in-group taxa, including clades corresponding to FOSC, FFSC, FLSC, and FIESC. The MLSA tree accordingly indicated that the D. officinale isolates in the present study clustered unambiguously with F. concentricum, F. fujikuroi, F. nirenbergiae, F. curvatum, and F. stilboides type strains with the bootstrap values of 99%, 99%, 96%, 95%, and 99%, respectively (Figure 3). The topologies of the trees obtained from each individual gene also resembled each other and were, above all, similar to the MLSA tree (Supplementary Figure S1). Nonetheless, the concatenated dataset, in addition to all the individual phylogenies, clearly determined the phylogenetic relationship and taxonomy of Tiepishihu isolates in the Fusariod taxa. Moreover, the F. equiseti strain DFS, as a pathogen of the dieback disease of D. officinale in China, fell strongly with the F. equiseti clade, which belonged to the FIESC (Supplementary Figure S1).

Pathogenicity Assays
Two weeks after inoculation, the pathogenicity test results revealed that the isolates from five Fusarium spp. had the typical black brown necrosis and the dieback symptoms on the tip of D. officinale (tie-pie variety) stems, which were congruent with the field observations, while no symptoms developed on the control plants inoculated with the agar media. All these fungal species were also re-isolated from the inoculated plants, and identified using the tef1 locus, thereby fulfilling Koch's postulates ( Figure 4A). Notably, the dieback disease symptoms incited by each pathogen were indistinguishable in the field. As such, all Fusarium spp. isolates were pathogenic on Tiepishihu and caused the dieback disease in these inoculation studies. Furthermore, F. concentricum and F. fujikuroi, among the assayed isolates, indicated higher virulence on pears than the others, followed by F. nirenbergiae and F. curvatum, which had relatively similar disease severity, whereas F. stilboides showed the lowest virulence in this respect ( Figure 4B; Table 3).  dieback disease symptoms incited by each pathogen were indistinguishable in the field. As such, all Fusarium spp. isolates were pathogenic on Tiepishihu and caused the dieback disease in these inoculation studies. Furthermore, F. concentricum and F. fujikuroi, among the assayed isolates, indicated higher virulence on pears than the others, followed by F. nirenbergiae and F. curvatum, which had relatively similar disease severity, whereas F. stilboides showed the lowest virulence in this respect ( Figure 4B; Table 3).

Discussion
The dieback disease has already plagued the D. officinale industry with a high incidence rate in Zhejiang Province, China. On the basis of the MLSA supported by morphological observations in this study, five distinct taxa, viz., F. concentricum, F. fujikuroi, F. nirenbergiae, F. curvatum, and F. stilboides, causing the dieback disease on Tiepishihu, were diagnosed. The study findings also suggested that the Fusarium species associated with this condition on Tiepishihu were more diverse than the ones previously recorded [8]. The Koch's postulates correspondingly showed that the Fusarium spp. Isolates were infective in nature, with slight variations in virulence.
The validity of morphological identification in this study was thus confirmed by the phylogenetic analyses derived from the molecular results. The tef1, rpb2, and tub2 genetic barcodes were thus selected for this purpose because they consisted of phylogenetically informative sequences for the differentiation and classification of Fusarium species [13,16,19,20].
The tef1 phylogeny accordingly demonstrated better resolution at the species level in comparison with rpb2 and tub2. For the concatenated gene analyses, the topologies of the trees inferred for individual genes were also evaluated visually to establish that the overall tree topology of the single locus datasets were similar to each other and to that of the tree acquired from the combined dataset alignment. In the MLSA tree, the Fusarium isolates from D. officinale in the present study were phylogenetically different from each other and were situated within the FOSC, FFSC, and FLSC clades with F. concentricum CBS 450.97, F. curvatum CBS 238.94, F. fujikuroi CBS 221.76, F. nirenbergiae CBS 840.88, and F. stilboides CBS 746.79 type strains, while none of them fell in the FBSC, FRSC, and FsaSC clades. Even if F. nirenbergiae and F. curvatum were placed in the same clade, they formed two distinct well-supported subclades, which correlated with clade VIII resolved by Lombard et al. [16]. Furthermore, the tree topology of the concatenated dataset in this study was closely similar to the trees developed by Lombard et al. [16], Crous et al. [13], and Yilmaz et al. [19].
Despite much effort, there was no success in defining the pathogen-specific diagnostic criteria for these five taxa, which induced the dieback disease on D. officinale. Further surveys are thus required to establish the species-specific symptoms of this condition. Additionally, the symptoms on some stem tips may slightly differ in respect of intensity or color, suggesting that such symptoms may have been due to secondary infections by saprophytic microbes or affected by environmental conditions, such as high humidity or unfavorable ventilation, as documented in previous research [21,22,35,[45][46][47][48][49][50].

Conclusions
In sum, the study data here confirmed that the losses in D. officinale yields were caused by F. concentricum, F. curvatum, F. fujikuroi, F. nirenbergiae, and F. stilboides. Notably, the incidence rate of these local outbreaks could be triggered by environmental factors, and it is expected to increase in the future because of both climate change and susceptible cultivars. Regarding the significance of this study, it provided information on the biodiversity and epidemiology of Fusarium spp. associated with the dieback disease, which can contribute to the development of breeding programs and disease management strategies.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jof8090919/s1, Figure S1: Phylogenetic tree of the Fusarium species resulting from maximum likelihood based on the sequence of tef1 gene. Bootstrap support values were given at the nodes. Isolates isolated in this study are shown by a black circle (•), while F. equiseti strain DFS is indicated by a black square ( ).