Colonization of human opportunistic Fusarium oxysporum (HOFo) isolates in tomato and cucumber tissues assessed by a specific molecular marker

Fusarium oxysporum is a large complex cosmopolitan species composed of plant pathogens, human opportunistic pathogens, and nonpathogenic isolates. Many plant pathogenic strains are known based on host plant specificity and the large number of plant species attacked. F. oxysporum is an opportunistic pathogen in humans with a compromised immune system. The objectives of this study were: (1) to develop a specific marker to detect human opportunistic F. oxysporum (HOFo) isolates; (2) to determine whether or not HOFo isolates can colonize and cause disease symptoms in plants; and (3) to assess Taiwan isolates sensitivity to two agro-fungicides. The primer pair, Primer 5/ST33-R, specifically amplifying Taiwan and international reference HOFo isolates was developed and used to detect and assess the distribution of a Taiwan isolate in inoculated tomato plants and tomato and cucumber fruit. Taiwan HOFo isolate MCC2074 was shown to colonize tomato roots, hypocotyls, and cotyledons, but did not show any visible symptoms. Four days after surface inoculation of tomato and cucumber fruit with the same isolate, MCC2074 was detected in the pericarp and locular cavities of both tomato and cucumber fruit and in columella of tomato fruit. Three Taiwan HOFo isolates were found to be moderately sensitive to azoxystrobin and highly sensitive to difenconazole.


Introduction
Filamentous fungal human pathogens (FFHPs), commonly found in clinics and hospitals [1], affect patients with suppressed immune systems, such as those with immune deficiency syndrome (AIDS), organ transplantation, and cancer chemotherapy [1,2]. Among FFHPs, Fusarium spp. have been shown to cause fusariosis with disseminated infections in immunosuppressed patients resulting in mortality rates from 50 to 75% [3].

Extraction of total DNA from HOFo isolates and other fungi
Fungal DNA was extracted by the method of Wang et al. [36] with some modifications. Mycelium (0.1 g) was transferred into a 1.5 ml microcentrifuge tube containing 500 μl lysis buffer (200 mM Tris-HCl, 50 mM ethylenediaminetetraacetic acid, 200 mM NaCl, 1% n-lauroylsarcosine sodium salt at pH 8.0) and incubated at 65˚C for 30 min. Phenol/chloroform/isoamyl alcohol (25:24:1) (500 μl) was added to each microtube and agitated. The mixture was centrifuged at 18,000 g for 10 min at 4˚C (Rotor: Nr. 12154, Sigma 3k20). The supernatant was

Colonization of Taiwan HOFo isolate MCC2074 in tomato plants
To assess the capability of Taiwan HOFo isolate MCC2074 to colonize tomato plants, they were inoculated by three methods as follows: (1) Hypocotyl inoculation, twenty 7-10-day-old cherry tomato seedlings ('Yu-Nyu', Known-You Seed Co. Ltd) were cut off just above the roots and the hypocotyls immersed in a 5x10 4 /ml conidial suspensions for 30 min, transplanted into sterile peat moss in 5 cm 3 pots, covered with humidity domes to maintain high humidity, and placed in the greenhouse at 25-35˚C. Domes were removed after 3 days and plants examined for disease symptoms and MC2074 colonization 7 days after transplanting. A parallel set of plants handled in the same way but inoculated with F. oxysporum f. sp. lycopersici (isolate of Fol146) and rated for disease severity as follows: 0 = health, 1 = cotyledon and first leaf with yellowing symptom, 2 = stunting or <1/2 leaves with the yellowing symptom, 3 = stem yellowing, vascular discoloration, and >1/2 leaves with wilt symptoms, and 4 = plant wilted and dead.; (2) Soil drench inoculation, seeds of 'Yu-Nyu' cherry tomato and 'Farmers 301' cooking tomato (Known You Seed, Kaohsiung, Taiwan) were sown into sterile peat moss in 5 cm 3 pots and 25 plants of each variety inoculated by pipetting 3 ml of a 1x10 5 /ml conidial suspension into the growth medium at the base of each plant at the 2-leaf stage. Plants were examined for colonization 7 days after inoculation; (3) Infested soil inoculation, inoculum was prepared by infesting a sterile mixture of oat grain, sand, and water (100:100:20, w/w/v) with a conidial suspension (10 ml with 1x10 6 spores/ml) and incubated for 2 weeks (shaken briefly each week) in the greenhouse at 25-28˚C. The inoculum was combined in equal quantities with sterile soil and incubated another 2 weeks in the same environment. Propagule density was estimated by suspending 1 g of the infested soil in 10 ml sterile water and spreading 1 ml aliquots of the suspension onto quintozene (PCNB) peptone agar plates. Infested soil was mixed with sterile soil to achieve 10 4 propagules per gram. Twenty-five 'Yu-Nyu' tomato plants were transplanted at the 2-leaf stage into the infested soil and examined for colonization after 2 wk in the greenhouse.

Inoculation of tomato and cucumber fruit by Taiwan HOFo isolate MCC2074, F. oxysporum f. sp. lycopersicum and F. oxysporum f. sp. cucumerinum
Fruit of 'Yu-Nyu' cherry tomato and 'Swallow' cucumber (Known-You Seed Co. Ltd.) were inoculated with Taiwan HOFo isolate MCC2074 to assess its ability to colonize the fruit. Prior to inoculation, tomato and cucumber fruit were soaked in 1% NaClO solution for 30 min then washed with sterilized water. Surfaces of fruit were abraded with carborundum by a cotton swab to mimic injuries that may occur during harvest and postharvest handling. The injured fruits were then immersed in a 1x10 6 /ml conidial suspension in water with 0.1% tween 20 for 30 sec. Inoculated fruits were dried in a laminar flow hood then incubated in a moist chamber at 28˚C without light for 5 days before interior fruit structures were assessed for fungal colonization. An additional set of tomato and cucumber fruit handled in the same way were inoculated with F. oxysporum f. sp. lycopersicum isolate Foly146 and F. oxysporum f. sp. cucumerinum isolate Foc0812, respectively.

Assays of tomato and cucumber tissues for fungal colonization by specific primers
Primer 5/ST33-R, specific for amplification of HOFo isolates, was used to assay various plant parts for colonization following inoculation. Tomato and cucumber fruit inoculated with Foly146 and Foc0812 were assayed for colonization using primer FIGS11/12 [36]. Parts of inoculated tomato plants assayed included root, hypocotyl, cotyledon, and stem. One-cm-long tissue samples of each part were collected and sterilized by immersing in a mixture 95% ethanol and 5.25% NaClO (

Sensitivity of Taiwan HOFo isolates to agro-fungicides
In clinical medicine, azole fungicides such as fluconazole, itraconazole, voriconazole etc. are often used to treat human fungal diseases [11,39]. Azoles are also important for control of plant fungal diseases in agriculture environment (FRAC Code List © � 2019). We examined sensitivities of Taiwan HOFo isolates to two agro-fungicides, azoxystrobin (C3, FRAC code 11) and difenoconazole (G1, FRAC code 3). Sensitivities of these isolates were compared to those of international HOFo isolates from NRRL and F. oxysporum f. sp. tracheiphilum isolate Fot60 (Huang, personal communication). Three mm diameter agar discs from 7-day-old Taiwan HOFo cultures grown on PDA at 28˚C with 12 h day/night were transferred twice onto three replicate plates of PDA amended with 1, 10, 100 and 500 mg a.i./L of each fungicide and incubated at 28˚C without light. Inhibition of fungal growth was based on colony diameters after a 3-day incubation. The EC 50 for each fungicide was calculated using log-linear model software kindly supplied by Dr. Hsin Chi (Taiwan)

Identification and sensitivity of HOFo isolates by specific primers
Fourteen strains of HOFo intergenic spacer (IGS) region sequences obtained from NCBI Gen-eBank database were compared with the HOFo isolated from Taiwan and other Fo formae special isolates. Comparison of IGS nucleotide variation with other plant pathogenic and nonpathogenic Fo isolates using Clustal X 1.81 alignment software, two primers, designated Primer 5 (5' -GAGATATGTGGTTCAGGGTAGG-3') and ST33-R (5' -TTTCAGCTA CCCTAGAGTGATG-3'), were designed to identify and detect HOFo isolates (S1 Fig). The primer pair Primer 5/ST33-R was found to amplify a 642 bp DNA fragment from three HOFo isolates originally cultured from Taiwan and six international HOFo reference isolates (Fig 1 and Table 1). No products were obtained from DNA prepared other 16 formae speciales and 11 nonpathogenic isolates of F. oxysporum using Primer 5 and ST33-R. The primer pair Primer 5/ST33-R failed to amplify any DNA product from genomic DNA prepared from 51 different formae speciales and 11 nonpathogenic isolates of F. oxysporum, 9 HOFo isolates, 8 isolates of other Fusarium spp. and 3 isolates of Aspergillus spp. (Table 1). The PCR sensitivity result showed that the Primer 5 and ST33-R could amplify the 642 bp fragment from as little as 100 pg (10 −1 ng) template DNA in a 25 μl reaction mixture (Fig 2).   (Table 2). Fungal virulence was also assessed by drench inoculation. Tomato seedlings cv. 'Yu-Nyu' (cherry tomato) and cv. 'Farmers 301' (cooking tomato) at the 2-leaf stage were inoculated by soil drenching with conidial suspensions prepared from MCC2074. No visible symptoms were observed. The 642 bp DNA fragment could be amplified from roots and lower hypocotyl tissues of both cultivars with the Primer 5/ST33-R primer pair 1, 2, 3, and 4 weeks after inoculation ( Table 3). Amplification of the 642 bp DNA fragment from samples prepared from cotyledonary node and hypocotyls tissue below the node was inconsistent over the 4-week period. The 642 bp DNA fragment was amplified from stem tissues collected from 'Yu-Nyu' cherry tomato samples 3 and 4 weeks after inoculation. Fungal virulence was assessed by planting tomato seedlings in soils infested with MCC2074. The results revealed that the 642 bp DNA fragment was amplified from the roots of 'Yu-Nyu' cherry tomato plants 4, 8, and 12 weeks and from hypocotyl and cotyledonary node tissue 12 weeks after transplant ( Table 4).

Colonization of tomato and cucumber fruit by HOFo isolate MCC2074, F. oxysporum f.sp. lycopersici (Fol) and F. oxysporum f.sp cucumerinum (Foc)
Tomato and cucumber fruits were inoculated with conidial suspensions prepared from MCC2074, Fol, or Foc. PCR with the 5/ST33-R primer pair amplified the 642-bp DNA fragment from pericarp, locular cavity, and columella of tomato fruit and pericarp and locular cavity of cucumber fruit 5 days after inoculation (Table 5). Fol and Foc were detected solely from pericarp and locular cavity tissues of tomato and cucumber, respectively. No fungus was detected in seed samples taken from inoculated tomato or cucumber.

Sensitivity of HOFo isolates to agro-fungicides
Fungicide sensitivity assays revealed that HOFo isolates were more sensitive to difenconizol than to azoxystrobin than other test Fusarium strains (Table 6). Mycelial growth of international

PLOS ONE
Human opportunistic F. oxysporum isolates on plant isolates from NRRL was completely inhibited at 10 mg a.i./L of difenconizol whereas 100 ppm was required to completely inhibit the growth of Taiwan HOFo isolates. All test fungal strains grew poorly on medium amended with 1, 10, 100, or 500 mg a.i./L of azoxystrobin. However, it appeared that Taiwan HOFo isolates were less sensitive to azoxystrobin compared to the international isolates. Taiwan isolate Foc0812 and Fot60, isolated from the conventional farming area with cucumber and asparagus bean production farms in Nantou and Pingtung county, respectively, displayed sensitivity to both test fungicides at levels similar to the international isolates.

Discussion
The sequence types show that the three Taiwan HOFo isolates belong to the sequence type ST33. Short et al. [37] reported that ST33 is the predominant clinical type of F. oxysporum Cumber and tomato fruits were laid paper disks (0.2 cm) with 50 μl fungal spore suspension (1x10 6 spores/ml) 2 The primers FIGS11/12 amplified a 700 bp DNA product of Foc0812 and Foly146, and the Primer 5/ ST33-R could amplified 650 bp DNA product of MCC2074. The symbol "+" means that target isolate could be detected from this part tissues of all three replicate samples, the symbol "−" means that no PCR product of target isolate were amplified of all three replicate samples, and the symbol "+/−"means that target isolate could be detected from this part tissues of some replicate samples. 3 "P" = Pericarp wall; "L" = Locular cavity; "C" = Columella; "S" = Seeds https://doi.org/10.1371/journal.pone.0234517.t005

PLOS ONE
species complex (FOSC). Although only three isolates were examined, we suspect that clinical ST33 is the predominant type in Taiwan. In this study, the primer pair (Primer 5/ST33-R) has been shown to be highly specific to the ST33 isolates. Other studies have demonstrated that sequence types could be differentiated by TEF-1α and IGS regions [37,40]. The two regions displaying high levels of nucleotide diversity are suitable for fungal identification, classification and specific primer design [36,41]. Here, the entire IGS region of the HOFo isolates was sequenced and compared with IGSs of other F. oxysporum. Consequently, two primers Primer 5 and ST33-R were designed to detect the HOFo isolates from Taiwan and ATCC isolates. The primers show higher degrees of specificity because they are unable to amplify any products from genomic DNA prepared from other formae speciales or nonpathogenic F. oxysporum. Tomato plants inoculated with the clinical HOFo isolate MCC2074 show no visible disease symptoms although the fungal isolate is able to colonize tissues in the lower part of tomato plants. Of three inoculation methods tested, hypocotyl inoculation after the roots are cut off results in the most rapid and extensive colonization of the fungus in the lower stem sections above the cotyledonary node (Table 2). Although the method of inoculation affects the rate and the extent of tissue colonization, the overall data show that colonization by the isolate MCC2074 is restricted to the lower part of tomato plants. This suggests that MCC2074 fails to move systemically through the vascular system as F. oxysporum f. sp. lycopersici does. Previous studies have shown that proteins secreted-in-xylem (SIX) are closely related to pathogenicity of F. oxysporum f. sp. lycopersici [42,43]. The primer pair SIX6-F/SIX6-R [44] fails to amplify the SIX6-coding gene (Chung, unpublished data) in any of the HOFo isolates used in this study, suggesting that the clinical HOFo isolates likely do not have the ability to cause diseases in plants. Nonpathogenic F. oxysporum has been shown to colonize crop tissues because of the lack of the SIX6-coding gene [36,45]. Fusarium spp. could be endophytic fungi commonly found in many different plants [46,47], and it has been noted that, if inoculum is available, there is a risk for human pathogens or human opportunistic pathogens to contaminate plant products or enter into plant tissues [48,49].
In this study, the tomato and cucumber fruit with small surface wounds were extensively colonized following surface inoculation with HOFo isolate MCC2074. Although interior tissues of the fruit were colonized, they exhibited no disease symptoms. These results show that there is a potential for symptomless tomato and cucumber fruit intended for human consumption to be infected with HOFo isolates. Our study also indicated that the tested HOFo isolates could survive in soils with chlamydospores. Previous studies have demonstrated that the human bacterial [48,49,50] or fungal [51,52] pathogens could adhesive on the surface of plants by the contamination of animal excrement, plant debris, and soils. Although the HOFo MCC2074 isolate could not cause symptoms in plants or fruit, the MCC2074 has the ability to overcome the physical and chemical barriers of unsusceptible plants. The HOFo isolates could directly penetrate or through natural openings into plant or fruit tissue [1,27]. Springer et al. [51] and Xue et al. [52] reported that the human pathogen Cryptococcus could colonize inside of the living plant tissues, such as leaves and vascular, and the characteristics play an important role for Cryptococcus to complete sexual cycles and produce infectious propagules. If HOFo could persist in plant tissue, they might have chances to encounter with pathogenic F. oxysporum or nonpathogenic F. oxysporum, raising a concern about gene flow or gene exchanging [53]. Thus, plants serve as a platform where HOFo may contact with other F. oxysporum strains. In addition, the possibility of HOFo isolates to colonize in plant or fruit tissue may increase the risk for immunosuppressed patients who come to contact or eat the contaminated fruits and plants.
The fungicidal sensitivity of HOFo isolates to difenoconazole and azoxystrobin indicated that HOFo isolates were more tolerant to azoxystrobin than difenoconazole. This is interesting as azoxystrobin is an agricultural fungicide belonging to aryloxypyrimidine, which has a 4,6-diphenoxypyrimidine skeleton. Azoxystrobin impacts mitochondrial respiration by interrupting electron transfer between cytochromes b and c1 [54]. Difenoconazole is also an agricultural fungicide belonging to triazoles, which are demethylation inhibitor (FRAC Code List © � 2019). Triazoles are commonly used to treat human fungal diseases [11,39]. Resistance to azole drugs has been reported in certain human fungal pathogens [10]. Thus, the drug resistance of HOFo isolates to difenoconazole should be higher than azoxystrobin. However, the results show an opposite reaction between difenoconazole and azoxystrobin. This result indicates that the azoxystrobin-resistance might be inherent in HOFo isolates, and this phenomenon has been observed in F. graminearum [causing Fusarium head blight] with natural resistance to trifloxystrobin in Europe [55]. Thus, the HOFo isolates show a lower sensitivity to fungicides might also be associated with natural resistance. Previous studies have revealed that the strobilurin-resistance is associated with the point mutation at codons 129 and 143 in the cyt b gene [56][57][58] and alternative oxidase (AOX) [59][60][61]. The resistant mechanism warrant more research in future. In this study, the sensitivities of the Taiwanese HOFo isolates to azoxystrobin and difenoconazole at 1 and 10 mg a.i./L are lower than the USA HOFo isolates. These results demonstrate that the drug reaction of the HOFo isolates might dependent on the geographic conditions. The Taiwanese HOFo isolates might have the chance to expose to triazoles with a higher concentration over a long period of time than the USA isolates. It is necessary to collect more Taiwanese isolates to confirm the sensitivity to difenoconazole and other triazole fungicides.