Molecular Characterization and Antifungal Susceptibility of Clinical Fusarium Species From Brazil

Fusarium is widely distributed in the environment and is involved with plant and animal diseases. In humans, several species and species complexes (SC) are related to fusariosis, i.e., F. solani SC, F. oxysporum SC, F. fujikuroi SC, F. dimerum, F. chlamydosporum, F. incarnatum-equiseti, and F. sporotrichoides. We aimed to investigate the susceptibility of Fusarium clinical isolates to antifungals and azole fungicides and identify the species. Forty-three clinical Fusarium isolates were identified by sequencing translation elongation factor 1-alpha (TEF1α) gene. Antifungal susceptibility testing was performed to the antifungals amphotericin B, itraconazole, voriconazole, posaconazole, and isavuconazole, and the azole fungicides difenoconazole, tebuconazole, and propiconazole. The isolates were recovered from patients with median age of 36 years (range 2–78 years) of which 21 were female. Disseminated fusariosis was the most frequent clinical form (n = 16, 37.2%) and acute lymphoblastic leukemia (n = 7; 16.3%) was the most commonly underlying condition. A few species described in Fusarium solani SC have recently been renamed in the genus Neocosmospora, but consistent naming is yet not possible. Fusarium keratoplasticum FSSC 2 (n = 12) was the prevalent species, followed by F. petroliphilum FSSC 1 (n = 10), N. gamsii FSSC 7 (n = 5), N. suttoniana FSSC 20 (n = 3), F. solani sensu stricto FSSC 5 (n = 2), Fusarium sp. FSSC 25 (n = 2), Fusarium sp. FSSC 35 (n = 1), Fusarium sp. FSSC18 (n = 1), F. falciforme FSSC 3+4 (n = 1), F. pseudensiforme (n = 1), and F. solani f. xanthoxyli (n = 1). Amphotericin B had activity against most isolates although MICs ranged from 0.5 to 32 μg mL-1. Fusarium keratoplasticum showed high MIC values (8–>32 μg mL-1) for itraconazole, voriconazole, posaconazole, and isavuconazole. Among agricultural fungicides, difenoconazole had the lowest activity against FSSC with MICs of >32 μg mL-1 for all isolates.


INTRODUCTION
The fungal genus Fusarium is widely distributed as saprobes in the environment but is also able to cause cross-kingdom disease in both plants and mammals (Gauthier and Keller, 2013;van Diepeningen and de Hoog, 2016). In humans, the disease may manifest in different ways, depending on the portal of entry and the host's immune status. Invasive fusariosis is the most severe manifestation that predominantly affects immunocompromised hosts with hematological malignancies, neutropenia, or glucocorticoid exposure (Nucci et al., 2003(Nucci et al., , 2004(Nucci et al., , 2019de Souza et al., 2014). In immunocompetent hosts, the fungus may cause onychomycosis (Guevara-Suarez et al., 2016), keratitis (Tupaki-Sreepurna et al., 2017a) or other (sub)cutaneous disorders. The most frequent fungal diseases caused by Fusarium species are onychomycosis and keratitis, although other clinical presentations are also observed, such as fungemia, mycetoma, skin infection, lung disease (including allergic disease, hypersensitivity pneumonitis, colonization of a pre-existing cavity, pneumonia in severely immunocompromised patients), and other rare infections (endocarditis, urinary tract infection, osteomyelitis, etc.) (Sierra-Hoffman et al., 2005;Su et al., 2007;Nucci et al., 2015;Kassar et al., 2016).
In attempts to reduce agricultural losses caused by fungal diseases, many strategies have been used, including augmentation of plant resistance, spraying of chemicals, biological control, integrated disease management (Singh et al., 2016), and fungicide use, especially azoles (Hof, 2001). The continuing uncontrolled use of fungicides may lead to selective pressure on environmental fungi (Deising et al., 2008). Due to the structural similarity of azoles used in agriculture and medicine, cross-resistance may be observed in clinical fungi Verweij et al., 2016). Studies have been performed to test the hypothesis whether fungicide use in agroecosystems may lead to antifungal resistance in Aspergillus fumigatus in the clinic (Snelders et al., 2008;Chowdhary et al., 2012Chowdhary et al., , 2013Meis et al., 2016;Alvarez-Moreno et al., 2017).
Here we aimed to investigate the susceptibility of Fusarium clinical isolates to commonly used antifungals and fungicides and identify the species. For this study, we used strains that were isolated from patients with fusariosis diagnosed in two tertiary Brazilian hospitals in southern Brazil.

DNA Isolation, PCR, and Sequencing
Fusarium isolates were cultured on Sabouraud dextrose agar plus chloramphenicol (SDA; Difco Laboratories, Detroit, MI, United States). Culture plates were incubated at 26 and 37 • C and observed daily for growth up to 7 days. Initial identification of Fusarium isolates was based on macroscopic colony morphology and microscopic features in a lacto-phenol wet mount preparation according to standard laboratory procedures. Final identification was done using molecular methods. DNA extraction was performed as described by Khodavaisy et al. (2016). Conidia were suspended in 400 µL bacterial lysis buffer (Roche Diagnostics, Almere, Netherlands) followed by mechanical lysis in a MagNA Lyser (Roche Diagnostics) for 30 s at 4,500 × g. Cells were inactivated for 10 min by heating at 100 • C and 200 µL of the solution was used for automated DNA extraction by using the MagNA Pure 96 platform (Roche Diagnostics) with a final elution volume of 100 µL.

Alignment and Phylogenetic Analyses
For preliminary identification, a homology search for the sequences of TEF1α was done using the BLAST tool in NCBI database, the CBS database, FUSARIUM-ID (Geiser et al., 2004)

Antifungal Susceptibility Profiles
MICs are shown in Tables  isolates, followed by propiconazole and tebuconazole. In contrast, the three fungicides showed activity against FFSC, with MIC ranges of 2-8 µg mL −1 .

DISCUSSION
Invasive fusariosis is a severe disease that affects immunocompromised patients, mostly those with underlying hematological malignancies (Nucci et al., 2003Nucci and Anaissie, 2007;Campo et al., 2010;Carlesse et al., 2017). In agreement with the literature, the present study found the majority of disseminated cases of fusariosis (11/16) occurring in patients with acute lymphoblastic leukemia and acute myeloid leukemia. Disseminated fusariosis in these patients has a poor prognosis and mortality rates are close to 75% (Nucci and Anaissie, 2007;Campo et al., 2010). The treatment of this infection is a challenge and in the absence of better alternatives, voriconazole and amphotericin B are the most recommended therapies (Nucci and Anaissie, 2007;Nucci et al., 2014;Tortorano et al., 2014;Al-Hatmi et al., 2017).
Results from our sequence analysis show that twelve phylogenetic species within the solani complex were involved in  (Figure 1)]; note that according to these authors (Sandoval-Denis and Crous, 2018) the entire Fusarium solani species complex phylogenetically constitutes a separate genus, Neocosmospora, but not all extant species have consistently been denominated, resulting in the use of two generic names for closely related species. One strain (Fu87) was identified as a novel phylogenetic lineage within FSSC and matched with LEMM 110739, which was previously reported by Guevara-Suarez et al. (2016) from an onychomycosis case. Numerous haplotypes and the newly reported lineage have remained yet unnamed. In the present study, F. keratoplasticum (FSSC 2) was the most often recorded species (28%), followed by F. petroliphilum (FSSC 1, 23.3%), which agrees with data of O'Donnell et al. (2007). In accordance with literature data (O'Donnell et al., 2007;Walther et al., 2017) we also encountered Fusarium solani sensu stricto (FSSC 5) causing keratitis. Members of FSSC with a significant role in clinical infections in our data set comprised F. falciforme (FSSC 3+4), F. keratoplasticum (FSSC 2), F. lichenicola (FSSC 16), F. metavorans (FSSC 6), F. petroliphilum (FSSC 1), F. pseudensiforme (FSSC 33), and F. solani sensu stricto (FSSC 5) (Al-Hatmi et al., 2018a;Boral et al., 2018). Another lineage associated with opportunistic infections in FSSC that has been named is FSSC 27 (Phialophora cyanescens = Cylindrocarpon cyanescens), which was recently recombined as Neocosmospora cyanescens, MB 813864 (Summerbell and Scott, 2016). This species of FSSC lacks a name in Fusarium, while conversely F. solani f. xanthoxyli has no name in Neocosmospora; thus, consistent naming of the fungi in FSSC is impossible. Recently, a study from Japan also reported that haplotypes FSSC 9 and FSSC 18 are associated with opportunistic infections and with mycotic keratitis (Muraosa et al., 2017), while a German report found FSSC 9 and FSSC 25 to be involved in endophthalmitis (Walther et al., 2017). Literature data indicate that species within FSSC are the main cause of fusariosis worldwide (Scheel et al., 2013;Hassan et al., 2016;Tupaki-Sreepurna et al., 2017a). Fusarium keratoplasticum has been reported as the etiologic cause of disseminated fusariosis in hematologic patients (García-Ruiz et al., 2015;Chiewchanvit et al., 2017), as well as keratitis (Tupaki-Sreepurna et al., 2017a), onychomycoses (Guevara-Suarez et al., 2016;Gupta et al., 2016) and eumycetoma (Al-Hatmi et al., 2017). In addition, F. keratoplasticum is an important veterinary etiologic agent, causing disease in equine and marine vertebrates as well as in invertebrates (O'Donnell et al., 2016).
In the present study, we identified additional species and haplotypes for the first time from clinical samples, including  (Figure 1), but confirmed case reports are as yet lacking. All these haplotypes are phylogenetically distinct from described species but remain unnamed as molecular siblings. Our data suggest that these additional species/haplotypes might be of importance for human health, although on the other hand it remains questionable whether formal description of the FSSC lineages as formal species is meaningful. Using TEF1α sequences strain Fu87 matched with an undescribed lineage (LEMM 110739) previously reported by Guevara-Suarez et al. (2016) from clinical samples in Colombia. The number of reports of Fusarium species that were previously considered to be exclusive plant pathogens but are now implicated in superficial and systemic infections in humans and animals is obviously increasing (Zhang et al., 2006). Fusarium is rather unique in having pathogenic strategies to infect plants as well as animals including humans. This trans-kingdom pathogenicity has been demonstrated for the molecular siblings F. falciforme, F. keratoplasticum and F. solani sensu stricto within FSSC (Nalim et al., 2011;Short et al., 2013). Thus, our findings support the concept that Fusarium might serve as good model for studying the genetic basis of trans-kingdom pathogenicity in fungi (Ortoneda et al., 2004).
Our findings agree with reports from different regions in the world where the most frequently identified species causing human infections belonged to the FSSC followed by the fujikuroi and oxysporum species complexes , 2016bTaj-Aldeen et al., 2016). In Brazil species of FSSC were the most commonly reported, followed by the fujikuroi species complex (Scheel et al., 2013) and oxysporum species complex (Dallé da Rosa et al., 2018). Future studies including larger numbers of isolates are warranted to establish the prevalence of rare Fusarium species in clinical settings. In our study, F. keratoplasticum showed high MIC values (8->32 µg mL −1 ) for most azoles tested and agricultural fungicides, with geometric mean MICs of 1.58 µg mL −1 for amphotericin B, 16 µg mL −1 for voriconazole and 64 µg mL −1 for posaconazole, the most effective drugs against Fusarium species (Lortholary et al., 2016). Rosa et al. (2017) observed that F. keratoplasticum was the species most frequently found in onychomycoses lesions and was more susceptible to amphotericin B and voriconazole than the other antifungals tested, with geometric mean MICs of 4.88 and 20.09 µg mL −1 , respectively, higher than those observed in the present study. A study performed with 89 Fusarium isolates obtained from patients with superficial infections revealed that 49 (55.1%) of isolates belonged to F. solani species complex and 40 belonged to F. oxysporum species complex. Most of isolates showed high MIC values to antifungals tested, with modal MIC values of >16 µg mL −1 to amphotericin B, itraconazole, voriconazole, and posaconazole (Guevara-Suarez et al., 2016). Itraconazole had no in vitro effect against the isolates tested, which agrees with Tupaki-Sreepurna et al. (2017b). Similarly, Gupta et al. (2016) observed high MIC values of flucytosine, itraconazole, posaconazole, anidulafungin, and caspofungin for clinical isolates of F. keratoplasticum.

CONCLUSION
In conclusion, F. keratoplasticum and F. petroliphilum were the most frequent species in this study. Amphotericin B showed lower MICs against Fusarium species whereas the antifungal azoles and the fungicide difenoconazole exhibited higher MICs against FSSC.

ETHICS STATEMENT
Samples were collected during routine patient care and the study was retrospective, therefore it was determined by the local Institutional Review Board of the Hospital de Clínicas, Federal University of Paraná and CAPES that ethical clearance was not indicated.

AUTHOR CONTRIBUTIONS
PH, AA-H, FQ-T, and JM designed the study. PH and AA-H performed the experiments and wrote the first draft. RP, MM, MN, FQ-T, GH, and JM analyzed the data and revised the manuscript. All authors contributed to the writing and approved the final manuscript.

FUNDING
The work of PH was supported by Coordination for the Improvement of Higher Education Personnel, but is currently supported by "Instituto Nacional de Ciência e Tecnologia de Inovação em Doenças de Populações Negligenciadas."