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, 2004, 2019; de 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).

Species belonging to Fusarium are distributed into several species complexes (SC), some of which are important in human and veterinary mycology, particularly F. solani SC, F. oxysporum SC, F. fujikuroi SC, F. dimerum, F. chlamydosporum, F. incarnatum-equiseti, and F. sporotrichoides (van Diepeningen et al., 2014; Salah et al., 2015; Al-Hatmi et al., 2016a; Hassan et al., 2016). Fusarium graminearum, F. culmorum, F. fujikuroi SC, F. solani SC, and F. oxysporum SC may additionally be found as plant pathogens in maize, wheat, rice, soybean, and tomato crops (Basler, 2016; Costa et al., 2016; Kim et al., 2016; Manzo et al., 2016). Some Fusarium species produce mycotoxins during growth in plant tissue, which may contaminate cereal grains and derivatives, making them unsuitable for consumption and causing great agricultural losses (Milicevic et al., 2010; Sobrova et al., 2010).

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 (Meis et al., 2016; 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., 2012, 2013; Meis et al., 2016; Alvarez-Moreno et al., 2017).

In the medical field, amphotericin B, voriconazole, and posaconazole are the main antifungal drugs recommended for prophylaxis and treatment of human fusariosis (Lortholary et al., 2010; Tortorano et al., 2014; Clark et al., 2015; Nucci et al., 2015; Taj-Aldeen et al., 2016; Al-Hatmi et al., 2018b). Most Fusarium species exhibit high minimal inhibitory concentrations (MICs) to currently used antifungals, especially azoles (Katiyar and Edlind, 2009; Fan et al., 2013; Al-Hatmi et al., 2015).

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.

Materials and Methods

Strains and Clinical Data

Forty-three clinical Fusarium isolates were available from the Laboratory of Mycology at the Federal University of Paraná Hospital, Curitiba, Brazil and Federal University of Rio de Janeiro Hospital, Rio de Janeiro, Brazil, recovered from 40 patients cared between 1985 and 2015. Three patients (32, 36, and 38) had each two isolates recovered, as specified in the Table 1. The patient’s medical records were reviewed to collect minimal clinical information such as age, gender, treatment, and outcome.

Table 1

Fusarium isolates data.

Isolate no.	Species complex	Species	Patient	Type of fusariosis	Underlying disease	Source	Treatment	GenBank accession no.
Fu02	FSSC25	Fusarium sp.	1	Disseminated	Unknown	Blood	VOR	MG738163
Fu14	FSSC2	F. keratoplasticum	2	Disseminated	AML	Skin	VOR	MG738189
Fu27	FSSC2	F. keratoplasticum	3	Cutaneous	Arterial insufficiency on legs	Skin	VOR	MG738193
Fu34	FSSC5	F. solani s.s.	4	Keratitis	None	Eye	VOR	MG738195
Fu37	FSSC2	F. keratoplasticum	5	Cutaneous	None	Skin	VOR	MG738184
Fu50	FSSC1	F. petroliphilum	6	Disseminated	Myelodysplasia	Skin	AMB	MG738167
Fu51	FSSC1	F. petroliphilum	7	Disseminated	AML	Blood	FLU	MG738168
Fu56	FFSC	F. napiforme	8	Cutaneous	Fanconi anemia	Blood	VOR + AMB	MG738202
Fu66	FSSC3+4	F. falciforme	9	Keratitis	None	Eye	VOR	MG738197
Fu71	FFSC	F. verticillioides	10	Disseminated	AML	Skin	VOR	MG738201
Fu72	FSSC7	N. gamsii	11	Cutaneous	ALL	Blood	VOR	MG738177
Fu73	FSSC7	N. gamsii	12	Disseminated	Non-Hodgkin lymphoma	Skin	VOR + AMB	MG738178
Fu75	FSSC1	F. petroliphilum	13	Keratitis	None	Eye	not done	MG738169
Fu77	FSSC2	F. keratoplasticum	14	Disseminated	Purpura amegakaryocytic	Skin	not done	MG738190
Fu78	FFSC	F. subglutinans	15	Disseminated	Aplastic anemia	Blood	VOR + AMB	MG738203
Fu80	FSSC7	N. gamsii	16	Unknown	Unknown	Skin	Unknown	MG738179
Fu86	FSSC25	Fusarium sp.	17	Unknown	Unknown	Skin	Unknown	MG738164
Fu87	FSSC	Fusarium sp.	18	Cutaneous	ALL	Blood	VOR	MG738166
Fu89	FSSC35	Fusarium sp.	19	Disseminated	Unknown	Blood	VOR	MG738162
Fu92	FSSC1	F. petroliphilum	20	Cutaneous	Aplastic anemia	Skin	VOR + AMB	MG738170
Fu93	FSSC20	N. suttoniana	21	Disseminated	CML	Skin	VOR + AMB	MG738198
Fu94	FSSC	F. xanthoxyli	22	Disseminated	Unknown	Skin	VOR	MG738182
Fu96	FSSC2	F. keratoplasticum	23	Disseminated	ALL	Skin	VOR	MG738185
Fu97	FSSC2	F. keratoplasticum	24	Disseminated	ALL	Endotracheal aspirate	VOR + AMB	MG738194
Fu99	FSSC1	F. petroliphilum	25	Cutaneous	Aplastic anemia	Skin	VOR + ISA	MG738171
Fu100	FSSC20	N. suttoniana	26	Keratitis	None	Eye	VOR	MG738199
Fu101	FSSC2	F. keratoplasticum	27	Disseminated	Myocardium revascularization	Skin	VOR	MG738183
Fu103	FSSC20	N. suttoniana	28	Keratitis	None	Eye	VOR	MG738200
Fu105	FSSC2	F. keratoplasticum	29	Disseminated	Myelodysplasia	Skin	VOR	MG738191
Fudm2	FSSC7	N. gamsii	30	Disseminated	ALL	Blood	VOR + AMB	MG738180
FuB302.1	FSSC2	F. keratoplasticum	31	Unknown	Rheumatoid arthritis	Skin	VOR	MG738192
FuB371	FSSC5	F. solani s.s.	32	Unknown	ALL	Skin	VOR	MG738196
FuB391	FSSC33	F. pseudensiforme	33	Unknown	Unknown	Skin	Unknown	MG738161
FuB478	FSSC2	F. keratoplasticum	34	Unknown	AML	Skin	AMB	MG738186
FuB560	FSSC7	N. gamsii	35	Unknown	CML	Skin	VOR + lipid AMB	MG738181
FuB604	FSSC1	F. petroliphilum	36	Unknown	ALL	Synovial fluid	VOR	MG738172
FuB665	FSSC1	F. petroliphilum	36	Unknown	ALL	Synovial fluid	VOR	MG738173
FuB817	FSSC1	F. petroliphilum	37	Unknown	Myelodysplasia	Skin	VOR	MG738174
FuB920	FSSC1	F. petroliphilum	32	Unknown	ALL	Synovial fluid	VOR	MG738175
FuB935	FSSC2	F. keratoplasticum	38	Unknown	AML	Skin	VOR + lipid AMB	MG738187
FuB936	FSSC2	F. keratoplasticum	38	Unknown	AML	Skin	VOR + lipid AMB	MG738188
FuH79A	FSSC18	Fusarium sp.	39	Unknown	AML	Blood	VOR + lipid AMB	MG738165
FuH05	FSSC1	F. petroliphilum	40	Unknown	Unknown	Blood	Unknown	MG738176

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.

Fragments of the translation elongation factor 1-alpha (TEF1α) gene were amplified and sequenced using PCR protocols following the methods published by Al-Hatmi et al. (2014) with primers EF1 (5′-ATGGGTAAGGA(A/G)GACAAGAC-3′) and EF2 (5′-GGA(G/A)GTACCAGT(G/C)ATCATGTT-3′) (O’Donnell et al., 1998). Sequencing reaction mixtures contained 1 ng/μL of template DNA, 1 pmol/μL, 0.7 μL of BigDyeTM terminator (Applied Biosystems, Foster City, CA, United States), 3 μL buffer and ultra-pure water to 10 μL final volume. Sequencing PCR was performed as follows: 95°C for 1 min, followed by 30 cycles consisting of 95°C for 10 s, 50°C for 5 s and 60°C for 2 min. Sequencing was done on an ABI 3730xL automatic sequencer (Applied Biosystems).

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) and the Fusarium MLST (O’Donnell et al., 2010) database down to species and haplotype level. DNA sequences were edited, and consensus sequences were assembled by the SeqMan package of Lasergene software (DNAStar, Madison, WI, United States). Retrieved alignments were manually corrected to avoid mis-paired bases. Sequences were exported as FASTA files. Sequences of TEF1α were aligned with MAFFT program[1] and adjusted in MEGA6 (Tamura et al., 2013). The best-fit model of evolution was determined by MEGA6. Maximum likelihood (ML) analysis was done with RAxML-VI-HPC v. 7.0.3 with non-parametric bootstrapping using 1000 replicates. GenBank accession numbers are shown in Table 1.

Antifungal Susceptibility Testing

Antifungal susceptibility testing by the broth microdilution method was performed according to the CLSI protocol M38-A2 (Clinical and Laboratory Standards Institute [CLSI], 2008). Antifungal agents tested were amphotericin B (Bristol Myers Squibb, Woerden, Netherlands), itraconazole (Janssen Pharmaceutica, Beerse, Belgium), voriconazole (Pfizer, Sandwich, United Kingdom), posaconazole (Merck, NJ, United States) and isavuconazole (Basilea Pharmaceutica, Basel, Switzerland). The fungicides used were difenoconazole, tebuconazole and propiconazole (all from Sigma-Aldrich, St. Louis, MO, United States). The concentrations of antifungals ranged from 0.031 to 32 μg mL-1. Fusarium isolates were cultured onto Sabouraud glucose agar until sporulation at 30°C and the inocula were adjusted to 1.8–3 × 106 CFU/mL in saline supplemented with 0.05% Tween 20 to perform the test. Microdilution plates were incubated at 35°C for 48 h and the MICs were defined as the lowest concentration able to complete growth inhibition when compared with the drug free growth control. Aspergillus flavus ATCC 204304, Candida parapsilosis ATCC 22019 and C. krusei ATCC 6258 reference strains were used as quality controls (Clinical and Laboratory Standards Institute [CLSI], 2008). Interpretation of the MIC values was based on Epidemiological Cutoff Values (ECV) according to previous literature data (Espinel-Ingroff et al., 2016). MIC50 and MIC90 were obtained by ordering the data for each antifungal in ascending order and selecting the median and 90th quantile, respectively. Geometric mean MICs were calculated using Microsoft Office Excel 2010 software (Microsoft, Redmond, WA, United States). When the MIC was more or less than dilutions tested, 1 log2 dilution higher or 1 log2 dilution lower was considered for calculating the geometric mean.

Results

Clinical Data

The median age of the 40 patients was 36 years (range 2–78 years) and 21 were female. Disseminated fusariosis was the most frequent clinical form (n = 16, 37.2%), followed by cutaneous infections (n = 7; 16.3%) and keratitis (n = 5; 11.6%). Fusarium strains were isolated most frequently from the skin (n = 24; 55.8%), blood (n = 10; 23.2%), and eye (n = 5; 11.6%). Acute lymphoblastic leukemia (n = 7; 16.3%) and acute myeloid leukemia (n = 6; 13.9%) were the most commonly underlying conditions. Twelve out of 16 cases of disseminated fusariosis occurred in patients with hematological malignancies. Voriconazole monotherapy was the treatment in 21 (48.8%) patients, 13 of which (61.9%) had a favorable response to therapy. Combination therapy with voriconazole and deoxycholate amphotericin B was given to 7 (16.3%) patients, and voriconazole plus liposomal amphotericin B in 3 patients (7%). Other therapies were deoxycholate amphotericin B alone (n = 2; 4.7%), fluconazole alone (n = 1; 2.3%), and voriconazole associated with itraconazole (n = 1; 2.3%). For 2 (4.7%) patients no therapy was given. Information about treatment was not available in 6 cases. The isolates and respective patients’ clinical data are shown in Table 1.

Molecular Identification and Phylogeny

Phylogenetic analysis based on TEF1α sequences was conducted in order to position the isolates in the Fusarium solani complex and their respective species complexes (Figure 1). The analysis included 55 sequences from different species, and one outgroup taxa (NRRL 22316 F. staphyleae). Within FSSC, F. keratoplasticum FSSC 2 (n = 12) was most often involved in cases of fusariosis, followed by F. petroliphilum FSSC 1 (n = 10), Neocosmospora 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. FSSC 18 (n = 1), F. falciforme FSSC 3+4 (n = 1), F. pseudensiforme (n = 1), and F. solani f. xanthoxyli (n = 1). One isolate clustered in a separate clade (unknown species/haplotype) forming a distinct, well-supported, unnamed lineage and which matched only with a single strain from Colombia (LEMM 110739, GenBank accession no. LN827969, misidentified as Fusarium solani). We also identified the following members of the Fusarium fujikuroi species complex (FFSC): F. subglutinans (n = 1), F. verticillioides (n = 1), and F. napiforme (n = 1) which are not included in the phylogenetic analysis.

FIGURE 1

Maximum likelihood (ML) phylogenetic tree created from TEF1 sequences of 55 Fusarium sequences. Total alignment length is 707 bp. Thousand bootstrap-replications. The tree was rooted with NRRL 22316 F. staphyleae.

Antifungal Susceptibility Profiles

MICs are shown in Tables 2, 3. Amphotericin B had relatively high activity with MICs ranging from 0.5 to 32 μg mL-1, except for the isolates Fu73 (novel lineage) and Fu80 (Neocosmospora gamsii FSSC7), which showed MIC values of 8 and 32 μg mL-1, respectively. All isolates exhibited high MICs to itraconazole with MICs >32 μg mL-1. The FSSC had MIC values of posaconazole and difenoconazole higher than 32 μg mL-1. Other azoles showed to be less effective against FSSC isolates with high MIC values of 8–>32 μg mL-1. Fusarium keratoplasticum showed high MIC values (8–>32 μg mL-1) for itraconazole, voriconazole, posaconazole and isavuconazole. In counterpart, azoles showed activity against FFSC with MIC values ranges of 1–8 μg mL-1 and with only one isolate of F. napiforme showing MIC of >32 μg mL-1 for posaconazole.

Table 2

Minimal inhibitory concentrations of Fusarium clinical isolates.

Species complex	Antifungal	No. of isolates per MIC value (μg mL-1)
		0.031	0.062	0.125	0.25	0.5	1	2	4	8	16	32	>32
FSSC (n = 40)	Amphotericin B					10	16	9	4			1
	Itraconazole												40
	Voriconazole							1	2	21	7	3	6
	Posaconazole												40
	Isavuconazole											10	30
	Difenoconazole												40
	Tebuconazole										3	4	33
	Propiconazole										1		39
FFSC (n = 3)	Amphotericin B						1	2
	Itraconazole												3
	Voriconazole							2	1
	Posaconazole					1	1						1
	Isavuconazole								3
	Difenoconazole								1	2
	Tebuconazole							1	2
	Propiconazole							1		2

Table 3

Individual minimal inhibitory concentration (μg mL-1) of all Fusarium spp. and Neocosmospora spp. isolates.

Isolate	Identification – EF	Minimal inhibitory concentration (μg mL-1)
		AMB	ITC	VOR	POS	ISA	DIF	TEB	PRO
Fu101	F. keratoplasticum (FSSC 2)	4	64	16	64	64	64	64	64
Fu77	F. keratoplasticum (FSSC 2)	2	64	16	64	64	64	64	64
Fu14	F. keratoplasticum (FSSC 2)	0.5	64	64	64	64	64	64	64
Fu105	F. keratoplasticum (FSSC 2)	2	64	8	64	64	64	64	64
FuB302.1	F. keratoplasticum (FSSC 2)	2	64	16	64	64	64	64	64
Fu97	F. keratoplasticum (FSSC 2)	1	64	16	64	64	64	64	64
Fu27	F. keratoplasticum (FSSC 2)	4	64	64	64	64	64	64	64
FuB936	F. keratoplasticum (FSSC 2)	1	64	8	64	64	64	64	64
Fu96	F. keratoplasticum (FSSC 2)	1	64	8	64	32	64	16	64
FuB935	F. keratoplasticum (FSSC 2)	2	64	8	64	64	64	64	64
Fu37	F. keratoplasticum (FSSC 2)	1	64	32	64	64	64	64	64
FuB478	F. keratoplasticum (FSSC 2)	2	64	8	64	64	64	64	64
	Range∗	0.5–4	64	8–64	64	32–64	64	16–64	64
	MIC50∗	2	64	16	64	64	64	64	64
	MIC90∗	4	64	64	64	64	64	64	64
	Geometric mean∗	1.58	64	16	64	60.4	64	57.01	64
FuB920	F. petroliphilum (FSSC 1)	1	64	8	64	64	64	64	64
Fu92	F. petroliphilum (FSSC 1)	1	64	8	64	32	64	64	64
Fu50	F. petroliphilum (FSSC 1)	2	64	4	64	32	64	64	64
Fu51	F. petroliphilum (FSSC 1)	0.5	64	16	64	64	64	64	64
Fu99	F. petroliphilum (FSSC 1)	1	64	8	64	32	64	32	64
Fu75	F. petroliphilum (FSSC 1)	0.5	64	8	64	64	64	64	64
FuB665	F. petroliphilum (FSSC 1)	1	64	8	64	64	64	64	64
FuB817	F. petroliphilum (FSSC 1)	0.5	64	8	64	32	64	64	64
FuB604	F. petroliphilum (FSSC 1)	1	64	8	64	32	64	64	64
FuH05	F. petroliphilum (FSSC 1)	0.5	64	8	64	64	64	64	64
	Range∗	0.5–2	64	4–16	64	32–64	64	32–64	64
	MIC50∗	1	64	8	64	32	64	64	64
	MIC90∗	1	64	8	64	64	64	64	64
	Geometric mean∗	0.81	64	8	64	45.25	64	59.71	64
Fu72	N. gamsii (FSSC 7)	1	64	64	64	64	64	64	64
FuB560	N. gamsii (FSSC 7)	2	64	8	64	32	64	32	64
Fudm02	N. gamsii (FSSC 7)	4	64	8	64	64	64	64	64
Fu80	N. gamsii (FSSC 7)	32	64	8	64	64	64	64	64
Fu73	N. gamsii (FSSC 7)	4	64	8	64	64	64	64	64
Fu93	N. suttoniana (FSSC 20)	1	64	64	64	64	64	64	64
Fu100	N. suttoniana (FSSC 20)	0.5	64	64	64	64	64	64	64
Fu103	N. suttoniana (FSSC 20)	1	64	32	64	64	64	64	64
Fu02	Fusarium sp. (FSSC 25)	0.5	64	32	64	64	64	32	64
Fu86	Fusarium sp. (FSSC 25)	1	64	8	64	64	64	64	64
FuB371	F. solani sensu stricto (FSSC 5)	2	64	8	64	64	64	64	64
Fu34	F. solani sensu stricto (FSSC 5)	0.5	64	16	64	64	64	64	64
FuH79A	Fusarium sp. (FSSC 18)	2	64	8	64	64	64	32	64
Fu89	Fusarium sp. (FSSC 35)	0.5	64	16	64	64	64	64	64
Fu87	Fusarium sp.	1	64	4	64	32	64	64	64
Fu66	F. falciforme (FSSC 3+4)	0.5	64	2	64	32	64	16	16
Fu94	F. solani f. xanthoxyli	1	64	64	64	64	64	64	64
FuB391	F. pseudensiforme	1	64	8	64	32	64	16	64
Fu78	F. subglutinans	1	64	2	0.5	4	8	4	2
Fu71	F. verticillioides	2	64	4	1	4	8	4	8
Fu56	F. napiforme	2	64	2	64	4	4	2	8

Among the agricultural fungicides, difenoconazole had the lowest activity against FSSC with MICs of >32 μg mL-1 for all 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., 2003, 2013; Nucci 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 40 cases and responsible for 93% of the fusariosis in this study. In addition, three species were identified as belonging to the fujikuroi complex (7%). Of the 12 species and haplotypes of FSSC where the 40 strains were distributed, six belonged to previously described Fusarium species or varieties (F. keratoplasticum, F. falciforme, F. petroliphilum, Fusarium solani sensu stricto, F. ambrosium, and F. solani f. xanthoxyli), three to known haplotypes (FSSC 25, FSSC 35, FSSC 18, respectively), while two clades were recently described in Neocosmospora [FSSC 7 = N. gamsii, FSSC 20 = N. suttoniana (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 F. pseudensiforme (FSSC 33), F. solani f. xanthoxyli (FSSC 22), N. gamsii (haplotype 7 – FSSC 7), N. suttoniana (haplotype 20 – FSSC 20), Fusarium sp. (FSSC 25), and Fusarium sp. (FSSC 35) (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 (Al-Hatmi et al., 2015, 2016b; Taj-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.

In view of the resistance of Fusarium spp. to several antifungal agents, some studies have tested its susceptibility to new antifungals. Abastabar et al. (2018) tested luliconazole, lanoconazole, and efinaconazole against clinical and environmental Fusarium isolates members of the F. fujikuroi species complex (n = 94), F. solani species complex (n = 14), F. oxysporum species complex (n = 11), F. lateritium species complex (n = 1), and F. graminearum species complex (n = 1). Overall, Fusarium species demonstrated lower MICs to luliconazole, lanoconazole and efinaconazole (geometric mean MICs of 0.005, 0.013, and 0.85 μg mL-1, respectively) when compared with voriconazole and amphotericin B (geometric mean MICs of 1.37 and 1.9 μg mL-1, respectively). In addition, Tupaki-Sreepurna et al. (2017b) tested the susceptibility of F. solani species complex (n = 18), F. dimerum species complex (n = 2), and F. incarnatum-equiseti species complex (n = 1) to efinaconazole. The concentrations of efinaconazole necessary to inhibited fungal growth vary from 0.031 to 2 μg mL-1, with geometric mean MICs varying from 0.08 to 0.7 μg mL-1 depending on Fusarium species. These data suggested that luliconazole, lanoconazole and efinaconazole are effective drugs that may be used against fusariosis.

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.

Conflict of Interest Statement

JM received grants from F2G and Merck. He has been a consultant to Scynexis and Merck and received speaker’s fees from Merck, United Medical, TEVA and Gilead Sciences. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.