ABSTRACT
Forty-eight Fusarium isolates morphologically identified as belonging to seven species of clinical interest (i.e., Fusarium chlamydosporum, Fusarium dimerum, Fusarium incarnatum, Fusarium napiforme, Fusarium nygamai, Fusarium proliferatum, and Fusarium sacchari) were characterized molecularly by the analysis of the sequences of the TUB region of the β-tubulin gene. F. chlamydosporum and F. dimerum were the most genetically heterogeneous species. A high degree of correlation between the morphological and molecular identification was shown among the isolates studied. A table with the key morphological features for the identification of these Fusarium species is provided. The antifungal susceptibilities of the Fusarium isolates to 11 antifungal drugs were tested; terbinafine was the most active drug against all the species tested with the exception of F. incarnatum, for which amphotericin B was the most active.
The most frequent species causing fusariosis are Fusarium solani, Fusarium oxysporum, and Fusarium verticillioides (1, 16, 47), but several other species are also found to cause human infections, although less frequently. Some of these species are Fusarium chlamydosporum, Fusarium dimerum, Fusarium incarnatum, and the following other species that are included into the Gibberella fujikuroi species complex: Fusarium napiforme, Fusarium nygamai, Fusarium proliferatum, and Fusarium sacchari (30, 31). These species have been associated with different types of infection, in particular with keratomycoses and other ocular infections (10) and with disseminated infections in immunocompromised patients (2, 6, 17, 20, 23, 24, 26, 39, 41, 43, 44). The real incidence of these species is unknown since they are poorly known and laboratorians and clinical microbiologists are not generally aware of their possible presence in human infections.
Since the species of Fusarium are generally resistant to all the available antifungal drugs (40), it could be considered that speciation of Fusarium is necessary only for epidemiological purposes. However, some in vitro data concerning particular species seem to be very promising and deserve to be investigated clinically. For instance, F. verticillioides isolates were susceptible to posaconazole and terbinafine and Fusarium thapsinum isolates to terbinafine (4). The identification of fusaria to the species level is not easy, and in numerous clinical cases the etiological agent is reported as being a Fusarium sp. However, several recent studies have demonstrated the usefulness of molecular methods for the identification of those Fusarium species that are difficult to distinguish morphologically (1, 4, 47). In recent years, the in vitro antifungal susceptibilities of the most frequent species of Fusarium have been evaluated (1, 3, 4, 40, 47), but only a few isolates of the less-common species have been studied. The objectives of our study were (i) to evaluate the correlation between the morphological and the molecular identification of less-frequent Fusarium species isolates received by our laboratory and (ii) to determine the antifungal susceptibilities of isolates representative of those less-common Fusarium species of clinical interest identified molecularly.
MATERIALS AND METHODS
Isolates and morphological identification.We included 48 Fusarium isolates in the study (Table 1). Twenty-nine of these isolates, which were mainly from clinical sources, were sent to our laboratory from different clinical centers for identification purposes. Those isolates were identified following the conventional morphological criteria (25, 28, 30). Briefly, to study the microscopic and colony features, the isolates were usually subcultured on potato-dextrose agar (PDA; Difco Laboratories, Detroit, MI) and on oatmeal agar (OA; 30 g oat flakes, 1 g MgSO4·7H2O, 1.5 g KH2PO4, 15 g agar, 1 liter tap water), incubated at 25°C in the dark. Sporodochia can be detected by examining the cultures under a stereoscopic microscope, usually after 7 days of incubation. The microscopic features were examined by making direct wet mounts with lactic acid from cultures on PDA and OA after 7 to 10 days of incubation. Growth rates were obtained from colonies on PDA at 25°C after 4 and 10 days of incubation in the dark. In addition, 19 reference strains from the Centraalbureau voor Schimmelcultures (CBS) and the Agricultural Research Service (ARS/NRRC) were used for a comparison.
Isolates included in this study and their origins
Molecular study.For the phylogenetic analysis, we sequenced the TUB region of the β-tubulin gene, which has proven to be highly phylogenetically informative in different molecular studies of the genus Fusarium (33, 34, 35). For DNA extraction, amplification, and sequencing, we followed the procedures previously described by Gilgado et al. (13), with some modifications. We used the primers TUB-F (9) and T22 (33) with an annealing temperature of 55°C, and the PCR products were purified using a GFX PCR DNA kit (Pharmacia Biotech, Cerdanyola, Spain). We included five reference sequences retrieved from GenBank corresponding to F. nygamai, F. napiforme, F. sacchari, Fusarium subglutinans, and F. proliferatum, respectively, since they are the only TUB sequences of the species included in the study that are available in the database (Table 1). The sequences were aligned using the ClustalX (version 1.8) computer program (46), followed by manual adjustments with a text editor. The phylogenetic analysis was performed using PAUP* version 4.0b10 (45). Maximum parsimony trees were obtained after 100 heuristic searches by using a random sequence addition and tree bisection-reconnection branch-swapping algorithms, collapsing zero-length branches, and saving all minimal-length trees (MulTrees).
Antifungal susceptibility study.We evaluated the in vitro activity of 11 antifungal drugs against 48 Fusarium isolates (Table 2), which were grown on PDA plates and incubated at 25°C for 7 days. The tests were performed by using a microdilution reference method (8), with some modifications. The inocula were adjusted to a final concentration of 4 × 103 to 5 × 104 conidia/ml with a hemocytometer and verified by quantitative colony counts on PDA plates. The antifungal agents tested included amphotericin B (AMB), albaconazole (ABC), voriconazole (VRC), itraconazole (ITC), ravuconazole (RVC), terbinafine (TBF), ketoconazole (KTC), posaconazole (PSC), micafungin (MFG), fluconazole (FLC), and flucytosine (5-FC). MFG, FLC, and 5-FC were diluted in sterile distilled water and the others in dimethyl sulfoxide. Final drug concentrations ranged from 64 to 0.12 μg/ml for FLC and 5-FC, from 128 to 0.25 μg/ml for MFG, and from 16 to 0.03 μg/ml for the rest. The MIC endpoints for AMB, TBF, and the triazoles ABC, ITC, PSC, RVC, and VRC were defined as the lowest concentration that produced complete inhibition of growth (MIC0). For FLC, KTC, 5-FC, and MFG, the endpoint was defined as the lowest concentration that produced 50% inhibition of growth (MIC50). Paecilomyces variotii ATCC 36257 was included as a quality control strain. The tests were performed twice on two different days and, in those cases where the results did not coincide, they were repeated a third time. For those strains, the MIC was considered the mode of the three MICs.
Activities of conventional and new antifungal drugs against isolates of seven Fusarium species of clinical interest
RESULTS
Correlation between morphological and molecular identification.The 29 Fusarium isolates received for identification were morphologically identified as F. chlamydosporum (n = 3), F. dimerum (n = 6), F. proliferatum (n = 8), and F. incarnatum (n = 8). For four of the isolates, it was difficult to differentiate between F. sacchari and F. subglutinans. However, based on the presence of septate microconidia, we tentatively identified the four isolates as being F. sacchari. The formation of sporodochia was variable according to each isolate, but when present, sporodochia were observed as cushion-like or slimy masses on the agar surface, at least in primary cultures on PDA or OA after 7 to 10 days. A molecular comparison with reference strains confirmed the identification of our isolates. Table 3 summarizes the most relevant morphological features useful for distinguishing these species, which can help those laboratories that have no molecular facilities for the identification of clinical isolates.
Key morphological features of seven Fusarium species of clinical interesta
Molecular study.With the primers we used we were able to amplify and sequence a fragment of 378 bp. Parsimony analysis of the data set yielded 225 phylogenetic trees of 157 steps in length. Three main clades with 100% bootstrap support were obtained (Fig. 1). The first of these comprised five different subclades, each representing the different species with the respective reference sequences. The first two included the isolates of F. nygamai (n = 7) and F. napiforme (n = 3), respectively. The F. sacchari (n = 5) subclade was genetically more distant; the fourth one grouped all isolates to F. proliferatum (n = 9), and the most distant subclade included only the reference sequence of F. subglutinans. The four species presented a very low intraspecific variability, showing only one haplotype each, with the exception of F. sacchari, which showed two haplotypes. The second main clade grouped the isolates of F. incarnatum (n = 9) and those of F. chlamydosporum (n = 8). These two species were clearly genetically differentiated and showed more genetic variability than those previously mentioned, with six and five haplotypes, respectively. Finally, the third main clade included the isolates morphologically identified as F. dimerum (n = 7), the most phylogenetically distant and heterogeneous species, which included three subclades with a total of six different haplotypes.
One of the 225 most parsimonious trees obtained from heuristic searches based on TUB sequences. Bootstrap support values are indicated at the nodes. CI, consistency index; HI, homoplasy index; RI, retention index; asterisks, accession numbers of sequences retrieved from GenBank.
Antifungal susceptibility study.The results of the susceptibility tests are shown in Table 2. TBF was the most active drug, showing a total geometric mean (GM) MIC of 0.60 μg/ml against all the species studied. One exception was F. incarnatum, which had a GM MIC of 10.08 μg/ml. AMB was the second most active drug, with a total GM MIC of 1.94 μg/ml. For this drug, F. proliferatum (GM MIC of 3.70 μg/ml), F. nygamai (GM MIC of 3.28 μg/ml), and F. napiforme (GM MIC of 3.18 μg/ml) were more resistant than the other species, whereas F. incarnatum showed the lowest GM MIC (0.93 μg/ml). Although VRC showed poor activity against all the strains tested, it was the third most active antifungal drug, with a total GM MIC of 4.24 μg/ml. PSC showed very variable results. Although it showed very high MICs in general, it exhibited significant activity for the majority of the strains of F. chlamydosporum, F. incarnatum, and F. sacchari. The rest of the azoles tested (RVC, ABC, ITC, and KTC) also showed high MICs for all the strains. FLC, 5-FC, and MFG were not active against any of the isolates (data not shown).
DISCUSSION
The use of the internal transcribed spacer rRNA gene sequences has proven to be useful for the identification of numerous fungal pathogens. However, this marker has been used only for the recognition of species complexes in the genus Fusarium (5). In contrast, β-tubulin has proven to be a good marker for species identification within the Gibberella fujikuroi species complex, although it is not very useful for other species because of paralogous or duplicated divergent alleles (5, 32). In general, we have obtained a high degree of correlation between the morphological and molecular identification of those clinical isolates that we have received for identification. The only species that was difficult to identify was F. sacchari. This is a species morphologically close to F. subglutinans (10, 25), and even Nelson et al. (28) considered them synonyms. Fusarium sacchari, F. subglutinans, and other related species constitute a morphologically similar group of species that can be differentiated practically only by the use of mating tests or molecular markers (25). Although several human infections have been attributed to F. subglutinans (10, 25), the identification of the case isolates is questionable. None of the clinical isolates included in this study was molecularly identified as F. subglutinans (Fig. 1). Our study confirmed that F. chlamydosporum and, especially, F. dimerum represent complexes of species (37), as has occurred in other more common species of Fusarium, such as F. solani, F. oxysporum, and F. verticillioides (3, 4, 32, 33, 35, 36). It would be interesting to do further studies of the phylogeny of F. chlamydosporum and F. dimerum, which have been associated with severe human infections (10, 25).
TBF showed good activity against all the species tested, which is in agreement with previous studies where this drug worked well in vitro against different Fusarium species (4, 22). Although there is no history of this drug in clinical cases caused by the species included in the present study, it was successfully used in two cases of onychomycosis by F. oxysporum (12, 18). Moreover, Gupta et al. (18) reported that onychomycosis infections caused by Fusarium species show response to this drug. Recently, the successful treatment of a disseminated cutaneous F. proliferatum infection using liposomal AMB and TBF was reported (29). This combination therapy was also effective in the treatment of a disseminated F. oxysporum infection (42).
AMB and VRC, the second and third most active drugs, respectively, are the drugs recommended for treating fusariosis (11). In fact, the use of AMB has shown good results in different clinical cases of systemic infection involving some of the species included in this study, such as F. chlamydosporum (23), F. sacchari (17), F. nygamai (24), F. dimerum (26), and F. proliferatum (6). Regardless of the antifungal used, for fusariosis neutrophil recovery is fundamental for resolving the infection, as demonstrated in practically all the cases. Numerous AMB failures also exist. For example, F. napiforme showed resistance to this drug in the only published case of disseminated infection caused by this species (27), as did F. proliferatum (44) and F. dimerum (2). VRC has demonstrated efficacy in the treatment of disseminated fusariosis, but most of the isolates involved in such infections were not identified to the species level (38). Recently, a stem cell transplant recipient with an invasive infection by F. dimerum (7) was also successfully treated with VRC. Despite the high MICs showed by PSC, this drug proved to be active in several invasive infections caused by F. proliferatum (21) or F. solani (19). The inactivity of FLC, 5-FC, and MFG against Fusarium isolates demonstrated in our study has already been reported by other authors (4, 14, 40).
In conclusion, we have demonstrated in this study that several Fusarium species of clinical interest can be identified through a detailed morphological exam. This fact can be very important for those clinical laboratories with no molecular facilities. However, there are still some species that are difficult to identify morphologically, such as F. sacchari or F. thapsinum (4), and in these cases molecular identification is mandatory. This is the first in vitro study that tests numerous strains of less-common clinical species of Fusarium and demonstrates the good in vitro activity of TBF. However, in vivo studies are needed to elucidate the potential clinical usefulness of this drug.
ACKNOWLEDGMENTS
We are indebted to K. O'Donnell (Agricultural Research Service, Peoria, IL), to the curators of the Centraalbureau voor Schimmelcultures (Utrecht, The Netherlands), and to A. Stchigel (Universitat Rovira i Virgili, Reus, Tarragona, Spain) for supplying strains. We thank Núria Pilas, Catalina Núñez, and Eduardo Álvarez for their contributions to this work.
This work was supported by the Spanish Ministerio de Ciencia y Tecnología grants CGL2005-07394/BOS and CGL 2007-65669/BOS.
FOOTNOTES
- Received 22 December 2008.
- Returned for modification 13 February 2009.
- Accepted 16 March 2009.
- Copyright © 2009 American Society for Microbiology
