Introduction

Sporothrix schenckii is a pathogenic dimorphic fungus that causes sporotrichosis, a cutaneous lymphatic or systemic mycosis particularly frequent in certain geographical areas such as Mexico, Brazil, Peru, and India [1–3]. However, infections due to S. schenckii have also been reported from other parts of the world, including Europe, where sporotrichosis is considered a rare disease [4–6]. Nevertheless, in recent years, several clinical autochthonous cases have been described in patients and animals that live in European countries [7–12], showing that this pathogenic fungus is more widespread than is now believed. In fact, a high prevalence of S. schenckii has been recently found in some kind of commercial soils which might represent an important vehicle for introducing the fungus in nature [13]. Therefore, it is not clear why, despite the occurrence of S. schenckii in environmental and commercial samples, sporotrichosis has a low incidence in Italy. Different assumptions were made in this regards, and one possible explanation could be the existence of cryptic species with reduced or absent pathogenicity [13]. This is in accordance with previous data obtained from phenotypic, virulence, and molecular studies which have suggested that more than one species could exist within S. schenckii [14–18]. In addition, more recently, Marimon et al. [18] have further supported this idea by examining three protein-coding loci, and on the bases of such studies, S. schenckii can now be recognized as a species complex comprising at least six sibling species: Sporothrix albicans, Sporothrix brasiliensis, Sporothrix globosa, Sporothrix luriei, Sporothrix mexicana, and S. schenckii [19, 20]. Furthermore, subsequent phylogenetic analysis based on sequence data of ribosomal DNA and β-tubulin regions from Sporothrix pallida, Sporothrix nivea, and S. albicans revealed a remarkably high genetic similarity, and therefore the last two species were synonymized with S. pallida, the first of the three species to be described [21]. However, apart from S. schenckii sensu stricto, only S. brasiliensis, S. globosa and S. luriei are associated with human infections, whereas S. mexicana and S. pallida have only been isolated from environmental samples even if S. pallida has been sometimes found as part of the intestinal flora of some insects [22]. It also is assumed that these two species of Sporothrix are not-infective to humans and other animals, and a recent experimental in vivo study has further confirmed that these fungi are non-pathogenic in a murine model of infection [23].

Molecular phylogeny, based on partial calmodulin sequence data, has significantly improved resolution among closely related Sporothrix species establishing six well-defined groups [19, 20]. Therefore, in this study, we amplify and sequence the calmodulin-encoding gene to evaluate the phylogenetic group to which the Italian Sporothrix isolates belong to. This is important for epidemiological investigations because it can improve our understanding of the distribution of these newly described Sporothrix species that have also showed significant differences in antifungal susceptibilities in vitro [24]. In fact, Marimon et al. [24] reported that the terbinafine was the most active antifungal agent against all Sporothrix species complex, whereas all others tested antimicotic drugs showed very poor activity against S. pallida, S. globosa, and S. mexicana.

Materials and Methods

Twenty-six environmental strains of S. schenckii, plus two clinical isolates, were examined in this study (Table 1). All strains were isolated in Italy and have already been well characterized, phenotypically and genotypically, in our previous study [13]. Nevertheless, however, their identity was further confirmed before use by standard phenotypic and molecular methods. S. schenckii ATCC 10268, S. schenckii KMU 975, S. luriei KMU 2787, S. mexicana FMR 9108, S. brasiliensis SS74, S. globosa SS49, and S. albicans CBS 302.73 were also included and used as reference strains.

Table 1 Strains, species, origin, and GenBank accession numbers of fungal isolates used in this study
Full size table

For demonstration of thermal dimorphism, one-week-old fungal cultures of the mycelial form on Sabouraud dextrose agar, grown at 25°C, were inoculated in brain heart infusion broth and incubated at 37°C. Fungal morphology of both yeast-like and mycelial phase was microscopically examined.

For molecular analysis, total genomic DNA was isolated by using the same extraction protocol as previously described in detail in Criseo and Romeo [13].

The protocol of Kano et al. [25], based on amplification of the partial chitin synthase 1 gene (Chs1-PCR), was used to confirm the identity of S. schenckii.

Amplification of the partial calmodulin-encoding gene was performed with primers CL1-GARTWCAAGGAGGCCTTCTC and CL2A-TTTTTGCATCATGAGTTGGAC as described by O’Donnell et al. [26]. Following PCR, amplicons were purified with QIAquick PCR purification kit (Qiagen, Milan-Italy) and sequenced at MWG-biotech (M-Medical S.r.l, Milan, Italy) with the same primers used for PCR.

The identity of our nucleotide sequences was verified by BLASTN search (http:\\www.ncbi.nlm.nih.gov/blast). Several published Sporothrix-calmodulin-related sequences (Table 1) were retrieved from GenBank, and computer-assisted multiple sequence comparisons were performed using ClustalW algorithm implemented in MEGA4 software [27].

The multiple nucleotide sequence alignment was inspected, visually adjusted and subsequently was used for neighbor-joining analysis performed using MEGA4 software; confidence was estimated using 1,000 rounds of bootstrapping.

DNA and protein sequence analyses, to detect nucleotide and aminoacidic changes in the calmodulin exons, were performed with a combination of Transeq, ClustalW and GeneWise computing packages (http:\\www.ebi.ac.uk).

Three representative nucleotide sequences: HQ686039 (strain SPA8; clade Va), HQ686040 (strain HOL3; clade Vb), and HQ692915 (strain BG6; clade Vb) have also been submitted to GenBank database.

Results

A total of 28 previously identified S. schenckii Italian strains (26 environmental and 2 clinical) [13] were examined in order to establish their genetic relatedness on the basis of the calmodulin sequences, the most phylogenetically informative locus found in recent studies [19, 20].

All S. schenckii isolates showed temperature dimorphism with the typical cigar-shaped form of the yeast-like cells grown at 37°C and distinctive micromorphology at 24°C. The identity of the fungus was further confirmed by Chs1-PCR that produced a DNA fragment of ~300 bp long from all tested strains including reference strains. These results point out that this molecular method does not discriminate among newly recognized Sporothrix species since it produces an identical DNA fragment independently from the species status of the examined strain (Fig. 1).

Fig. 1
figure 1

Agarose gel electrophoresis of DNA fragments obtained by amplification of the CHS1 gene. M molecular size marker; Lanes 1–3 S. schenckii ATCC 10268, CDM18 and SPO1, respectively; Lane 4 S. luriei KMU 2787; Lane 5 S. mexicana FMR 9108; Lane 6 S. brasiliensis SS74; Lane 7 S. globosa SS49; Lane 8 S. pallida CBS 302.73; Lanes 9–13 Italian environmental isolates: GER1, SPA1, SAM1, BG, and HOL3, respectively

Full size image

The amplified calmodulin genes yielded DNA fragments of approximatively 800 bp in size.

BLAST searches, using calmodulin sequences as query, revealed that all 26 environmental isolates had a high level of sequence similarity with S. albicans (99–100%), whereas the two clinical isolates showed high-score matches (99%) with S. schenckii.

The results of our phylogenetic analysis agree with those obtained in other previous studies where six well-defined and supported groups were found [19, 20] (Fig. 2).

Fig. 2
figure 2

Phylogenetic tree generated by neighbor-joining analysis using partial nucleotide sequences of the calmodulin-encoding gene. Bootstrap support values above 85% are indicated at the nodes

Full size image

The two Italian clinical strains were co-clustered with S. schenckii sensu stricto isolates; in particular, they fall within sub-clade IIa that includes S. schenckii isolates from different geographical origin, including both European strains: CBS 938.72 and CBS 359.36T [19] (Fig. 2).

Conversely, all 26 environmental isolates, significantly clustered (bootstrap confidence value of 100%) with clade V isolates (CBS 111110 and CBS 302.73) (Fig. 2), were previously shown to belong to S. albicans [19] whose name has been recently replaced by S. pallida [21]. Furthermore, our results showed that clade V could be further subdivided into two sub-clades, designated Va and Vb, consisting of fungal isolates from different European countries (Fig. 2). More precisely, 8 of 10 S. pallida strains of sub-clade Va were previously isolated from commercial amended soil purchased in Italy, but produced and packaged in Spain (4 isolates), Germany (4 isolates) and imported; the remaining two isolates were found in natural soil in Reggio Calabria, Italy [13]. The 16 isolates of sub-clade Vb were, indeed, recovered from environmental and commercial samples from Italy (5 isolates) and commercial soils imported from Austria (5 isolates), Spain (3 isolates), and Holland (3 isolates) [13]. This latter group (Vb) was found to be more heterogeneous than the sub-clade Va, suggesting that these two sub-clades, which contain three different types of strains, could likely correspond to the previously synonymized species: S. pallida, S. nivea, and S. albicans. Unfortunately, it was not possible for us to test this hypothesis because of the lack of published calmodulin-related sequences from these species. In fact specific searches in GenBank database revealed only the presence of ribosomal and β-tubulin sequences, which have been previously used as genetic markers to evidence the synonymy of these fungi [21].

Bioinformatic analysis revealed that genetic differences between members of the two sub-clades were due to mutations located mostly in intronic regions; only isolates of sub-clade Vb, except the two reference isolates and HOL3 strain (Fig. 2) showed a silent mutation (Arg74; CGC → CGT) in the third exon of the calmodulin.

Discussion

In past years, several molecular studies have clearly demonstrated that S. schenckii isolates displayed genetic characteristics so different that they appear not to be the same species [16, 28–31]. Hereafter, the taxonomy of S. schenckii has undergone significant changes due to the description of new closely related cryptic species [19, 20]. In these cases, as it has happened for other clinically relevant pathogenic fungi [32–34], retrospective and prospective epidemiological studies are necessary to better know the incidence and clinical significance of these new species. However, it remains difficult, nowadays, to identify correctly the members of the S. schenckii complex in clinical laboratories due to lack of rapid and specie-specific molecular methods. In fact, the results obtained here, and in our previous study [13], indicate that the amplification of the chitin synthase gene [25] as well as that of the DNA topoisomerase II gene [35] produces identical DNA patterns in S. schenckii sensu stricto and its related species; hence, these methods are not reliable tests for discrimination of members of the S. schenckii group. In effect, our data suggest that all molecular methods, for specific identification of S. schenckii, developed before the recent description of the new Sporothrix species should be used with caution because the proper identification of the species causing the infection is needed in order to choose the most effective antifungal treatment. These observations are strongly supported by the recent study of Marimon et al. [24] who examined 92 isolates belonging to the S. schenckii complex and found that there were significant differences in antifungal susceptibilities between species tested. Therefore, only sequencing of highly informative genetic loci such as the calmodulin-encoding gene is, at present, useful for elucidating relationships and differentiates among species of the S. schenckii complex.

Although in general, ribosomal DNA genes evolve cohesively within a single species and exhibit normal levels of sequence divergence between rDNA copies from different species, in the case of S. schenckii group, the use of this genetic marker could lead to wrong conclusions as already happened in other studies [13, 16, 21].

The inefficacy of ITS regions for resolving the phylogeny of S. schenckii has already been demonstrated by de Meyer et al. [21]. In their study, these authors reclassified some previous environmental S. schenckii strains [16] by using sequencing of the β-tubulin gene because genetic data based on ITS sequences alone were unable to resolve phylogenetic relationship among the strains [16, 21]. Likewise, in our previous phylogenetic study based on sequencing of the D1-D2 region of 28S rDNA, we have found two strongly supported groups: one group composed entirely of isolates of clinical origin and a second one containing only environmental isolates [13]. The observed genetic differences, between these two groups, led us to hypothesize the presence of cryptic related species that has been demonstrated in this study. In fact, all 26 members of the environmental group [13] have, here, been identified as S. pallida. The exclusive presence of this non-pathogenic species in our environmental samples, as well as the absence of S. schenckii sensu stricto strains, may explain in part why the sporotrichosis has a low incidence in Italy and in Europe in general.

Interestingly, in this study, the S. pallida (clade V) was found to be quite heterogeneous and was further subdivided into two subgroups (Va and Vb). Bioinformatic analysis revealed that the genetic differences between members of these groups were due to nucleotide substitutions mainly located within intronic regions (near exon/intron junctions). However, all isolates of the subgroup Vb, except CBS 111110, CBS 302.73T and HOL3 strains, also showed one characteristic nucleotide transition (C → T) in the third exon of the calmodulin, but this mutation was silent and not accompanied by any amino acid alteration.

Finally, only two isolates were recognized as S. schenckii sensu stricto, and this was consistent with their origin and their pathogenic nature; in fact, these strains represent the only examples of authentic clinical S. schenckii isolates from Italy.

In conclusion, the data reported in this study not only suggest further environmental and clinical investigations to reveal the occurrence of these newly described Sporothrix species in Europe but also bring to our attention the need to develop rapid and specific molecular methods to differentiate these new Sporothrix species in the clinical laboratories.