Molecular Identification and Susceptibility of Clinically Relevant Scedosporium spp. in China

As various new sibling species within the Scedosporium spp. have been described recently, this study was conducted to investigate distribution and antifungal susceptibility profiles of the different species of Scedosporium spp. in China. Twenty-one clinical strains of Scedosporium from China and two strains from Japan were reidentified by MLSA. The analysis included BT2, CAL, RPB, SOD, and ACT and the combination of the five loci. Pseudallescheria boydii complex (17 strains) and S. apiospermum (6 strains) were identified. P. boydii complex included four closely related subgroups: P. boydii (9 strains), P. ellipsoidea (6 strains), P. fusoidea (1 strain), and P. angusta (1 strain). There were no significant differences in MICs for neither VOR, POS, nor AMB over all the five species in study. For itraconazole, intraspecific diversity was evident.


Introduction
Scedosporium spp. is one of the species of the opportunistic pathogenic fungi that can always be found in environment, especially in sewerage. It infects immunocompromised patients and drowning men, involving lungs, sinuses, bones, joints, eyes, and brain [1].
In recent years, members of the genus Scedosporium are increasingly recognized as opportunistic agents of disease, for example, in transplant recipients. Scedosporium species have been identified as the second most prevalent mold after Aspergillus colonizing the lungs of patients with cystic fibrosis [2]. Scedosporium infections occur worldwide. In European countries, USA and Australia, Scedosporium species were always found in patients with chronic lung diseases, cystic fibrosis (CF), lung or allogenic bone-marrow transplantation, and hematologic malignancies [1,3,4]. Accordingly, the common types were pulmonary, nasal sinuses, skin/soft tissues, CNS, and disseminated infection [1,3,4].

Strains.
From 1990 to 2014, twenty-one Scedosporium strains isolated from patients were reserved in Research Center for Medical Mycology at Peking University. The clinical samples were collected from 14 Chinese hospitals which located mainly in central and south of China. All the isolates were identified as S. apiospermum or P. boydii by morphology. A total of 23 isolates (including two strains from Japan) as shown in Table 1 were investigated in this study. Furthermore, in vitro susceptibility was performed on the same set strains. * Strains from Japan.
The PCR assay (25 L) included 2 L of fungal DNA extract, 1 M of each gene-specific primer, 2.5 mmol dNTP Mix 1 L, 10x PCR buffer 2.5 L, and LA Taq polymerase 0.25 L (Fermentas, St. Leon-Rot, Germany). The amplification for all targeted genes was performed in a Eppendorf PCR machine (AG22331) as follows: 5 min of initial denaturation at 95 ∘ C, followed by 35 cycles at 95 ∘ C for 30 s, gene-specific annealing temperature for 30 s, and 72 ∘ C for 1 min (for RPB2 the annealing time was 2.5 min). The PCR products were visualized by electrophoresis on a 1% (w/v) agarose gel. Both strands of the PCR fragments were sequenced using the above-mentioned primers. The consensus sequences were obtained using SeqMan (DNAStar-Lasergene, Madison, WI, USA) software. Newly obtained sequences were deposited in GenBank under accession numbers KP 981107 to KP 981221 (Table 1). They were used to conduct alignment analysis for preliminary species identification in the NCBI genomic database (http://blast.ncbi.nlm.nih.gov/) and CBS database (http://www.cbs.knaw.nl/). The sequences were aligned using MUSCLE. For the maximum likelihood analysis, the distances between sequences were calculated using the best parameter model found by MEGA 6.0 6 (http://www.megasoftware.net/). A bootstrap analysis was conducted with 1000 replications.

Susceptibility Test.
The in vitro susceptibility of the 23 Scedosporium isolates against four antifungal agents was evaluated by using the Clinical and Laboratory Standards Institute (CLSI) M38-A2 broth microdilution method [15]. The inocula suspensions were prepared in new sterile tubes and adjusted to 0.4−5 × 10 6 colony-forming units per milliliter (CFU/mL) by counting spores in a hemocytometer and subsequently verifying them through quantitative colony counts on PDA plates. The nongerminated spore suspensions were diluted 1 : 100 in an RPMI 1640 to achieve a final inoculum concentration of 0.4-5 × 10 4 CFU/mL. The following antifungal agents were used: voriconazole (VOR; Shouguang Fukang Pharmaceutical Co., Ltd., China), posaconazole (POS; Merck, Rahway, NJ, USA), itraconazole (ITR; Shouguang Pharm), and amphotericin B (AMB; Sigma-Aldrich Co., St. Louis, USA). They were all diluted in 100% dimethyl sulphoxide as a stock solution with a concentration of 1.600 mg/L. Final drug concentrations ranged from 16   concentrations (MIC) endpoints were defined as the lowest concentration at which there was a complete inhibition of growth. Aspergillus flavus ATCC 204304 served as a quality control strain. The microtiter panels were incubated at 35 ∘ C and the results were read after 72 h. All tests were performed in triplicate on three different days.

Molecular Phylogeny.
We were able to amplify and sequence 470 bp, 645 bp, 935 bp, 775 bp, and 396 bp of the BT2, CAL, RPB, ACT, and SOD loci, respectively. Of the 3221 nucleotides sequenced, 209 (6.5%) were informative for parsimony in the different Scedosporium isolates. For identification, reference sequence of Scedosporium species available in public database were used including BT2 and CAL, but no sequences for RPB, SOD, and ACT were available. The sequences for BT2, CAL, RPB, SOD, and ACT yielded phylogenetic trees with the same topology ( Figures  1-5). In the combination phylogenetic trees based on BT2 and CAL (Figure 6), 23 strains in our study were reidentified to the species level according to the reference strains, which were analyzed in the Gilgado literature. P. boydii (9/23) and its closely related subtypes P. ellipsoidea (6/23), P. fusoidea (1/23), and P. angusta (1/23) were the most common, and the other 6 of 23 strains were identified as S. apiospermum.
The topology of the combined tree (Figure 7) of all five loci was similar to those observed in the trees of individual locus and the combined tree of BT2 and CAL. Four principal clades were obtained. The four clades were the P. boydii clade, P. ellipsoidea clade, P. fusoidea/P. angusta clade, and S. apiospermum clade. P. fusoidea and P. angusta always assemble together, and P. boydii and P. ellipsoidea were very closely related. S. apiospermum has the highest intraspecies variability (genetic distance = 0.008), which is comparable to the interspecies variability between P. fusoidea and P. angusta (genetic distance = 0.008).

In Vitro Susceptibility Test.
The MIC values for the four antifungal agents examined by the CLSI M-38A2 microdilution method against the 23 strains are presented in Table 2. VOR was the most active agent against all 23 strains with a MIC range from 0.25 to 1 g/mL and a 0.46 g/mL GM. POS was the second most active agent with MIC values of 2 or 4 g/mL. For ITR, the intraspecific diversity was obvious as most strains had a MIC between 2 and 4 g/mL, but five strains were higher at a MIC of 32 g/mL. AMB is the least effective with high MIC values ranging from 4 to 32 g/mL in vitro.
The        presented in Table 3. There were similar MIC 50 and MIC 90 for VOR, POS, or AMB of the three species above. For ITR, the MIC 90 of P. boydii (4 g/mL) was lower than that of P. ellipsoidea (32 g/mL) and S. apiospermum (32 g/mL).
Because cystic fibrosis is extremely rare in the Chinese population, no scedosporiosis in CF was reported in China. Based on molecular identification, we found that P. boydii complex and S. apiospermum were the only two species in our study. Actually, P. boydii and S. apiospermum demonstrated a closely phylogenetic relationship. This is also reflected by their undifferentiated morphological characteristics, assimilation of different sugars, and growth temperature (data not shown). In P. boydii complex, four closely related subgroups, P. boydii (9 strains), P. ellipsoidea (6 strains), P. fusoidea (1 strain), and P. angusta (1 strain), were present in the study.
In a set of clinical and environmental strains from Austria, Germany, and Netherlands, S. apiospermum was the most prevalent, followed by P. boydii [32,33]. In Australia, S. apiospermum and S. aurantiacum were the Scedosporium species with a high incidence except for L. prolificans [4]. Based on our study, we found that in China the P. boydii complex represents the most prevalent species (16/21) followed by S. apiospermum (5/21), which is the same as in Northern     Spain and France [34,35]. In the Chinese strains, we could not find S. aurantiacum, which had a high incidence in Australia and a relatively low incidence in Europe.
Antifungal susceptibility profiles for clinical breakpoint of Scedosporium species are still under study. Of the four antifungal tested in this study, VOR was found to be the most active agent against Scedosporium species; POS was the second, and AMB had limited antifungal activity. For ITR, the intraspecific diversity was obvious as the range of MIC from 2 to 32 g/mL and P. boydii had a lower MIC 90 than that of P. ellipsoidea and S. apiospermum. This in vitro susceptibility data is consistent with previous reports [32,34,[36][37][38]. In China, most cases caused by Scedosporium were treated with either ITR or VOR. Although Scedosporium strains have relatively high MIC values for ITR (GM = 5.25) in vitro, studies have verified that they are clinically effective [7].

Conclusion
The P. boydii complex and S. apiospermum were the main Scedosporium species in clinical samples in China. Scedosporium species can be distinguished unambiguously by all the five loci in this study. VOR is the most active agent in vitro against the set of Scedosporium species in this study, followed by POS and ITR.