Identification of Pseudallescheria and Scedosporium Species by Three Molecular Methods

Abstract

The major clinically relevant species in Scedosporium (teleomorph Pseudallescheria) are Pseudallescheria boydii, Scedosporium aurantiacum, Scedosporium apiospermum, and Scedosporium prolificans, while Pseudallescheria minutispora, Petriellopsis desertorum, and Scedosporium dehoogii are exceptional agents of disease. Three molecular methods targeting the partial β-tubulin gene were developed and evaluated to identify six closely related species of the S. apiospermum complex using quantitative real-time PCR (qPCR), PCR-based reverse line blot (PCR-RLB), and loop-mediated isothermal amplification (LAMP). qPCR was not specific enough for the identification of all species but had the highest sensitivity. The PCR-RLB assay was efficient for the identification of five species. LAMP distinguished all six species unambiguously. The analytical sensitivities of qPCR, PCR-RLB, and LAMP combined with MagNAPure, CTAB (cetyltrimethylammonium bromide), and FTA filter (Whatman) extraction were 50, 5 × 10³, and 5 × 10² cells/μl, respectively. When LAMP was combined with a simplified DNA extraction method using an FTA filter, identification to the species level was achieved within 2 h, including DNA extraction. The FTA-LAMP assay is therefore recommended as a cost-effective, simple, and rapid method for the identification of Scedosporium species.

INTRODUCTION

Pseudallescheria species and their anamorphs in the genus Scedosporium classically are known as traumatically inoculated molds causing subcutaneous infections; in both of these genera, systemic diseases, eventually with neurotropic involvement, have become significant (16). The fungi are regularly encountered in the lungs of patients with cystic fibrosis (CF) or chronic suppurative lung disease (5, 7, 8, 17, 37), being the second most frequent fungal colonizer. However, upon isolation with routine media these species are outcompeted by Aspergillus and Candida spp. (7, 31), and their frequency is severely underestimated. In addition to selective isolation (31), the development of non-culture-based detection methods would provide better insight into their clinical relevance. For example, using counterimmunoelectrophoresis, scedosporiosis was found in 21.1% of patients serologically (7), while with an oligonucleotide array a 30.7% frequency of Scedosporium apiospermum was recently noted (5). Diabetes is a common complication of cystic fibrosis, and lung transplantation remains the ultimate treatment for the disease; in both situations there is a risk of dissemination of fungal colonizers in the lungs, such as Scedosporium (35). Overall, mortality for systemic scedosporiosis is 46.9%, but with disseminated disease and concomitant cerebral infection (33) the mortality may be as high as 87.5% (32).

Several recent studies demonstrated a significant phylogenetic distance within the Microascaceae between two main groups with anamorphs in Scedosporium. Scedosporium aurantiacum, S. apiospermum, and S. dehoogii were members of a clade (I) with Pseudallescheria teleomorphs (P. boydii and P. minutispora), whereas S. prolificans and relatives were found in a clade (II) with Petriella teleomorphs (M. Lackner, A. H. G. Gerrits van den Ende, G. S. de Hoog, and J. Kaltseis, submitted for publication). The great majority of subcutaneous and systemic strains are concentrated in four species located in these main clades, including S. aurantiacum, S. apiospermum, P. boydii, and S. prolificans (12, 14). Lackner (26) found that S. apiospermum and P. boydii had very similar antifungal susceptibility profiles, while profiles of S. aurantiacum, S. prolificans, and S. dehoogii were deviant from each other. Given the differences in virulence, metabolic abilities, and antifungal susceptibility between most species (3, 13, 15, 20), it is significant to detect and identify individual taxa within the clades. Morphological recognition of Scedosporium species is impossible due to a lack of phenetic characters and a high degree of morphological plasticity. New methods for routine identification are therefore needed. Several molecular methods targeting the internal transcribed spacer (ITS) region have been described, but with insufficient resolution to differentiate all clinical species of the two clades. The methods applied thus far include the use of sequencing (9, 12), random amplified polymorphic DNA (8), restriction fragment length polymorphism (9; Lackner et al., submitted), M13 PCR fingerprinting (9), amplified fragment length polymorphism (9), quantitative PCR (6), microarrays (5), specific PCR (Lackner et al., submitted), and rolling-circle amplification (27, 38). Lackner et al. (submitted) designed a set of ITS primers specific for most of the species, but for the identification for the most common taxa, S. apiospermum and P. boydii, partial β-tubulin gene (BT2) primers were needed. Gilgado et al. (12) noted that β-tubulin provided more information than ITS as a target for the identification of Scedosporium species.

In the present study, we developed three methods to identify closely related species of the Scedosporium species complex down to the species level using quantitative real-time PCR (qPCR), PCR-based reverse line blotting (PCR-RLB), and loop-mediated isothermal amplification (LAMP) combined with different DNA extraction methods, including a MagNAPure LC Isolation station, cetyltrimethylammonium bromide (CTAB), and Whatman FTA filter papers (FTA).

MATERIALS AND METHODS

Strains.

Sixty strains covering the entire diversity of the species of Scedosporium and nine strains from other related Pseudallescheria and Scedosporium species, as well as 10 non-Pseudallescheria and non-Scedosporium fungal isolates were obtained from the collection of the Centraalbureau voor Schimmelcultures Fungal Biodiversity Centre (CBS) (Table 1). The Pseudallescheria and Scedosporium species tested included S. aurantiacum, S. apiospermum, P. boydii, P. angusta, P. ellipsoidea, P. fusoidea, S. prolificans, S. dehoogii, and P. minutispora; Parascedosporium tectonae, Lophotrichus fimeti, Petriellopsis africana, Petriella setifera, Petriella sordida, Petriella guttulata, Petriellopsis desertorum, Pithoascus langeronii, Graphium penicillioides, Aspergillus fumigatus, Aspergillus flavus, Aspergillus nidulans, Aspergillus niger, Aspergillus terreus, Candida albicans, Fusarium solani, Rhizopus oryzae, Cryptococcus neoformans, and Exophiala dermatitidis were also tested. Since P. angusta, P. ellipsoidea, and P. fusoidea could not be clearly distinguished from P. boydii, we reduced them to the P. boydii complex. All strains belonging to Pseudallescheria and Scedosporium species were sequenced for both the ITS regions and the partial BT2 gene which comprised exons 2 to 4, generated and sequenced with primers Bt2a and Bt2b.

Table 1.

Summary of the strains used in this study

Species	Straina	Country of isolation	Source
Scedosporium aurantiacum	CBS 116910 (T)	Spain	Clinical
	CBS 101726	Netherlands	Mud
	CBS 117423	France	Sputum, cystic fibrosis
	DH 16589	Austria	Soil
	DH 16457	Austria	Soil
	DH 16443	Austria	Soil
	DH 16452	Austria	Soil
	DH 16532	Austria	Soil
	DH 16634	Austria	Soil
	DH 16495	Austria	Soil
Scedosporium apiospermum	CBS 116899 (T)	France	Sputum, cystic fibrosis
	CBS 101725	Netherlands	Mud
	CBS 695.70	Ukraine	Pig, nasal cavity
	CBS 948.87	India	Human (male), mycetoma pedis
	CBS 116403	Germany	Human (male), fatal cerebral infection, brain
	CBS 116779	China	Human (male), sinus
	DH 16618	Austria	Soil
	DH 16481	Austria	Soil
	DH 16473	Austria	Soil
	DH 16544	Austria	Soil
Pseudallescheria boydii	CBS 101.22 (T)	United States	Human (male), mycetoma
	CBS 116894	Thailand	Soil
	CBS 120157	France	Human (male), lung, leukemia
	CBS 116892	France	Sputum, cystic fibrosis
	DH 16566	Austria	Soil
	DH 16549	Austria	Soil
	DH 16539	Austria	Soil
	CBS 117390	Zaire	Forest soil
Pseudallescheria angusta	CBS 254.72 (T)	United States	Sewage half digestion tank
	CBS 108.54	Zaire	Soil
Pseudallescheria ellipsoidea	CBS 418.73 (T)	Tajikistan	Soil
Pseudallescheria fusoidea	CBS 106.53 (T)	Panama	Soil
Scedosporium prolificans	CBS 467.74 (T)	Belgium	Greenhouse soil
	CBS 100391	Germany	Human (male), disseminated infection in patient with AIDS and Burkitt lymphoma
	CBS 114.90	United States	Human (male), bone biopsy
	CBS 116900	Spain	Human (male), blood
	CBS 116901	Spain	Soil
	CBS 116902	Spain	Human (male), blood
	CBS 116903	Spain	Air
	CBS 116904	Spain	Human (male), blood
	CBS 116906	France	Sputum, cystic fibrosis
	CBS 116908	Spain	Human (male), blood
Scedosporium dehoogii	CBS 117406 (T)	Spain	Soil
	CBS 101723	Netherlands	Mud
	CBS 101720	Netherlands	Sandy soil of polluted ditch
	CBS 101721	Netherlands	Mud
	CBS 499.90	Netherlands	Mud of tropical pond
	CBS 100.26	Unknown	Clinical
	DH 16629	Austria	Soil
	DH 16645	Austria	Soil
	DH 16644	Austria	Soil
	DH 16451	Austria	Soil
Pseudallescheria minutispora	CBS 116911 (T)	Spain	River sediment
	CBS 116595	Belgium	Fuel oil
	DH 16625	Austria	Soil
	DH 16628	Austria	Soil
	DH 16629	Austria	Soil
	DH 16569	Austria	Soil
	DH 16631	Austria	Soil
	DH 16357	Austria	Soil
Parascedosporium tectonae	CBS 108.10 (T)	France	Human (male), foot
Lophotrichus fimeti	CBS 129.78 (T)	India	Dung of goat
Petriellopsis africana	CBS 311.72 (T)	Namibia	Brown sandy soil
Petriella setifera	CBS 265.64	Japan	Soil
Petriella sordida	CBS 385.87	Finland	Human (male), nail
Petriella guttulata	CBS 362.61 (T)	Germany	Dung of partridge
Petriellopsis desertorum	CBS 489.72 (T)	Kuwait	Salt-marsh soil
Pithoascus langeronii	CBS 203.78 (T)	India	Dung of herbivore
Graphium penicillioides	CBS 320.72	Solomon Islands	Forest soil
Aspergillus fumigatus	CBS 133.61	United States	Chicken, lung
Aspergillus flavus	CBS 100927	Pacific Islands	Cellophane
Aspergillus nidulans	CBS 565.70	United Kingdom	Cotton yarn
Aspergillus niger	CBS 554.65	United States	Fermentation
Aspergillus terreus	CBS 469.81	Thailand	Human (male), cardiac valve
Candida albicans	CBS 8758	unknown	Disseminated candidiasis
Fusarium solani	CBS 108942	Netherlands	Human (male), big toe
Rhizopus oryzae	CBS 112.07 (T)	Netherlands	Unknown
Cryptococcus neoformans	CBS 10513	United States	Unknown
Exophiala dermatitidis	CBS 207.35	Japan	Human (male), facial chromomycosis

^aAbbreviations: (T), ex-type strain; CBS, Centraalbureau voor Schimmelcultures, Utrecht, Netherlands.

Primer design.

Primers and probes (Table 2) were designed on the basis of ∼300 BT2 sequences in an internal research database of Pseudallescheria and Scedosporium available at CBS which included all known ex-type strains. For qPCR, four species-specific primers and 5′-FAM-labeled TaqMan probes were designed. For RLB, a group-specific primer PS_F specific for S. aurantiacum, S. apiospermum, P. boydii, S. dehoogii, P. minutispora, and Petriellopsis desertorum and Pro_F specific for S. prolificans were designed as forward primers with 5′-biotin-labeled T2_Bas reverse primer. Six species-specific probes and a group-specific probe PS_P specific for S. aurantiacum, S. apiospermum, P. boydii, S. dehoogii, P. minutispora, and Petriellopsis desertorum were labeled with C6-amino linker at the 5′ end. The melting temperatures of the probes were optimized at ca. 60°C by Sigma-Aldrich (http://www.sigma-genosys.com/calc/DNACalc.asp). Briefly, the design of the two outer primers, F3 and B3, is the same as that of regular PCR primers, while the design of the forward inner primer (FIP; F1c + F2) and backward inner primer (BIP; B1c + B2) is different from that of PCR (Fig. 1). Primers for six species of interest were designed by using PrimerExplorer v.4 (http://primerexplorer.jp/e/). All primers and probes were BLAST searched against nucleotide databases available at GenBank, as well as the CBS database, to verify specificity.

Table 2.

Primers and probes designed in this study^a

Primer or probe	Oligonucleotides (5′–3′)	Position	Specificity
Primers and probes used for qPCR
Forward primer	TGGCGAGCACGGTCTTG	136–152a
Aura_P	FAM-TAGCAACGGAATGTATGGACTCCCCCT-BBQ	154–180a	S. aurantiacum
Aura_R	CGTCGACTGCTGCTGCT	215–199a	S. aurantiacum
Apioboy_P	FAM-TAGCAACGGAGTGTACGGAACCACCC-BBQ	120–145b	S. apiospermum and P. boydii
Apioboy_R	ACATTCACGGCAGACACTGATT	216–195b	S. apiospermum and P. boydii
Pro_P	FAM-CAGCAATGGCGTGTATGGCTTCCC-BBQ	121–144d	S. prolificans
Pro_R	GCCTTTACCTCTGGTGATTTGG	176–155d	S. prolificans
Minu_P	FAM-TAGCAACGGAGTGTACGGCCCCC-BBQ	72–94f	P. minutispora
Minu_R	TAATGCCAGTCACATCGACTAGTG	134–111f	P. minutispora
Primers and probes used for PCR-RLB
PS_F	CAAAGTTACAATGGCACTTCTG	269–290a
Pro_F	GTTACAATGGAACATCCGAACTCC	239–262d
T2_B	biotin-TAGTGACCCTTGGCCCAGTTG	586–566a
Aura_P	amino-CAGCCTATCTATGTTGCGG	375–393a	S. aurantiacum
Apio_P	amino-GAGGTAAGTTTTTGGCTAAAGC	286–307b	S. apiospermum
Boy_P	amino-CGAGGTAAGTTTTTGGTTCAAA	272–293c	P. boydii
Pro_P	amino-ATAGCGAGCCATAAATCCG	313–331d	S. prolificans
Deho_P	amino-CAGAAATACTAACGTCTTGGTTACCT	339–364e	S. dehoogii
Minu_P	amino-GGTATGTTTTTGGTTAAAGCCAT	237–259f	P. minutispora
PS_P	amino-TAGGCTTCGGGCAACA	426–441a	S. aurantiacum, S. apiospermum, P. boydii, S. dehoogii, P. minutispora, and P. desertorum
Primers used for LAMP
Aura_F3	CCATTTCTGGCGAGCACG	129–146a	S. aurantiacum
Aura_B3	ACCTCGTTGAAGTAGACGCT	330–311a	S. aurantiacum
Aura_F2	TGTATGGACTCCCCCTTTCC	165–184a	S. aurantiacum
Aura_F1c	GAGAGCAGCAGCAGCAGCAG	233–214a	S. aurantiacum
Aura_B2	AGCTGGAGTTCAGAAGTGC	300–282a	S. aurantiacum
Aura_B1c	AACCGCAGAGACTGATTGTCCC	239–260a	S. aurantiacum
Apio_F3	ATGGCACTTCTGAACTCCAG	221–240b	S. apiospermum
Apio_B3	GGGCTCGAGATCTACAAGGA	419–400b	S. apiospermum
Apio_F2	CTTGAGCGCATGAGCGTC	241–258b	S. apiospermum
Apio_F1c	CAACCGGCCCGTGGCTTTAG	302–283b	S. apiospermum
Apio_B2	TTTGTTGCCCGAAGCCTAT	383–365b	S. apiospermum
Apio_B1c	GTGTCATCCGGCCTCCGTTG	308–327b	S. apiospermum
Boy_F3	ATGGCACTTCTGAACTCCAG	226–245c	P. boydii
Boy_B3	CGAGATCTACAAGGACAGCG	419–400c	P. boydii
Boy_F2	CTTGAGCGCATGAGCGTT	246–263c	P. boydii
Boy_F1c	AACCAGCCCGTGGTTTGAACC	306–286c	P. boydii
Boy_B2	TTTGTTGCCCGAAGCCTAT	388–370c	P. boydii
Boy_B1c	GTGTGGTGTCATCCAGCCTCC	308–328c	P. boydii
Pro_F3	ACAGGCAAACCATTTCCGG	89–107d	S. prolificans
Pro_B3	GCTCAAGCTGGAGTTCGG	273–256d	S. prolificans
Pro_F2	CGAGCACGGTCTTGACAG	108–125d	S. prolificans
Pro_F1c	AGAGGTAAAGGCGGGGGTGG	168–149d	S. prolificans
Pro_B2	ACACGCAGAGGAAAATCAGT	236–217d	S. prolificans
Pro_B1c	GGTGATTTGGCGTATTGGGCTG	169–190d	S. prolificans
Deho_F3	AGGCAAACCATTTCTGGCG	78–96e	S. dehoogii
Deho_B3	ACCTCGTTGAAGTAGACGCT	280–261e	S. dehoogii
Deho_F2	AGCACGGTCTTGATAGCA	97–114e	S. dehoogii
Deho_F1c	TGGCATTAGGGGGGTTTAAGGG	159–138e	S. dehoogii
Deho_B2	CAAGCTGGAGCTCAGAAG	252–235e	S. dehoogii
Deho_B1c	TGCTGCTCTCGCATTAACGACA	174–195e	S. dehoogii
Minu_F3	GCACGGTCTTGATAGCAACG	60–79f	P. minutispora
Minu_B3	GGTCGGATGACACCACATC	287–269f	P. minutispora
Minu_F2	CCTCATCCCTTACCCCCTA	94–113f	P. minutispora
Minu_F1c	AGGGACAATCAGTGTCTGCCGT	170–149f	P. minutispora
Minu_B2	CAGCCCATGGCTTTAACCA	265–247f	P. minutispora
Minu_B1c	TGGCACTTCTGAACTCCAGCTT	189–210f	P. minutispora

^aAbbreviations: Aura, S. aurantiacum; Apio, S. apiospermum; Boy, P. boydii; Pro, S. prolificans; Deho, S. dehoogii; Minu, P. minutispora; PS, Pseudallescheria/Scedosporium; F, forward primer; B, backward primer; P, probe; c, complementary; BBQ, BlackBerry Quencher. Superscript letters denote the following: a, β-tubulin sequence of S. aurantiacum, GenBank accession number GU126389; b, β-tubulin sequence of S. apiospermum, GenBank accession number FJ904910; c, β-tubulin sequence of P. boydi, GenBank accession number AJ889863; d, β-tubulin sequence of S. prolificans, GenBank accession number AB362937; e, β-tubulin sequence of S. dehoogii, GenBank accession number GU126407; and f, β-tubulin sequence of P. minutispora, GenBank accession number AM409107.

Fig. 1.

Nucleic acid sequences of minimum P. boydii-specific LAMP reaction unit of different haplotypes. Abbreviations: H, haplotype; F, forward primer; B, backward primer; c, complementary. Sequences: H1, CBS 101.22 (T); H2, CBS 117390; H3, DH 16566; H4, CBS 254.72; H5, CBS 116892. LAMP inner (FIP and BIP) and outer (F3 and B3) primer pairs are indicated by arrows. The FIP (BIP) primer consists of F1c (or B1c) and F2 (B2). In the first step of a LAMP reaction, new DNA strands were synthesizes from the F3 and B3 primers. In the next step, the newly synthesized strands would be recognized by FIP and BIP which bound both sense and antisense strands of target DNA in order to start loop-mediated autocycling amplification.

DNA extraction.

A MagNAPure LC isolation station was used for the DNA extraction for qPCR. 200 μl of the conidial suspension was transferred to MagNA Lyser Green bead tubes (Roche Applied Science) and homogenized for 20 s at 6,500 rpm by using the MagNA Lyser instrument and subsequently processed with a total nucleic acid isolation kit according to the manufacturer's protocol (Roche Applied Science) (1). Nucleic acids were resuspended in 50 μl of elution buffer. The DNA extraction for RLB was the CTAB (cetyltrimethylammonium bromide) method as previously described (11). Briefly, this method combines enzymatic (proteinase K) and mechanical steps (glass beads) in the presence of CTAB. For DNA extraction for LAMP, a simple method with Whatman FTA filter papers was used. Aqueous suspensions (100 μl) containing conidia and hyphal fragments were first treated by three to four freeze-thaw cycles coupled with vortex mixing for 30 s in the presence of 20 to 40 acid-washed glass beads (diameter, 425 to 600 μm; Sigma). The suspensions were then applied directly to FTA filters, and punches (2 mm in diameter) were removed from dried FTA filters and washed twice with 100 μl of sterile water. After the washing step, the filters were again air dried, and LAMP reagents were added directly to the FTA filters (4).

qPCR.

Quantitative real-time PCR was performed in a volume of 25 μl, consisting of 12.5 μl of LightCycler 480 master mix (Roche Applied Science), 2 μM concentrations of each primer, 0.8 μM concentrations of each probe, and 2.5 μl of template. The PCR thermal profile consisted of an initial incubation at 95°C for 10 min, followed by 45 cycles 95°C for 15 s and 64°C for 30 s. Species were detected separately. Amplification, detection, and data analysis were executed by the LightCycler 480 system (Roche Applied Science). A Cp (crossing point) value more than 40 was considered negative. The master mix was prepared in a biosafety cabinet. The qPCR was run in a room separate from where the PCR master mix or DNA was prepared.

PCR-RLB.

Duplex-PCR amplification of BT2 was performed at 94°C for 5 min, followed by 35 cycles 94°C for 45 s, 56°C for 45 s, and 72°C for 90 s, followed in turn by an elongation step at 72°C for 7 min in a mixture of of 1× GoTaq Green Master Mix (Promega), 400 nM primers, and 1 μl of template. RLB was modified as described by Bergmans et al. (2) using a top-down hybridization temperature and a critical washing temperature, as well as adjustment of sodium dodecyl sulfate (SDS) and probe concentrations. Briefly, solutions with 5′-amino-linked oligonucleotide probes ranging from 50 to 2,500 pmol/lane were coupled covalently to an activated Biodyne C membrane in a line pattern with a miniblotter (Immunetics). The membrane was incubated in 0.1 M NaOH for 10 min at room temperature, washed in 2× SSPE (360 mM NaCl, 20 mM Na₂HPO₄ · H₂O, 2 mM EDTA) with 0.1% SDS at 42°C. Then, 10-μl portions of the biotin-labeled PCR products were diluted in 150 μl of 2× SSPE–0.1% SDS, denatured at 100°C for 10 min, and applied to the membrane immediately in a perpendicular way, without chilling on ice. Afterward, the membrane was incubated in the miniblotter for 10 min at 70°C to maintain denaturation of the high-GC-content PCR products and subsequently hybridized for 30 min at 50°C. The membrane was removed from the miniblotter and washed twice in 2× SSPE–0.1% SDS for 10 min at 60°C exactly. Subsequently, the membrane was incubated for 30 min at 42°C with streptavidin-peroxidase (Boehringer), diluted 1:4,000 in 2× SSPE–0.5% SDS, washed twice for 6 min in 2× SSPE–0.5% SDS, and twice for 6 min in 2× SSPE. The membrane was then incubated with ECL detection liquid (GE Healthcare) and exposed to an ECL hyperfilm (GE Healthcare) for 2 to 10 min. After routine development and fixing, hybridization signals were visible by naked eye. The membrane with probes could be reused at least 20 times after stripping off the PCR products by incubation of the membrane for 2 × 10 min at 80°C in 1% SDS.

LAMP.

The LAMP reaction was conducted in a reaction mixture with 0.4 μM concentrations (each) of F3 and B3, 1.6 μM concentrations (each) of FIP and BIP, 200 μM deoxynucleoside triphosphates, 0.8 M betaine (Sigma), 20 mM Tris-HCl (pH 8.8), 10 mM KCl, 10 mM (NH₄)₂SO₄, 4 mM MgSO₄, 0.1% Triton X-100, 8 U of the Bst DNA polymerase large fragment (New England Biolabs), and one punch of an FTA card or 1 μl of genomic DNA as a template. The reaction mixture without Bst DNA polymerase was first incubated at 95°C for 5 min and chilled on ice; 8 U of Bst DNA polymerase was then added, followed by incubation at 63°C for 60 min and heating at 85°C for 2 min to terminate the reaction (34). The reactions were carried out either in a water bath or in a PCR machine. Species-specific reactions were carried out in separate tubes. Temperatures ranging from 60 to 70°C were tested to find the optimal condition, which was determined to be 63°C for all sets. The resulting amplicons were detected either by visual observation after the addition of 1 μl of 1/10 dilution of SYBR green I (Cambrex Bioscience) or by gel electrophoresis in 1% agarose gels stained with ethidium bromide. Careful precautions against cross-contamination were taken during sample collections and preparations by using separated rooms and filtered tips.

RESULTS

qPCR.

Identification down to the species level was possible with the probes for S. aurantiacum, S. prolificans, and P. minutispora. Since S. apiospermum and P. boydii share the same primers and probe, differentiation between them is impossible. The probes for S. aurantiacum and S. prolificans did not cross-react with other Pseudallescheria and Scedosporium spp., but S. apiospermum, P. boydi, and P. minutispora cross-reacted with each other (data not shown).

PCR-RLB.

Duplex-PCR amplification of BT2 yielded clear Pseudallescheria and Scedosporium-specific bands (data not shown). Although the closely related species Petriellopsis desertorum was also amplified, it gave a signal only with the group-specific probe PS_P on the blot (data not shown). This assay was specific for five species except S. dehoogii. Nonspecific signals were found with S. dehoogii strains with probes of S. apiospermum, P. boydii, and P. minutispora. Nonspecific signals varied among different haplotypes due to nucleotide variation at the probe region. However, each S. dehoogii strain was identified correctly according to the strongest signal. The group-specific probe matched well with the five species-specific probes (Fig. 2).

Fig. 2.

Results of reverse line blot of PCR products. Abbreviations: P, probe. Horizontal lanes indicate the species-specific probes and a group-specific probe (PS_P) (three lanes per probe). Vertical lanes indicate PCR products from isolates (seven strains per species): lanes 1 to 7, S. dehoogii; lanes 8 to 14, S. aurantiacum; lanes 15 to 21, S. apiospermum; lanes 22 to 28, P. boydii; lanes 29 to 35, S. prolificans; and lanes 36 to 42, P. minutispora.

LAMP.

All primers hybridized specifically with their respective DNA targets from each haplotype (data not shown). In all cases, we could distinguish LAMP-positive samples from LAMP-negative samples simply by visual observation after the addition of SYBR green I or gel electrophoresis. There was an agreement in the detection of the amplification product by gel electrophoresis and the addition of SYBR green I. The identification was shortened to 2 h starting from DNA extraction by FTA filters.

Analytical specificity.

To determine the analytical performance of these methods, DNAs from 60 CBS reference strains covering each haplotype of the species of interest were tested. Nine strains from other related Pseudallescheria and Scedosporium species were analyzed in order to examine the specificity. For comparison, genomic DNAs of 10 non-Pseudallescheria and Scedosporium fungal isolates, including Aspergillus, Candida, and Exophiala spp., were subjected to the detection methods outlined above. No cross-reaction with other related Pseudallescheria and Scedosporium species or non-Pseudallescheria/non-Scedosporium fungal isolates was observed.

Analytical sensitivity.

Analytical sensitivity was evaluated both by conidial spiking and DNA dilution. Ex-type strains of each species of interest were cultured in potato dextrose agar (PDA) and incubated at 37°C for 1 week. Conidia were harvested in 5 ml of 0.85% sterile saline–0.01% Tween 80 solution and filtered through an 11-μm-pore-size filter (6). Serial 10-fold dilutions of conidia and DNA were prepared in duplex in culture-negative and PCR-negative sputum samples from CF patients to achieve final concentrations ranging from 5 × 10³ to 5 cells/μl and 10 ng/μl to 10 fg/μl. The detection limit was 5 × 10³ cells/μl, as well as 20 pg of pure DNA for PCR-RLB and 50 cells/μl for qPCR. The LAMP assay reliably yielded positive signals from one punch of FTA paper with spiked sputum samples containing 5 × 10² cells/μl conidia, as well as 5 × 10³ cells/μl conidia and 20 pg of pure DNA extracted by CTAB (Fig. 3). No difference between species was found (data not shown). The qPCR combined with MagNAPure appeared to be the most sensitive method, followed by LAMP and PCR-RLB combined with FTA and CTAB, respectively.

Fig. 3.

Electrophoresis and visible assay of LAMP products. Genome DNAs from 10-fold diluted conidia of S. aurantiacum (CBS 116910) extracted by FTA cards were used as templates. (A) Lane 1, smart DNA marker; lanes 2 to 6, 5 × 10³ cells/μl, 5 × 10² cells/μl, 50 cells/μl, 5 cells/μl, and a tube without DNA template, respectively. (B) Tubes 1 to 5, visual appearance by the naked eye, 5 × 10³ cells/μl, 5 × 10² cells/μl, 50 cells/μl, 5 cells/μl, and a tube without DNA template, respectively; tubes 6 to 10, under UV transillumination, 5 × 10³ cells/μl, 5 × 10² cells/μl, 50 cells/μl, 5 cells/μl, and a tube without DNA template, respectively.

DISCUSSION

Scedosporium strains are known to exhibit a high degree of genetic variability (8, 39). Different numbers of haplotypes of these species are known to exist, with a maximum of 16 within P. boydii (Lackner et al., submitted). Only S. prolificans is relatively homogeneous, being recognizable by a set of stable diagnostic features (36). Due to the limited difference between species and significant variation within species, the ITS region appeared to be inadequate for primer and probe design for the methods applied in the present study. Even with BT2, the regions selected for potential primers and probes differed by only one or few nucleotides for some of the species pairs. Particularly, the complex of P. boydii and its heterothallic sister species S. apiospermum are very close to each other. Conversely, there is significant variation within both species. Five haplotypes were found within P. boydii at the primer region of LAMP (Fig. 1). With qPCR no specific primers and probes could be found in ITS or BT2 to distinguish between them, Pseudallescheria boydii sharing the same set of primers and probes with S. apiospermum. In addition, S. apiospermum, P. boydii, and P. minutispora cross-reacted with each other. For RLB, we found a unique potential probe region with three nucleotides difference between P. boydii and S. apiospermum. Thus, they were clearly distinguished by each of the probes. Pseudallescheria angusta, P. ellipsoidea, and P. fusoidea were recognized by the P. boydii probe. Scedosprium dehoogii was the most difficult species for identification. Two different primers were necessary to amplify all ITS haplotypes within this species (Lackner et al., submitted). For qPCR also two sets of primers and probes were required. With PCR-RLB, all nonspecific signals were due to S. dehoogii. Reliable differentiation of S. prolificans and S. aurantiacum was achieved by all three methods. LAMP provided very specific results, recognizing each haplotype of all six species despite low ratios of interspecific diversity and intraspecific variability. The efficiency and sensitivity of LAMP to recognize different, closely related entities with a single primer agreed with what was found in previous publications (18, 19, 24).

Castelli et al. (6) developed two real-time PCR-based assays using molecular beacon probes targeting the ITS1 region of ribosomal DNA for the detection of S. prolificans and S. apiospermum DNA. When Pseudallescheria and Scedosporium species other than S. prolificans and S. apiospermum were encountered, false-negative results could be obtained. The reason a false-negative result was obtained by microarrays developed by Bouchara et al. was that the probes designed could not detect all of these species (5). In our study, with the limitations in adequate probe design, qPCR was unable to distinguish all species. This included S. apiospermum and P. boydii, which are the prevalent clinical species of the genus. However, the sensitivity is the highest. It is probably better to use qPCR as a monitoring technique once the pathogen has been diagnosed, taking advantage of the high sensitivity and quantification. The PCR-RLB assay was able to identify five species except S. dehoogii. The latter species has not been proven to have clinical relevance, and thus PCR-RLB is sufficient for use in clinical diagnostics. If S. dehoogii strains are also present, correct signals were always stronger than nonspecific signals. The nonspecific signals caused by S. dehoogii were due to insufficient differences with S. apiospermum, P. boydii, and P. minutispora at the probe regions. Thus, every species could easily be distinguished, making the technique also suitable for identification of environmental isolates. Up to 43 targets in 43 individual specimens can be compared simultaneously (23). LAMP is one of the nucleic acid amplification tests applicable to microbial identification in various fields, such as diagnosis of infections (22). Using a set of two specifically designed inner primers and two outer primers that recognize six distinct regions of the target DNA, LAMP was easily performed isothermally for ∼1 h. The target sequence specificity of a LAMP reaction appears to be higher than that of PCR (29). Also, LAMP has the advantages that the specificity and sensitivity of detection appears not to be affected by the presence of known PCR inhibitors (10, 21, 29, 30). The LAMP assays developed in the present study readily distinguished six species of interest with high specificity.

To simplify DNA extraction for LAMP, we used commercially available Whatman FTA filters. The filter matrices are fibrous cards impregnated with chelators and denaturants that lyse and inactivate most microorganisms. Released large nucleic acids become physically trapped within the fibers of the FTA matrix and are preserved intact, while cellular debris can be removed by simple washes of the inoculated card. A recent study (4) evaluated the possibility of using prepunched Whatman FTA filter paper with fungal suspensions subjected to PCR. This material was sequenced successfully, targeting the partial β-tubulin gene (TUB). We found that with prepunched FTA filters the sensitivity was lower due to the small volume of samples added. A 100-μl droplet at the center of FTA filters and a punch removed from the center afterward was recommended for the detection of a few targets (28). FTA-LAMP was found to be significantly more sensitive than FTA-PCR (24) and ten times more sensitive than CTAB-LAMP.

Useful characteristics of FTA-LAMP as described here particularly concern the combination of highly sensitive and specific DNA amplification, as well as extraction simplicity for individual strains. PCR and other sensitive molecular techniques are best conducted in well-equipped laboratories only. In contrast, FTA-LAMP may also be used at local clinics. Thus, we conclude that—based on economics, practicality, and need—the FTA-LAMP assay is recommended for the identification of Pseudallescheria and Scedosporium species.