Rapid Identification of Seven Waterborne Exophiala Species by RCA DNA Padlock Probes

The black yeast genus Exophiala includes numerous potential opportunistic species that potentially cause systematic and disseminated infections in immunocompetent individuals. Species causing systemic disease have ability to grow at 37–40 °C, while others consistently lack thermotolerance and are involved in diseases of cold-blooded, waterborne vertebrates and occasionally invertebrates. We explain a fast and sensitive assay for recognition and identification of waterborne Exophiala species without sequencing. The ITS rDNA region of seven Exophiala species (E. equina, E. salmonis, E. opportunistica, E. pisciphila, E. aquamarina, E. angulospora and E. castellanii) along with the close relative Veronaea botryosa was sequenced and aligned for the design of specific padlock probes for the detection of characteristic single-nucleotide polymorphisms. The assay demonstrated to successfully amplify DNA of target fungi, allowing detection at the species level. Amplification products were visualized on 1% agarose gels to confirm specificity of probe–template binding. Amounts of reagents were reduced to prevent the generation of false positive results. The simplicity, tenderness, robustness and low expenses provide padlock probe assay (RCA) a definite place as a very practical method among isothermal approaches for DNA diagnostics.


Introduction
Exophiala is a member of the ascomycete order Chaetothyriales (fungi), comprising the black yeasts and allies, which are frequently encountered as causative agents of disorders in humans and animals [1][2][3][4][5][6][7]. Human infections vary from commensalism or moderate cutaneous infection to fatal neurotropism with serious mutilation. Infections frequently involve patients without known immune disorder or underlying metabolic disease. Outside humans, especially cold-blooded waterborne vertebrates are susceptible to a variety of Exophiala species [2], some of these seem to be specific to certain host taxa [1]. As virulence factors, the capability to absorb alkylbenzenes, within sweat and nervous tissues of mammals and in the poisonous skin of amphibians, has been proposed [2,8,9]. Studies on epizootics from the older literature obviously show that black yeast infection is a relatively popular phenomenon in cold-blooded vertebrates [10][11][12][13][14][15]. Recent molecular reports demonstrate that various pathogenic species are concerned [1,2,16], which morphologically are extremely similar.
Some species can be classified by physiological characteristics, for example, temperature tolerance and nitrate assimilation, however, for many taxa molecular characterization are needed. [17]. Sequencing of the rRNA ITS location is generally adequate for routine species distinction in the genus Exophiala [18]. This method is relatively costly and timeconsuming and less suitable for large numbers of strains in case of monitoring of epizootics.
Rolling circle amplification (RCA) is an isothermal DNA amplification technique applying so-called padlock probes. The method has been proven to be fast, cost-effective and specific for identification of human and plant pathogenic fungi [6,[19][20][21][22][23][24][25], including black yeasts and relatives [26][27][28]. The 3 0 -and 5 0 -end strands of the probes hybridize next to one another at the target strand, leading to circularization of the molecule upon ligation. The circular molecule is consequently amplified isothermally with a DNA polymerase that lacks exonuclease activity, and the resulting product subsequently can be utilized with a second primer causing a cascade of amplifications. Because of the necessary accurate base pairing, the padlock probes have the ability to identify single position mutations [29][30][31].
In the present paper, we developed eight padlock probes on the basis of the ITS location to identify the most relevant species of Exophiala in animal infection and epizootics, viz. E. equina, E. salmonis, E. opportunistica, E. aquamarina, E. angulospora and E. castellanii, together with Veronaea botryosa as outgroup. The objective of the current study was to evaluate the practical applicability of the method and to assess its limitations.

Strains
The 62 isolates of Exophiala and Veronaea included 13 strains of E. equina, 3 of E. salmonis, 6 of E. opportunistica, 6 of E. pisciphila, 8 of E. aquamarina, 10 of E. angulospora, 7 of E. castellanii and 9 of V. botryosa (Table 1); together these strains formed the 'salmonis-clade' of waterborne mesophilic species [2]. Isolates originated from cold-blooded animals, from human infections and from the environment. Cultures are preserved on slants of 2% malt extract agar (MEA) and oatmeal agar (OA) (Difco, Brunschwig, Amsterdam, the Netherlands) at 24°C in the reference collection of the Centraalbureau voor Schimmelcultures (housed at Westerdijk Fungal Biodiversity Institute, Utrecht, the Netherlands). Affiliation to Exophiala was verified using a phylogenetic tree constructed with sequences of the partial SSU gene. Species identity was confirmed by sequencing rDNA internal transcribed spacer (ITS), partial btubulin (BT2), partial elongation factor 1-a (TEF1) and actin (ACT1) genes [2]. To evaluate the specificity of padlock probes, we tested four closely related

DNA Extraction and Amplification
DNA extraction and quality tests were executed using glass beads (Sigma G9143) based on the methods described formerly [32]. DNA concentration and quality were tested spectrophotometrically at 260 and 280 nm (Shimadzu, Kyoto, Japan). ITS amplicons were produced with primers V9G and LS266 as described earlier [33]. PCR conditions were as follows: 95°C for 5 min, followed by 35  . In order to optimize joining efficiency to target DNAs, the padlock probes were designed with minimal secondary structure and with Tm of the 5 0 end probe binding arm near to or above ligation temperature (63°C). To improve their discriminative specificity, the 3 0 -end binding arm was designed with a Tm 10-15°C under ligation temperature. The linker part of each Exophiala species-specific probe obtained from [22] and the 5 0 -and the 3 0 -binding arms were designed in that article. Sequences of the both primers used for RCA and the oligonucleotide padlock probes are shown in Table 2. The oligonucleotide probes applied were c. 92-99 bp in length and contained two adjacent target complementary sequences (14-20 bp) with a spacer region (63 bp) to help loop formation and provide a template for RCA primer binding. Specificity of the probes was proved by BLAST examination in GenBank and in a validated database of filamentous fungi available at CBS for research purposes. Ligation of Padlock Probe One microliter of ITS amplicon was mixed with 2 U pfu DNA ligase (Epicentre Biotechnologies, Madison, WI, U.S.A.) and 0.1 lmol l -1 padlock probe in 20 mmol l -1 Tris-HCl (pH 7.5), 20 mmol l -1 Cl, 10 mmol l -1 MgCl 2 , 0.1% Igepal, 0.01 mmol l -1 rATP and 1 mmol l -1 DTT, with an overall total reaction volume of 10 ll. Padlock probe ligation reaction was performed as described by [34] by one denaturation cycle for 5 min at 94°C, followed by five cycles of 30 s. at 94°C and 4 min ligation at 63°C.

Exonucleolysis
Exonucleolysis is needed to eliminate unligated padlock probe and template PCR product and therefore minimize subsequent ligation-independent amplification events. It had been done in a 20-ll vol by addition of 10 U every one of exonuclease I and III (New England Biolabs, Hitchin, UK) to the ligation mixture and incubation at 37°C for 30 min, followed by 94°C for 3 min to inactive the exonuclease reaction.

Rolling Circle Amplification (RCA) Reaction
Rolling circle amplification was executed in 50 ll reaction mixture containing 8 U Bst DNA polymerase  (New England Biolabs), 400 lmol l -1 deoxynucleoside triphosphate mix and 10 pmol of every RCA primer in distilled water with 2 ll ligation product as template. Probe signals were amplified by incubation at 65°C for 60 min, and accumulation of double stranded DNA products was visualized on a 1% agarose gel to validate the specificity of probetemplate binding. Positive responses revealed a ladder-like structure, while negative responses showed a clean background. When the exonuclease step is omitted, some weak signal might be visible in gel electrophoresis [24].

Results
The general fungal primers ITS1 and ITS4 amplified the ITS location of all studied isolates. The ITS alignment revealed appropriate positions for the development of padlock probes that were able to differentiate between seven Exophiala species and the Veronaea out-group tested in this study. Each of the eight infectious species had several distinctive nucleotide positions. The duration of the RCA assay was 2 h. The tested species were unambiguously distinguished from each other and from different black yeast and relatives in the order Chaetothyriales compared by ITS sequence analysis and included as negative controls: the four closely related, clinically relevant species used for comparison, i.e., Cladophialophora bantiana (CBS 678.79), Exophiala dermatitidis (CBS 525.76), E. bergeri (CBS 526.76) and Rhinocladiella mackenziei (CBS 650.93) yielded negative results with the species-specific padlock probes (data not shown). Positive responses proved to be consisted and highly specific in all strains; all individual strains responded with respective probes being and were correctly identified as Exophiala equina, E. salmonis, E. opportunistica, E. aquamarina, E. angulospora and E. castellanii, as well as Veronaea botryosa, the nearest neighbor of the salmonis-clade [2]. No cross reaction was observed between any of the Exophiala species (Fig. 1). Products of the RCA responses were visualized by electrophoresis on 1% agarose gels. Positive reaction showed ladder-like patterns after RCA, while with negative results the background stayed clean. When the exonucleolysis step was deleted, a single poor band was apparent on the gel representing a non-specific band that did not interfere with the RCA reaction (data not shown). The concordance of RCA results and identification by multilocus sequencing was 100%.

Discussion
Thermotolerance is generally considered as a prime condition for vertebrate pathogenicity. During the last decades, several black yeasts have been described in Exophiala that constantly lacked thermotolerance, but still were associated with animal disease, indicating that these fungi have other, intrinsic, temperatureindependent infectious abilities [2]. Infections were especially within fish and amphibians, but sometimes also in invertebrates [14]. Such infections seem to be relatively regular, at least in captive and farmed fish and amphibians. Outbreaks in farmed and aquarium animals could cause serious losses to aquaculture and fishery industries [34], but because of the spread nature of reports it is hard to estimate the magnitude of the problem. Rolling circle amplification is a powerful and easy, isothermal in vitro DNA amplification method emerging as a tool for quick detection of specific nucleicacid sequences in DNA samples [35]. The use of a padlock probe to circularize oligonucleotides was produced by Nilsson [36]. The technique is on the basis of the replication of a short, single-stranded DNA circle by Bst DNA polymerases at constant temperature. Sequencing of the internal transcribed spacer (ITS) is the gold standard for species recognition of black yeast and relatives, as it provides sufficient resolution between species [18]. For analysis of large numbers of isolates in case of outbreaks and epidemiological monitoring, sequencing is nevertheless costly, time-consuming and impractical [22]. Furthermore, validated databases for comparison are needed, as GenBank information is polluted with wrongly identified sequences; we used a research database on black fungi housed at the Westerdijk Fungal Biodiversity Institute and of which a selection has been deposited in the ISHAM ITS Database (www.its.mycologylab.org). The RCA reaction is relatively free of requirement for high priced laboratory equipment and could be done within 2 h isothermally at 65°C in a water bath, thermocycler, heat block or microwave. Nevertheless, positive signals are often visible 15 min after commencement of the RCA reaction when recognized by real-time PCR [19,22]. The amplification product can be visualized by agarose gel electrophoresis, but can also be visualized in gel-free methods applying fluorescence staining of amplified product by SYBR Green in combination with a UV transilluminator. The progress of RCA probes to distinguish single species or groups of species depends on the presence of adequate sequence information and useful species-specific polymorphisms in genes of precisely identified species.
The objective of the current study was to begin a screening technique based on RCA for highly specific and rapid detection of waterborne Exophiala species  which repeatedly occur in the form of outbreaks in farmed fish, enabling their unambiguous differentiation from related melanized fungi. The RCA method performed well elsewhere in the fungal Kingdom, e.g., in Candida, Aspergillus, Scedosporium [22], Cryptococcus [37], Trichophyton [19], Fusarium [21], human-pathogenic Exophiala [28], Talaromyces marneffei [24], Scedosporium [20], Rhizopus [23] and Fonsecaea [27]. A low-cost alternative to RCA might be loop-mediated isothermal amplification (LAMP). This technique uses a set of six oligonucleotide primers with eight joining sites hybridizing particularly to various parts of a target gene [38]. Najafzadeh et al. [27] compared RCA and LAMP detection for human-pathogenic Fonsecaea species and discovered that LAMP was extremely sensitive, but RCA become more specific.
In conclusion, RCA is a very fast (less than 1 working day), specific (down to the single-nucleotide level) and economical (no additional equipment required) method for specific and rapid identification of fungal pathogens where large numbers of strains need to be processed. Our results show a considerable potential of the method in the future in laboratories for fungal outbreak control, e.g., in farmed fish. The establishment of the test is relatively expensive, but with high throughput applications, the final result per strain will be rapid and cost-effective.