New Molecular Markers Distinguishing Fonsecaea Agents of Chromoblastomycosis

The species belonging to the genus Fonsecaea are the main causative agents of chromoblastomycosis. The invasive potential of Fonsecaea differs significantly among its various sibling species. Moreover, the lack of clarity on the virulence and availability of precise markers to distinguish and detect Fonsecaea species is attributed to the different ways of dissemination and pathogenicity. Therefore, the present study aimed to propose new molecular tools to differentiate between sibling species causing chromoblastomycosis. We used an infection model of chromoblastomycosis in BALB/c to study species-specific molecular markers for the in vivo detection of Fonsecaea species in biological samples. Specific primers based on the CBF5 gene were developed for Fonsecaea pedrosoi, Fonsecaea monophora, Fonsecaea nubica, and Fonsecaea pugnacius. In addition, a padlock probe was designed for F. pugnacius based on ITS sequences. We also assessed the specificity of Fonsecaea species using in silico, in vitro, and in vivo assays. The results showed that markers and probes could effectively discriminate the species in both clinical and environmental samples, enabling bioprospecting of agents of chromoblastomycosis, thereby elucidating the infection route of the disease.


Introduction
Human chromoblastomycosis (CBM) is a skin disease that is exclusively caused by the members of order Chaetothyriales (black yeasts and relatives). The infection occurs through inoculation of the pathogen by a trauma or injury caused by sharp natural materials, such as plant thorns or wooden splinters that carry the respective opportunistic pathogen. The agents of the disease may be morphologically indistinguishable from their environmental counterparts. The disease is one of the most prevalent implantation mycoses in the world and occurs particularly in the (sub)tropical regions of South America, Africa, and China [1]. The genus Fonsecaea comprises cryptic species related to the disease, known as F. pedrosoi, F. monophora, and F. nubica [2]. All these species are found in the environment and infect humans, and F. pugnacius is a recently described agent of CBM [3].
The invasive potential of Fonsecaea differs significantly among the species [3,4]. Species such as F. pedrosoi, F. nubica, and Cladophialophora carrionii are exclusively associated with skin diseases caused by implantation, whereas F. monophora and Cladophialophora bantiana may also cause primary brain infection by implantation and inhalation, respectively. The single strain known as F. pugnacius behaved differently, as it causes a chronic condition of CBM that finally leads to secondary cerebritis by dissemination to the brain, despite the apparent intact immunity of the patient [3]. This type of dissemination and the apparent conversion to another invasive morphology has not been observed in F. monophora and C. bantiana, and the question arises whether this virulent ability is characteristic of F. pugnacius [2,3].
The traditional methods for the identification of pathogenic fungi are based on morphological characteristics and antigen detection. However, these methods are time-consuming and have low specificity [5]. In general, rapid and accurate recognition of fungal pathogens to the species or strain level is essential for disease surveillance and implementation of disease management strategies, especially for deep and disseminated infections. The development of direct detection assays is challenging because fungal pathogens and opportunists may be present in human and natural environments at a very low density [6]. In this regard, Abliz et al. [7,8] and Andrade et al. [9] reported certain specific primers based on the rDNA ITS spacer region of the CBM agents F. pedrosoi and C. carrionii. Likewise, padlock probes have been used in rolling circle amplification (RCA) to supplement the diagnosis of several mycotic diseases [10]. These methods demonstrated good reproducibility and sensitivity with CBM agents such as F. monophora, F. nubica, F. pedrosoi, Cladophialophora, and Exophiala [11][12][13]. Moreover, the method has also been applied for the diagnosis of infections caused by Candida, Penicillium, Aspergillus, Scedosporium, Cryptococcus, and Trichophyton [14][15][16][17].
The present study aims to propose new molecular tools for discriminating the Fonsecaea sibling species related to chromoblastomycosis. Novel specific primers using variations in the centromere microtubulebinding gene (CBF5) were developed for pathogenic members of the genus Fonsecaea. In addition, a probe based on the ITS sequence variations was proposed for a rapid and sensitive assay using RCA to detect F. pugnacius, a recently described species causing CBM that took a fatal turn.

Fungal Strains and DNA Extraction
Fonsecaea and related species associated with CBM that were used as references are presented in Table 1 [19].
GenBank Accession Numbers for the CBF5 Gene and Species-Specific PCR Primers Design for Fonsecaea Species In order to design specific primers, the gene for centromere/microtubule-binding protein cbf5 (CBF5) of the species was used. The sequences were deposited in GenBank (NCBI) as follows: F. erecta (XM_018834995), F. nubica (XM_022646888), F. monophora (XM_022652626), F. pedrosoi (XM_013424470), F. pugnacius (in progress), and C. carrionii (KB822705) ( Table 2). The sequences were aligned using MAFFT v7 [20] (a multiple sequence alignment program) and were used for in silico screening. Based on the nucleotide polymorphisms, short informative regions with variations of 18-25 nucleotides were selected manually. The software Primer3 [21] was used to evaluate melting temperatures, %GC contents, dimer sequences, and mismatches in the sequences. Subsequently, the primers were evaluated using the Mfold software [22] for potential secondary structures, which could reduce the amplification efficiency. The primers that could generate products of different sizes for each species to facilitate visualization via electrophoresis were selected.

Species-Specific PCR and Assay Sensitivity
The total DNA of Fonsecaea and the related species associated with CBM and other pathogens described to date in the literature was used as templates (Table 1) in the PCRs for each pair of specific oligonucleotides for the CBF5 gene. The reactions were performed in a final volume of 25 lL, including 10 lL of Master Mix PCR buffer containing 2.0-2.5 mM MgCl 2 , 400 mM dNTPs each, and 50 U/mL Taq polymerase, 5.6 lL of water, 1 lL of each of the forward and reverse oligonucleotides (20 pmol/lL), and 2 lL of target DNA. PCR was conducted as described by Rodrigues et al. [19]. The PCR conditions were as follows: an initial denaturing step of 5 min at 95°C, followed by 35 cycles of 1 min at 95°C, 1 min at the annealing temperature, and 1 min at 72°C, followed by a final step of 10 min at 72°C. The sensitivity of primers was established using 10-fold serial dilutions of PCR products, starting with 30 ng/lL and ending with 0.03 fg/lL [23].

Padlock Probe Design for Fonsecaea pugnacius
The design of padlock probes for F. pugnacius (FPgP) was based on the ITS sequence variations among Fonsecaea species (Table 3). ITS sequences of closely related species in Capronia, Cladophialophora, Cyphellophora, Exophiala, Fonsecaea, Phialophora, and Rhinocladiella were used as references (Table S1). The sequences of ITS regions were accessed from NCBI, aligned by MAFFT online [20], and adjusted manually using MEGA7 software [24] to identify single-nucleotide polymorphisms (SNPs). To optimize the joining efficiency to target DNAs, padlock probes were designed with minimal secondary structure and with T m of the 5 0 end probe binding arm near to or above the ligation temperature  KCl, 10 mmol L -1 MgCl 2 , 0.1% IGEPAL, 0.01 mmol L -1 rATP, and 1 mmol L -1 DTT, with a total reaction volume of 10 lL. Ligation conditions were as follows: 94°C for 5 min, followed by five cycles of 94°C for 30 s, and 4 min of ligation at 63°C. The ligation was visualized by electrophoresis on a 1% agarose gel. Exonucleolysis was used to remove the unligated padlock probes and template PCR products to reduce ligation-independent amplification events. This was performed in a 10 lL reaction mixture containing 10 U each of exonucleases I and III (New England Biolabs; Hitchin, UK) with incubation at 37°C for 30 min, followed by 94°C for 3 min to inactivate the exonuclease reaction [12].

Rolling Circle Amplification
Rolling circle amplification (RCA) was executed for all strains tested (Table 1) in a 12 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 the template. RCA conditions were as follows: 65°C for 60 min, followed by 85°C for 2 min to inactivate the enzyme and stop the amplification. The results of RCA reactions were visualized by electrophoresis in a 2% agarose gel to validate the specificity of the probetemplate binding. Positive responses revealed a ladderlike structure, whereas negative responses showed a clean background. The sensitivity of padlock probe was evaluated to ensure reliable amplification at low levels of target DNA by 10-fold serial dilutions of PCR products, starting with 2 ng/lL and ending with 0.02 pg/lL [7].

Fonsecaea pugnacius in Animal Infection Model
An infection model was developed in immunocompetent 6-8-week-old male BALB/c mice infected with F. pugnacius (CBS 139214). The animals were maintained under standard laboratory conditions with controlled temperature (23 ± 2°C) and ad libitum access to food and water. This model was used to evaluate the efficiency of padlock probe in vivo to validate the padlock probe (FPgP) specificity in biological samples. The mice were divided into four groups of six animals each; three groups were inoculated with F. pugnacius, and one negative control group was inoculated with sterile phosphatebuffered saline (PBS). Subsequently, the animals were infected via intradermal (ID) and intraperitoneal (IP) routes with 100 lL of 1 9 10 6 F. pugnacius propagules or sterile PBS. The mice were monitored weekly for up to 21 days post-inoculation and were killed with CO 2 anesthesia in an appropriate chamber at 7, 14, and 21 days post-infection [8]. The samples of the brain, lungs, liver, kidneys, footpad, and blood were aseptically collected from the infected mice and evaluated with in vitro hybridization using F. pugnacius specific primers (FOPU) and the padlock probe (FPgP).

Results
The oligonucleotide primers used for F. monophora (FOMO), F. nubica (FONU), F. pedrosoi (FOPE), and F. pugnacius (FOPU) were specific to the target species studied. The primers amplified fragments of different sizes under varying reaction conditions of temperature and MgCl 2 concentration ( Table 2).  (Fig. 1). The RCA padlock probe for F. pugnacius (FPgP) was designed based on the rDNA ITS region, as described by Najafzadeh et al. [21] for F. pedrosoi, F. monophora, and F. nubica. The padlock probe designed had 94 bp as follows: 5 0 P-cagggcggtc-ctccagcggatcaTGCTTCTTCGGTGCCCATtacgag-gtgcggatagctacCGCGCAGACACGATAgtctagagaa cacccgtga3 0 , with annealing and melting temperatures of 47.4°C and 62.4°C, respectively ( Table 4). The in silico assays demonstrated that the RCA padlock probe designed was specific to F. pugnacius (Fig. 2). Optimal results for RCA padlock probe specificity were obtained at a concentration of 10 -5 lM using PCR products diluted at 2 ng/lL (Fig. 2a, b). The RCA reaction amplification limit was observed at a dilution of 10 -5 PCR product (0.03 pg/lL), representing 10 9 DNA copies according to Avogadro's number (Fig. 2c).
Concerning the animal model infection by F. pugnacius, none of the infected animals presented neurological symptoms. The BALB/c mice infected via the ID route presented necrosis and swelling in the footpad tissues (Fig. 3a). The fungus was recovered only from the lungs (Fig. 3b). The DNA tissue samples (lungs, kidneys, liver, brain, footpad, and blood) of all animals tested were evaluated in vitro using the RCA probe and CBF5 primers. All collected samples reported positive results for DNA viability using the b-actin gene amplification (Fig. 3c). In vitro assays with the RCA probe demonstrated positive results for the brain, lungs, and footpad samples from the mice (Fig. 3d). It was not possible to detect the presence of the F. pugnacius DNA in biological samples using CBF5 primers, which can be explained by the low concentration of DNA in the samples.

Discussion
Chromoblastomycosis (CBM) is considered to be the second most prevalent fungal implantation mycosis worldwide. It is characterized by chronic cutaneous and subcutaneous lesions that develop at the site of previous transcutaneous trauma. The causative agents of CBM are found in decaying plant debris, including  Table 4 Rolling circle amplification padlock probes used in this study Nomenclature probe: FPgP-F. pugnacious padlock probe; the underlined sequences are the binding arm of padlock probe   Recently, a new causative agent, F. pugnacius, was described that resulted in cerebral dissemination, and thus was a variation to the primary cerebral cases caused by F. monophora [4]. However, infection sources and routes, and clinical evolution have not been entirely elucidated, partly owing to the lack of detection markers.
In general, diagnosis of CBM involves clinical and laboratory evaluation of tissue samples by direct microscopy to visualize muriform cells, followed by identification by sequencing [9,26], usually involving the ITS fungal barcoding region [27,28]. Moreover, serological tests have been developed to aid in CBM diagnosis although these are not routinely applied because of relatively low sensitivity and specificity, making new techniques more rapid and cost-effective for the diagnosis of CBM and other diseases caused by the black yeast-like fungi [29].
Currently, non-culture methods are being used to improve the sensitivity and specificity of mycological diagnosis [29]. Several studies have been published on the identification of specific CBM agents [12,14,30]. Although the ITS region efficiently detects specific causative agents, it is not always recommended because molecular identification may be hampered by sequence variability in the ITS domain caused by difficult-to-sequence homopolymeric regions [25]. In the present study, we developed primers to partially amplify the CBF5 gene encoding a small nucleolar ribonucleoprotein required for the correct processing of rRNA. Kendall et al. [31] reported that CBF5 gene mutation is relates to RNA-binding and pseudouridine (w) synthase activity. This mutation was considered as an evolutionarily conserved mutation in Saccharomyces cerevisiae [32].
Pseudouridine (w), a C-glycoside isomer of uridine, is the most common single-nucleotide modification found in functional RNA and often appears in highly conserved regions of homologous RNAs [33]. Several lines of evidence suggest an important role for this base modification in ribosome activity [34]. It has been hypothesized that w is an essential component of proper RNA folding and function [33]. In addition, this protein has been associated with pathogenicity and virulence in several microorganisms. When incorporated into RNA, w alters the RNA structure, increases the base stacking, improves base-pairing, and rigidifies the sugar-phosphate backbone. Earlier studies have also linked w, either directly or indirectly, to several human diseases [35].
We used the single mutation observed in this region to develop primers. The specific primers demonstrated specificity for the target variable region among the clinically cryptic Fonsecaea species. Moreover, the primers demonstrated high specificity in the regions selected for primer designing. No cross-reactions were observed with any of the 15 reference strains analyzed in the comparison (Fig. 2a). Sensitivity assays reported specific primers to detect low DNA concentrations (Fig. 2b). The novel molecular markers for Fonsecaea agents of CBM are very accurate, as demonstrated by 100% specificity of unambiguous profiles using gel electrophoresis.
The advantage of isothermal amplification methods is that these systems do not require a thermal cycler to produce, but just a simple platform, such as a heating block or water bath. Detection by loop-mediated isothermal amplification (LAMP) or RCA has become increasingly popular for SNP detection to rapidly identify the fungus in a variety of samples [6]. RCA has been proven to be useful for ultra-high specificity reactions. In this isothermal enzymatic process, a short DNA or RNA primer is amplified to form a long, single-stranded DNA fragment using a circular DNA template and special DNA polymerases. The method is particularly useful to differentiate closely related species or genotypes within species, which may differ by a single SNP only. The RCA padlock probe detects SNPs by creating a circular form to bind to the target DNA and to efficiently decrease the risk of nonspecific sequence replication. Accordingly, RCA is more potent compared to its linear alternative, yielding 10 9 or more copies of a circular sequence within an hour [36].
The RCA padlock probe (FPgP) in the present study demonstrated greater sensitivity than previously reported probes developed for Fonsecaea species [13,15]. Given the specificity of the RCA padlock probes, RCA was judged to be suitable for Fonsecaea detection and species differentiation without sequencing in a wide range of biological samples. The establishment of the test is relatively expensive; however, with high-throughput applications, the performance of testing will be rapid and cost-effective. Moreover, the developed specific primers could be useful for routine testing in clinical laboratories that screen a large number of samples.