Population differentiation, antifungal susceptibility, and host range of Trichophyton mentagrophytes isolates causing recalcitrant infections in humans and animals

Dermatophytes were detected from skin scrapings and hairs of 24 humans and 35 animals who had previously received oral and topical treatment for a period of at least 28 days (Table 1). None of the patients at that time took any other medications or received immunosuppressive therapy. Cases of infection were diagnosed between 2018 and 2019 in Poland. Tinea capitis: n = 15 (62.5%) was the predominant clinical form in the human infections followed by tinea corporis: n = 7 (29.2%) and tinea unguium: n = 2 (8.3%). In the animals, ringworm located around the head and neck: n = 17 (48.6%), on the torso: n = 8 (22.9%), and in multiple sites on the body: n = 9 (25.7%) was reported. In turn, dermatophytes were isolated from 17 (70.8%) human and 27 (77.1%) animal samples that were positive in the real-time PCR tests. The sample was collected from lesion margins. Detection and identification of dermatophytes in the dermatological material from the patients was conducted using the real-time PCR assay according to Ohst et al. [35]. Reactions with pan-dermatophyte primers (F: 5′-AGCGCYCGCCGRAGGA-3′; R: 5′-GATTCACGGAATTCTGCAATTCAC-3′) and species-specific primers (F: 5′-CGGCGAGCCTCTCTTTAGT-3′; R: 5′-GATTCACGGAATTCTGCAATTCAC-3′) targeting ITS1, ITS2, and 5.8S rDNA gene sequences in combination with TaqMan probes (Derm5.8S: CGCATTTCGCTGCGTTCTTCATC) were made to identify clinical isolates. DNA isolation and real-time PCR were carried out using a DNeasy Blood & Tissue Kit and a Quanti Tect SYBR green PCR master mix (Qiagen, Hilden, Germany), respectively. A samples with dermatophyte identified in the real-time technique were directed to isolation of cultures. The species identification for full confirmation of the taxonomic position was based on macro- and microscopic examination according to de Hoog et al. [36] (Fig. 1).

The genetic diversity of the clinical isolates was performed with the melting point PCR (MP-PCR) method optimized and modified for dermatophytes [31]. Briefly, the first step was the digestion of the genomic DNA isolated from the cultures (phenol-chloroform method [37]) using HindIII (Thermo Fisher, Waltham, USA) endonuclease. Next, an adaptor, i.e. a mixture of two oligonucleotides Helper and Ligated (Helper: 5′-AGCTGTCGACGTTGG-3′, Ligated: 5′-CTCACTCTCACCAACAACGTCGAC-3′), was ligated to DNA fragments. The PCR was made with a primer terminated with the AGCTT adapter sequence (PowaAGCT 5′-CTCACTCTCACCAACGTCGACAGCTT-3′). Electrophoresis of all PCR products was carried out in 3% agarose gels. T. mentagrophytes CBS570.80 and CBS677.86 were used as the reference strains. All analyses were made in triplicate.

The production of virulence factors was evaluated using specific test media. The following tests were performed: production of keratinase (1), elastase (2), phospholipase (3), lipase (4), protease (5), gelatinase (6), and detection of haemolytic activity (7). Keratinase production (1): the medium was prepared as described by Scott and Untereiner [38]. The basal layer containing 0.5 g/l MgSO₄·7H₂O, 0.05 g/l KCl, 0.5 g/l K₂HPO₄, 0.1 g/l ZnSO₄·7H₂O, 0.1 g/l FeSO₄·7H₂O, 0.03 g/l CuSO₄, and 25.0 g/l agar was dispensed horizontally in sterilized test tubes. The upper layer was supplemented with 4 mg/ml keratin azure. Clinical isolates inoculated with the medium grew for 1 month in the dark at 37 °C. A change of the colour of the basal layer from milky white to blue was a positive result. Chrysosporium keratinophilum CBS104.62 was used as a positive control. The elastin activity (2) was determined on two-layer plates as described by Rippon and Varadi [39]. The basal layer containing 8.0 g/l nutrient broth and 20.0 g/l Agar Noble was poured in a thin layer onto Petri dishes; the upper layer was supplemented with 3.4 g/l elastin from bovine neck ligament. Brighter zones around the colonies after incubation at 37 °C for 14 days indicated elastinase activity. Pseudomonas aeruginosa ATCC15152 was used as a positive control. The phospholipase assay (3) was performed according to the method described by Gnat et al. [40]. Briefly, the medium containing (g/l) 10.0 g/l peptone, 20.0 g/l dextrose, 57.3 g/l NaCl, 0.005 g/l CaCl₂, 20.0 g/l agar, and 50 ml egg yolk was punctiform inoculated with a single dermatophyte colony in the centre of the plate and incubated for 14 days at 28 °C. A clear halo zone around the colony indicated phospholipase production. Candida albicans ATCC10231 was used as a positive control. The lipase activity (4) was tested on medium containing 10.0 g/l peptone, 5.0 g/l NaCl, 0.1 g/l CaCl₂, 20.0 g/l agar, and 10 ml Tween 80 as described by Muhsin et al. [41]. Isolates of dermatophytes were punctiform inoculated in the centre of the plate and incubated for 14 days at 28 °C. A clear halo zone of precipitation around the colony indicated lipase production. Malassezia furfur ATCC14521 was used as a positive control. Protease activity (5) was tested on casein medium with bromocresol green (BCG) as described by Vijayaraghavan et al. [42]. The medium containing 5.0 g/l meat peptone 1.5 g/l beef extract, 1.5 g/l yeast extract, 5.0 g/l NaCl, agar, 15, 10.0 g/l casein, 0.0015% (w/v) BCG, and 15.0 g/l agar was punctiform inoculated in the centre of the plate and incubated for 14 days at 28 °C. A zone of proteolysis around the colony was a positive result. Bacillus subtilis ATCC6633 was used as a control. The gelatinase assay (6) was performed according to the method described by Gnat et al. [40]. The medium containing 5.0 g/l neopeptone, 3.0 g/l beef extract, and 120.0 g/l gelatin was poured into tubes, inoculated, and incubated at 25 °C for a month. After this period, all tubes were cooled to 4 °C and tilted to check if the gelatin was liquefied. The gelatinase secretion was determined on the basis of the physical state of the medium: liquid medium was a positive result. Staphylococcus aureus ATCC29213 was used as a positive control. The haemolytic activity (7) was assessed using Columbia agar medium (BioMaxima, Lublin, Poland) supplemented with 5% defibrated sheep blood as described by Schaufuss et al. [43]. A needle was used for punctiform inoculation with a single colony, and then incubation at 28 °C lasted for 7 days. The presence of a translucent halo around the colony was a positive result. Staphylococcus aureus ATCC25923 was used as a positive control for β-haemolysis, Streptococcus pneumoniae ATCC6305 for α-haemolysis, and Clostridium perfringens ACTCC13124 for double zone haemolysis. The catalase activity was determined on medium containing 3% hydrogen peroxide and 2% agar. Wells cut out on a Petri dish were supplemented with the dermatophyte suspension. The plates were incubated at 28 °C for 6 h aseptically. The staining solution containing 2% of K₃Fe(CN)₆ and FeCl₃·6H₂O each was prepared ex tempore. The staining solution was poured into Petri dishes containing the samples, which were then shaken gradually until a green colour appeared. The staining solution was then filtered off, and the plate was rinsed and filled with distilled water. The appearance of a yellow ring around the well was considered a positive result. Staphylococcus aureus ATCC6538 was used as a positive control. For each test, the clinical isolates of the dermatophytes were incubated on MM-Cove medium with keratin azure (Sigma-Aldrich, Missouri, USA) for 14 days at 28 °C. Unless otherwise described, a halo zone diameter exceeding 50% of the dermatophyte colony diameter was regarded as a positive result. Each test was made in triplicate.

The host range of the dermatophytes was determined with the method described previously by Gnat et al. [44]. The basis for the determination was the diverse keratinolytic activity against species-specific keratin substrates. Cultures used for this test were inoculated onto liquid Sabouraud glucose agar (BioMaxima, Lublin, Poland) at 37 °C for 15 days with shaking at 110 rpm (Multitron, Infors, Switzerland). Next, the mycelium was collected and mechanically homogenized, resuspended to a concentration of 1 in McFarland, and used as the source of enzymes. Hairs and fur obtained from Bos taurus (cow), Canis familiaris (dog), Cavia porcellus (guinea pig), Equus caballus (horse), Felis catus (cat), Homo sapiens (human), Ovis aries (sheep), Sus domesticus (pig), and Vulpes vulpes (fox) during routine hygiene operations were cut separately into 1-mm pieces, washed, and defatted. After drying, they were used as substrates. Keratin azure (Sigma-Aldrich, Missouri, USA) was used as a positive control and indicator substrate. The substrates and sources of the enzyme were added to a liquid minimal medium (MM-Cove) designed to determine keratinolytic activity at a concentration of 0.06% and in a volume 100 μl, respectively. The enzymatic activity (Uh⁻¹) was determined spectrophotometrically at a wavelength of 550 nm (SmartSpec, BioRad, USA) as mycelium growth per hour of incubation on incubation days 4, 7, 10, 15, 30, and 60. The test was performed in three replicates simultaneously, and differences between enzymatic activity in different periods of analysis as well as various substrates were assessed by Student’s t test using the R program version 3.6.3 (R Core Team, Missouri, USA).

In vitro testing of the susceptibility to allylamine, polyene, imidazole, triazole, and pyridinone derivatives as well as phenyl morpholine derivatives was performed according to the Clinical and Laboratory Standards Institute (CLSI) document M38-A3 (CLSI, 2018). Reagent-grade amorolfine (AMR), amphotericin B (AMB), ciclopirox (CPO), enilconazole (ENC), fluconazole (FLC), griseofulvin (GRE), itraconazole (ITC), ketoconazole (KTC), miconazole (MCZ), naftifine (NFT), terbinafine (TRB), and voriconazole (VRC) were obtained in the powder form (Sigma-Aldrich, Missouri, USA). Drug stock solutions were prepared in dimethyl sulfoxide (DMSO) to reach the final DMSO concentration in the wells below 1%. The drugs were analysed at the final concentration in the range of 0.002–2 μg/ml for allylamine, pyridinone derivatives, and phenyl morpholine derivatives, 0.004–4 μg/ml for polyenes, imidazoles, itraconazole, and voriconazole, and 0.06–64 μg/ml for fluconazole. The dermatophyte isolates were cultured on Sabouraud glucose agar (BioMaxima, Lublin, Poland) for 21 days, and conidial suspensions were prepared by gentle scraping mature colonies into sterile physiological saline containing 0.002% Tween 80. Homogeneous inoculum supernatants were collected, and their optical density (OD) at 530 nm was adjusted spectrophotometrically to an OD of 0.11 to 0.13, which ranged from 65 to 70% transmission, and the final density of inoculum was 1 × 10³ to 3 × 10³ CFU/ml. The inocula were diluted 1:50 in RPMI 1640 medium and incubated with the indicated concentrations of the antifungals in 96-well plates at 35 °C for 72 h. Minimum inhibitory concentrations (MICs) were determined visually using a reading mirror as complete inhibition of observable growth. Trichophyton mentagrophytes ATCC4439 and T. rubrum ATCC4438 served as quality controls for every new series of MIC plates. All tests were performed in triplicate, and differences between mean values were assessed by Student’s t test using the R program version 3.6.3 (R Core Team, Missouri, USA).

In this study, the dermatophytes isolated from human and animal infections in Poland were characterized for genomic diversity and relationships with the commonly used DNA fingerprinting technique based on PCR reaction, i.e. MP-PCR. Six genotypes of the 17 human isolates of T. mentagrophytes were distinguished (Fig. 2). In turn, nine genotypes were identified in the animal isolates. In each case, similar genomic diversity values of 35.3% and 33.3% were obtained for the clinical isolates from humans and animals, respectively. No relationship was demonstrated between the antifungal drug resistance profile and the MP-PCR profile. In addition, no relationship between the profile and the host species was found for the animal isolates. There were no profiles specific for human or animal isolates either, but specific electrophoretic patterns appeared in T. mentagrophytes isolated from both types of cases.

The enzyme profiles obtained for the human and animal isolates were not significantly differentiated (Table 2). All the tested clinical isolates of T. mentagrophytes showed keratinase, phospholipase, protease, and catalase enzymatic activity as well as haemolytic activity (Fig. 3). In turn, extracellular elastase was produced only by the clinical isolates obtained from humans, and lipase was produced by the animal isolates. Furthermore, no strains diverging from the relationship were found: when the enzymatic activity was positive, it was exhibited by 100% of the tested isolates, as in the case of negative results.

Particularly important is the species-specific keratinolytic activity, which is recognized as a determinant of the host range of dermatophytes. The degree of keratinolytic activity ranged from 0.8 Uh⁻¹ and 1.0 Uh⁻¹ on day 4 of incubation to 5.4 Uh⁻¹ on day 30 of incubation and 5.8 Uh⁻¹ on day 15 of incubation for the T. mentagrophytes clinical isolates from humans and animals, respectively (Fig. 4). The highest activity of keratinase enzymes of the tested isolates was recorded in media containing keratin from the fox (Vulpes vulpes), guinea pig (Cavia porcellus), and human (Homo sapiens) hairs. In these three cases, the degree of keratinolytic activity was statistically significantly higher (p < 0.05) than in the medium with keratin azure; it was 5.8 Uh⁻¹, 5.7 Uh⁻¹, and 4.9 Uh⁻¹ for the fox, human, and guinea pig hairs on day 15 of incubation for the animal isolates, and 5.2 Uh⁻¹, 4.95 Uh⁻¹, and 4.6 Uh⁻¹ for the human isolates, respectively. In turn, this value for keratin azure was 2.05 Uh⁻¹. Additionally, the geometric mean of keratinolytic activity on day 15 of incubation for three most active species was statistically significantly higher than for all the other species-specific hairs, i.e. 5.46 Uh⁻¹ vs. 3.13 Uh⁻¹ and 4.92 Uh⁻¹ vs. 2.98 Uh⁻¹ for the animal and human isolates, respectively. Interestingly, the animal T. mentagrophytes isolates exhibited higher keratinolytic activity against human hairs than all the others on day 30 of incubation. On the contrary, this relationship was not revealed for the human isolates, where the keratinolytic activity was at the highest level for the fox keratin in all periods.

The sensitivity to the antifungal substances was similar in the clinical isolates of T. mentagrophytes obtained from humans and animals. The MIC ranges, geometric means of MICs, MIC₅₀, MIC₉₀, and mode ratios of the 12 antifungal drugs used are summarized in Table 3. Phenyl morpholine derivatives, i.e. amorolfine, exhibited superior activity against strains obtained from both humans and animals with the lowest MIC₅₀ values. In turn, griseofulvin was found to exert the weakest in vitro effect and had the highest MIC₅₀ values in the tested isolates. Additionally, fluconazole had the widest MIC range, i.e. 0.06–32 μg/ml for the human isolates and 0.125–32 μg/ml for the animal isolates, respectively. Remarkably, MIC₉₀ values above 1 μg/ml were obtained for grisofulvin, enilconazole, ketoconazole, and fluconazole in both groups of the tested strains and for terbinafine and itraconazole in the case of the human T. mentagrophytes isolates. Interestingly, the lowest statistically significant geometric mean of the MIC values (p < 0.05) for the dermatophytes isolated from humans were noted for amorolfine. Furthermore, the lowest geometric mean MIC value (without statistical significance) was obtained for naftifine in the case of the animal isolates.