Introduction
The yeasts of the genus Malassezia are classified in the phylum basidiomycota, order Malasseziales, family Malasseziaceae (Crespo et al., 2008a; Gaitanis et al., 2012). These yeasts are considered part of the normal skin microbiota in humans and animals (Hernández, 2005). This genus is physiologically characterized for being lipid-dependent due to its incapacity to synthesize saturated fatty acids that is manifested by requiring aexogenous source of these fatty acids to grow; however, M.pachydermatis is an exception given the lack of need for lipids to grow in cultures (Giusiano, 2006). Based on these characteristics, the media commonly used are Dixon’s and Leeming, and Notman agars that allow an adequate isolation of these yeasts (Hernández, 2005; Kaneko et al., 2007).
Seventeen Malassezia species have been described so far, being lipid-dependent the following ones: M. furfur, M. globosa, M. obtusa, M. restricta, M. slooffiae, M. sympodialis, M. dermatis, M. japónica, M. yamatoensis, M. nana, M. caprae, M. equina, M. cuniculi, M. brasiliensis, M. psittaci and M. vespertilionis (Guého et al., 1996; Sugita et al., 2003; Hirai et al., 2004; Cabañes et al., 2007; Cabañes et al., 2011), with the exception of M. pachydermatis (Giusiano, 2006; Lorch et al. 2018) as already mentioned.
In animals, M. slooffiae, M. globosa, M. sympodialis and M. pachydermatis have been reported in the ear canal of dogs and cats (Pulido et al., 2010; Salah et al., 2010); M.nana in cattle with or without external otitis and also in the ear canal of healthy horses (Hirai et al., 2004; Aldrovandi et al., 2016), and M. sympodialis, M. slooffiae, M. furfur and M. pachydermatis in the ear canal of pigs (Nardoni et al., 2010).
There have been few studies on the presence of Malassezia spp. as part of the normal microbiota in horse skin; however some reports have shown them to be present in different areas of the skin surface, such as the inguinal area, back, perineum and ear canal, where species such as: M. pachydermatis, M. furfur, M. restricta, M. slooffiae, M. obtusa and M. globosa have been isolated from; in addition, M. furfur, M. restricta, M. sympodialis and M. globosa have been isolated from the axillar region (Zia et al., 2014; White et al., 2006; Crespo et al., 2002), and M. equina from the udder cleft of mares as well as from the preputial fossa of stallions and geldings (Cabañes et al., 2007; White et al., 2006).
Given the limited number of studies worldwide and the lack of studies on Malassezia spp. as part of the normal equine skin microbiota in tropical regions as Colombia, the main objective in this study was to isolate, identify and characterize these yeasts from different body regions of horses with no dermatological diseases.
Materials and Methods
Ethical Considerations
This research project was approved by the Bioethical committee of the science faculty of the Pontificia Universidad Javeriana by resolution No 14 on 08-10-14. It was also approved by the University Animal Care committee, by resolution C-023-14 on 01-10-14.
Sampling population
During a study period of 4 months, 22 horses, mares, geldings and stallions with ages that ranged between 26 month and 12 years, admitted to the Large Animal Clinic-Veterinary Teaching Hospital at the National University of Colombia were studied. These horses were presented with various clinical diseases none involving the skin, but these diseases did not represented a risk factor for yeast colonization of the skin or ear canal. The patients did not have any antibiotic, antimycotic nor costicosterioid treatments.
Sampling
The skin of the twenty-two horses was sampled using swabs. Samples were collected from the preputial area (n=10), mammary gland (n=12), inguinal area (n=16) and both ear canals (n=44). Eighty-two samples were collected and transported in sterile tubes at room temperature to the microbiology laboratory at the Pontificia Universidad Javeriana within a two hours-range after sampling.
Sample processing
Each sample was examined directly by making a slide smears using Gram staining to identify any yeast-like structure, oval or rounded blastoconidia of different sizes, filament like structure and bacteria. All samples were cultured on modified Dixon’s agar (malt extract 36 g, Peptone 6 g, Ox bile 20 g, Tween 40-10 ml, Glycerol 2 ml, Oleic acid 2 ml, Agar 12 g, deionized water 1,000 ml) (Resusta et al., 2007), and Sabouraud agar (Oxoid Hampshire, United Kingdom) supplemented with chloramphenicol (Sigma-Aldrich - St. Louis, MO, USA), and were incubated at 32 °C during 5 days (Hernández, 2005; Ashbee 2007; Pulido et al., 2010; Cafarchia et al., 2011; Aldrovandi et al., 2016).
All colonies morphologically compatible with Malassezia spp. were plated again to obtain pure colonies and describe them macroscopically along with the different morphological characteristics such as size, texture, color, shape, as well as the margin and surface of minimum 10 colonies. In order to describe adequately the microscopic appearance, 30 cells were used to measure length and width; such measurements were made using light microcopy (software Leica Microsystems®.DM 100 LED version 2.1.0.).
The colonies were considered to be Malassezia spp. (positive to Urea), underwent a series of biochemical and physiological tests that included assimilation of lipid supplements as Cremophor-EL (Sigma-Aldrich St. Louis, MO, USA), and Tweens (Merck Darmstadt, Germany/Sigma-Aldrich St. Louis, MO, USA), growth at 37 and 40 °C, enzymatic tests such catalase and β-glucosidase (Cafarchiaand Otranto, 2004; Kindo et al., 2004; Ashbee, 2007; Crespo et al., 2008b; Guého-Kellermann et al., 2010). Additionally, phospholipase activity was evaluated using 10% egg-yolk Sabouraud agar (Oxoid Hampshire, United Kingdom) on 4 points plating manner. The results were evaluated after 21 days using the averaging of the Pz values (Cafarchia and Otranto, 2004; Coutinho, 2005; Hurtado-Suarez et al., 2016).
The colonies that resulted morphologically and biochemically compatible with Candida spp. (Urea negative) were confirmed by chromogenic Chromagar Candida® (Becton Dickinson GmbH Heidelberg, Germany).
Culture quality control and strain storage
Reference strains of M. furfur CBS 7019, M. pachydermatis CBS 1879, M. sympodialis CBS 7222, M. slooffiae CBS 7956 and M. globosa CBS 7966 were used. The control strains Candida albicans ATCC 90028, Candida krusei ATCC 6258 and Candida parapsilosis ATCC 22019 were used as positive controls for phospholipase activity. All strains, including the isolated ones, were preserved in skim milk at -20°C.
Molecular identification
DNA extraction was performed using the fungi/yeast genomic DNA kit (Norgen® Thorold, ON, Canadá). The extracted DNAs were treated with RNAse (Promega, Madison, USA), 2 mg/mL at 37 °C for 4 hours (Gemmer et al., 2002), followed by amplification of region 5.8S DNAr, using primers ITS3 (5’-GCATCGATGAAGAACGCAGC-3’) and ITS4 (5’-TCCTCCGCTTATTGATATGC-3’) (Gaitanis et al., 2002; Hernández, 2005). The amplification was done by using PCR at reaction volumes of 50 μL having 45 μL of PCR SuperMix (1.1X) (Invitrogen CA, USA), 1 μL of each one of the primers (10 pmol/μL) and 3 μL of genomic DNA (20 ng/mL). The conditions of the reaction were an initial denaturation cycle at 95°C for 5 minutes, 30 cycles at 95 °C for 1 minute, 55 °C for 1 minute and at 72 °C for 1.5 minutes, and an extension final cycle at 72 °C for 5 minutes (Gaitanis et al., 2002; Hernández, 2005). The amplification products were detected by 1.5% agarose gel electrophoresis (Promega Madison, USA) with buffer TBE 1X (Promega Madison, USA), they were stained with ethidium bromide (Invitrogen CA, USA). The gen 5.8S RNAr amplification products were sent for purification and sequencing to Macrogen® Inc (Korea). The sequences were analyzed using BLASTn (https://blast.ncbi.nlm.nih.gov/Blast. cgi).
Results
Cytological examination of the direct Gram stained smears showed yeasts of different forms and sizes in 17.07% (n=14) of all swab samples. The macroscopic morphology of the colonies that were identified as Malassezia spp. were cream in color, flat and small with smooth margins with diameters between 1-2.1 mm. These yeasts were isolated mainly from left ear canals (n=5) and udders (n=2). The microscopic measurements showed different sizes of cells that varied from 1.5-2.4 μm wide and 2.2-3.2 μm in length.
Biochemical identification test allowed to determine that 50% (n=7) of the yeast isolates belonged to the Malassezia genus. The species that were identified included M.pachydermatis, M.globosa, M.obtusa, M slooffiae and M.sympodialis in 14.3% (n=1) for each isolate. The remaining two isolates were only identified as belonging to the Malassezia genus and were reported as Malassezia spp. (see Table 1). However, the molecular identification tests did not agree with the biochemical tests. The molecular tests identified M.nana as the most frequent species being 57.1% (n=4) of the isolates, followed by M. slooffiae with 28.6% (n=2) of the isolates; the remaining isolate was identified as belonging to a different genus Trichosporon asahii (see Table 1). The observed agreement between the two tests was 85.7% when genus identification was tested, but only 14.3% agreement regarding species identification. The rest of the yeasts isolated were determined to be Candida spp. in 35.7% (n=5), which were isolated from the preputial and inguinal areas, and Cryptococcus spp. was isolated from ear canal in 14.3% (n=2) of the cases.
On the other hand, Malassezia spp. was only determined to be present in only 4 out of 22 horses, showing a prevalence of 18.2% in this sample.
In this present study, 28.57% (n=2) of the isolates were phospholipase positive, which was very high for M.nana and high for M. slooffiae. The remaining 57.1% (n=4) of the isolates did not present this activity.
Candida spp. was isolated in 5 samples presenting negative urease activity and grew in Sabouraud agar. When using Medio CHROMagar Candida BD Becton-Dickinson® media Candida glabrata was determined in 80% (n=4) of these isolates and Candida tropicalis in 20% (n=1) of the cases. Finally, two samples yielded yeast belonging to genus Cryptococcus, basing their classification on positive urease activity and the presence of capsule, when stained with indian ink (Resusta et al., 2007).
Due to the small numbers of isolates, it was not possible to establish any association between isolates and variables such as age, sex or breed.
Discussion
Microscopic characteristics of the colonies identified as Malassezia spp. that included shape, margin, size, volume and texture are in accordance with the characteristics reported by several authors (Ashbee, 2007; Arenas, 2008; Crespo et al., 2008a; Guého-Kellermann et al., 2010; Hurtado-Suárez et al., 2016); however, there were differences in colony sizes, which could be attributed to possible limitations when measuring the organisms due to the distribution of the colonies on the agar. Regarding the microscopic characteristics based on the observed differences in the cell length, they were also partially similar to what is reported. This latter observation could also be due to limitations in the measuring as consequence of the cell distribution on the smear. These observations clearly suggest that both methodologies have a limited value when these yeasts are to be identified (Guého et al., 1996; Sugita et al., 2003; Hirai et al., 2004; Hernández, 2005; Cabañes et al., 2007; Crespo et al., 2008b).
The Malassezia isolates were determined to be M. pachydermatis, M. globosa, M. obtusa, M. slooffiae and M. sympodialis with prevalence of 14.3% (n=1) for each one of them. These species have also been reported by other studies to be commonly isolated from normal equine skin (White et al., 2006; Crespo et al., 2002); however, isolation of M. sympodialis and M.pachydermatis from mammary gland contrasts with the reports by White (2005) and White et al.(2006), who isolated M. slooffiae and/or M.equina from this same anatomical site of healthy mares. According to biochemical test results, the Malassezia species was not identified (2.43%) in two isolates, because their results did not coincide with what has been reported in the literature for the different species (Guého et al., 1996).
The identification of one of the isolates as M. pachydermatis was based on its ability to grow on Sabouraud Agar at 32 °C, and to be able to assimilate all the Tweens (Guého-Kellermann et al., 2010; Cafarchia et al., 2011). Despite the fact that M. globosa and M. obtusa share similar characteristics such as their incapacity to grow at 40 °C and the ability to assimilate tweens, the β- glucosidasa test allowed to differentiate one from the other due to the lack of the enzyme by M. globosa (Hernández, 2005; Giusiano, 2006). In a similar manner, M. sympodialis was differentiated from M. slooffiae because it has the β- glucosidase enzyme, and also due to the capacity to assimilate Tweens 40, 60 and 80, while M. slooffiae assimilates Tweens 20, 40 and 60, according to the biochemical characteristics reported, it was possible to determine the species of the isolates (Guého et al., 1996; Sugita et al., 2003; Hirai et al., 2004; Giusiano, 2006; Kaneko et al., 2007; Salah et al., 2010).
In order to identify M.pachydermatis, the phospholipase activity was evaluated given that it has been reported to be present in higher levels in either isolates from healthy skin or with skin lesions. This previous test and the biochemical characteristic allowed identifying this species (Cafarchia & Otranto, 2004; Juntachai et al., 2009; Pini & Faggi, 2011; Ortiz et al., 2013). In the case of M. obtusa, M. slooffiae, M. globosa and M. restricta, it has been reported that they do not show phospholipase activity in healthy animals (Pini & Faggi 2011). In the present study, M. slooffiae and M.globosa showed phospholipase activity in contrast to what has been reported. The presence of this activity could not be regarded as a virulence factor given that they came from healthy equine skin samples.
M. slooffiae, M. obtusa, M. globosa, M. pachydermatis and M. sympodialis were the isolated species from the healthy skin in horses using phenotypical and biochemical tests. These findings are similar to several studies that have reported them as the most common species isolated as normal microbiota of the skin of healthy horses (Crespo et al., 2002; Giusiano, 2006; White et al., 2006).
It is important to highlight that the phenotypical and the molecular identification methods agreed in only one isolate. The molecular identification technique was able to identify the species of the two isolates that were classified only as Malassezia spp.by phenotype classification. These low agreement level could be due to the variations in the results of the biochemical test as seen in the isolates from subject 2 (see Table 1), suggesting the need to use molecular techniques to identify the Malassezia species (Gaitanis et al., 2006; Ko et al., 2011; Zia et al., 2015).
Based on the molecular tests used, M.nana (80%) was the most frequent species isolated from the ear canal of the horses in the study; it was followed in occurrence by M. slooffiae (20%). Similar findings were reported by Aldrovandi et al. (2016) and Crespo et al. (2002). They reported the same species from this anatomical site, but it is different from the findings by Shokri (2016), who isolated M.pachydermatis from the ear canal. An interesting finding was that Malassezia spp. was only isolated from ear canals and udders; however, it was not isolated from the other anatomical sample sites despite reports indicating that it has been isolated from those areas in other studies (Crespo et al., 2002; Paterson 2002; White et al., 2006; Shokri, 2016).
The frequency of isolation observed in this study was 18.2% and only one species for each horse was detected. Comparing with other studies, this frequency was lower than the reported by Crespo et al.(2002), White et al., (2006) and Shokri (2016), that found prevalences of 34,9, 54 and 60%, respectively, in healthy horses.
Candida spp. was also isolated from some of the study subjects. This yeast is considered as part of the normal microbiota of both the skin of horses and its environment (Sgorbini et al., 2008; Cafarchia et al., 2013; Różański et al., 2013). This genus may produce opportunistic infections, despite being part of the normal microbiota, in patients with compromise of their immunological status or alteration of the normal microbiota (Cafarchia et al., 2013). It has been also reported that Cryptococcus spp. has been detected in the environment where horses are kept, which may explain why it was isolated in this study (Różański et al., 2013); however, it has been implicated in mycotic pneumonia in horses (Higgins & Pusterla, 2006).
Malassezia spp. is considered a normal inhabitant of the horse skin, but it has the potential to become a pathogen in cases of abnormal skin microenvironment or cutaneous infections (Paterson, 2002). In conclusion, the prevalence of Malassezia spp. in the studied population was 18.2% (4/22), and the species M.nana and M. slooffiae were identified as part of the skin microbiota in these horses using molecular tests. This is the first report that identifies Malassezia spp. as part of normal horse skin microbiota in Colombia.