Utilization of matrix-assisted laser desorption and ionization time-of-flight mass spectrometry for identification of infantile seborrheic dermatitis-causing Malassezia and incidence of culture-based cutaneous Malassezia microbiota of 1-month-old infants

Authors

  • Mikachi Yamamoto,

    1. Laboratory of Space and Environmental Medicine, Graduate School of Medicine, Teikyo University, Tokyo, Japan
    Search for more papers by this author
  • Yoshiko Umeda,

    1. Teikyo University Institute of Medical Mycology, Teikyo University, Tokyo, Japan
    2. General Medical Education Center, Teikyo University, Tokyo, Japan
    Search for more papers by this author
  • Ayaka Yo,

    1. Laboratory of Space and Environmental Medicine, Graduate School of Medicine, Teikyo University, Tokyo, Japan
    Search for more papers by this author
  • Mariko Yamaura,

    1. Laboratory of Space and Environmental Medicine, Graduate School of Medicine, Teikyo University, Tokyo, Japan
    Search for more papers by this author
  • Koichi Makimura

    Corresponding author
    1. Laboratory of Space and Environmental Medicine, Graduate School of Medicine, Teikyo University, Tokyo, Japan
    2. Teikyo University Institute of Medical Mycology, Teikyo University, Tokyo, Japan
    3. General Medical Education Center, Teikyo University, Tokyo, Japan
    • Correspondence: Koichi Makimura, M.D., Ph.D., Laboratory of Space and Environmental Medicine, Graduate School of Medicine, Teikyo University, 2-11-1 Kaga, Itabashi, Tokyo 173-8605, Japan. Email: makimura@med.teikyo-u.ac.jp

    Search for more papers by this author

Abstract

Matrix-assisted laser desorption and ionization time-of-flight mass spectrometry (MALDI-TOF-MS) has been utilized for identification of various microorganisms. Malassezia species, including Malassezia restricta, which is associated with seborrheic dermatitis, has been difficult to identify by traditional means. This study was performed to develop a system for identification of Malassezia species with MALDI-TOF-MS and to investigate the incidence and variety of cutaneous Malassezia microbiota of 1-month-old infants using this technique. A Malassezia species-specific MALDI-TOF-MS database was developed from eight standard strains, and the availability of this system was assessed using 54 clinical strains isolated from the skin of 1-month-old infants. Clinical isolates were cultured initially on CHROMagar Malassezia growth medium, and the 28S ribosomal DNA (D1/D2) sequence was analyzed for confirmatory identification. Using this database, we detected and analyzed Malassezia species in 68% and 44% of infants with and without infantile seborrheic dermatitis, respectively. The results of MALDI-TOF-MS analysis were consistent with those of rDNA sequencing identification (100% accuracy rate). To our knowledge, this is the first report of a MALDI-TOF-MS database for major skin pathogenic Malassezia species. This system is an easy, rapid and reliable method for identification of Malassezia.

Introduction

The use of matrix-assisted laser desorption and ionization time-of-flight mass spectrometry (MALDI-TOF-MS) in clinical applications related to bacteria and yeast identification has been increasing over the last decade.[1-4] This method allows rapid and reliable identification and classification of microorganisms at low running cost, which helps clinicians to make accurate diagnoses and administration of appropriate treatments in a timely manner.

Malassezia species are lipophilic yeasts that can be found on the skin of humans and other animals; they are normal inhabitants on healthy skin, but have also been reported to be causal or exacerbation factors of certain dermatological diseases, such as pityriasis versicolor, folliculitis and atopic dermatitis.[5-7] In particular, Malassezia restricta is the species most commonly associated with seborrheic dermatitis (SD).[8]

The genus Malassezia is currently classified taxonomically into 14 species based on ribosomal DNA (rDNA) sequences. Combinations of morphological, physiological and biochemical methods, such as colony and cell shapes and sizes, lipid requirements and catalase activity, are also used for species identification, but the process is laborious and the results are sometimes ambiguous.

In this study, we developed a new database identification system for major etiological Malassezia species using MALDI-TOF-MS. We also analyzed the composition of Malassezia microbiota on the skin of infants with or without infantile seborrheic dermatitis (ISD) using this technique.

Methods

Sample collection and preparation

Standard strains

Table 1 lists Malassezia standard strains, including type strains, used to create MALDI-TOF-MS reference spectra. Strains were subcultured on CHROMagar Malassezia (Kanto Chemical, Tokyo, Japan)[9] and modified Leeming and Notman culture media (containing 1% w/v peptone, 1% w/v glucose, 0.2% w/v yeast extract, 0.8% w/v ox bile, 0.05% w/v monostearate, 1% v/v glycerol, 0.5% v/v Tween-60, and 1.5% w/v agar, overlaid with olive oil),[10] and incubated at 32°C until sufficient numbers of colonies for TOF-MS analysis were visible (usually 2–7 days depending on species). Colonies a few millimeters in diameter are usually sufficient for the analysis.

Table 1. Type and standard strains used to obtain MALDI-TOF-MS reference spectra
SpeciesStrain
  1. ATCC, American Type Culture Collection (USA); CBS, The Convention on Biological Diversity (the Netherlands); MALDI-TOF-MS, matrix-assisted laser desorption and ionization time-of-flight mass spectrometry; NBRC, Biological Resource Center, National Institute of Technology and Evaluation (Japan); T, type strain.

Malassezia restricta NBRC103918T (CBS7877, ATCC96801, IFM55992), NN070, MrestSD
Malassezia globosa IFM51946T (CBS7966, ATCC96807), NBRC101597
Malassezia furfur CBS1878T, 068-2, gifu02, PV3-1, PV3-2
Malassezia pachydermatis CBS1879T, 003-1
Malassezia sympodialis CBS7222T
Malassezia obtusa CBS7876T
Malassezia slooffiae CBS7956T, CHR090-2, S-5
Malassezia japonica M9966T

Clinical isolates

Clinical isolates were collected from 50 infants at their regular 1-month checkup at the Pediatric Outpatient Department of Teikyo University Hospital, Tokyo, Japan, between November 2012 and August 2013. Subjects included 25 infants with healthy skin and 25 infants with ISD. The diagnosis of ISD was based on the characteristic red, flaking, greasy appearance of the skin at typically affected sites of the scalp, eyebrows and paranasal areas,[11] and confirmed by an experienced pediatrician. Using 6 cm × 7 cm OpSite transparent dressings (Smith & Nephew Medical Limited, Hull, UK), the forehead, cheek and upper arm were tape-stripped three times each to obtain clinical samples. Each dressing was cut into a strip of 1.5 cm × 3.5 cm, placed on a CHROMagar Malassezia plate, and incubated at 32°C until colonies were visible, which usually took up to approximately 2 weeks. The number of colonies grown from each tape strip was recorded as scant (1–10 colony-forming units [CFU]), moderate (11–100 CFU) or heavy (>100 CFU). Colonies were then subcultured on either CHROMagar Malassezia or modified Leeming and Notman culture media to purify colonies before MALDI-TOF-MS analysis. For each infant, written informed consent was obtained from the parent. The protocol was approved by Teikyo University Review Board.

Creation of MALDI-TOF-MS database for the genus Malassezia

The main spectra (MSP) for eight major Malassezia species were acquired as reference spectra for the MALDI-TOF-MS database using the standard strains listed in Table 1. Protein extraction and creation of new database entries were carried out according to the manufacturer's recommendations (Bruker Daltonics, Kanagawa, Japan). One to three fresh colonies for each standard strain were suspended in 300 μL of distilled water in 1.5-mL microfuge tubes. Then, 900 μL of absolute ethanol (Sigma-Aldrich, St Louis, MO, USA) was added to the tubes and vortexed. After centrifugation at 15 490 g for 2 min, the supernatant was removed and the pellet was allowed to air dry completely. The cell pellet was resuspended in 50 μL of 70% formic acid (Wako, Osaka, Japan), followed by the addition of 50 μL of acetonitrile (Sigma-Aldrich), and vortexed. The suspension was centrifuged at 15 490 g for 2 min, and 1 μL of the supernatant was spotted on a MALDI MSP 96 polished target steel plate (Bruker Daltonics) and allowed to air dry. Then, 1 μL of matrix solution (α-cyano-4-hydroxycinnamic acid; Bruker Daltonics) was applied to cover each sample spot, and allowed to air dry. Escherichia coli protein (Bruker Bacterial Standard) was used as the calibration standard for each set of analysis. Measurements were performed with a Microflex mass spectrometer (Bruker Daltonics) using the flexControl software (version 3.0; Bruker Daltonics). Mass spectra were acquired in linear positive extraction mode ranging 2000–20 000 Da. The spectra were imported into the flexAnalysis software (version 3.0; Bruker Daltonics) for quality check, selected for MSP creation, and added to the existing Biotyper database.

MALDI-TOF-MS identification of clinical isolates

For the clinically isolated Malassezia strains, MALDI-TOF-MS identification was carried out using fresh colonies obtained after 48–72 h of incubation. Incubation time was extended as necessary to obtain colonies of sufficient size for analysis. Samples were prepared for TOF-MS analysis with the same formic acid/acetonitrile extraction procedure as described above for standard strains. The spectra generated for clinically isolated strains were imported into the Biotyper software (version 3.0; Bruker Daltonics), and compared to the reference spectra of the newly created database. The degree of spectral concordance was expressed as a logarithmic identification score (LogScore) ranging 0–3; according to the manufacturer's instructions, a score of 2.0 or higher indicates species identification, 1.7 or more indicates identification of a genus, and less than 1.7 indicates no reliable identification.

DNA sequencing of the 28S rDNA (D1/D2) region

The identity of each Malassezia strain was confirmed by DNA sequencing. DNA was extracted by phenol-chloroform extraction[12] and stored at −30°C until analysis. The 28S region of rDNA was amplified by polymerase chain reaction (PCR) from a 1:50 dilution of template DNA in 10 × PCR buffer (Takara, Shiga, Japan), 0.025 U of Ex Taq Polymerase (Takara) per μL, 200 μmol/L of each deoxyribonucleotide triphosphate (Takara), forward primer 28SF1 (5′-AAGCATATCAATAAGCGGAGG-3′),[13] and reverse primer 635 (5′-GGTCCGTGTTTCAAGACGG-3′)[13] in sterile water. The PCR Thermal Cycler (Takara) parameters were as follows: initial denaturation at 94°C for 4 min, followed by 30 cycles at 94°C for 1 min, 55°C for 2 min, and 72°C for 1 min and 30 s, with a final extension at 72°C for 10 min. The amplification products were purified using a QIAquick PCR Purification Kit (Qiagen, Hilden, Germany) and sequenced using a BigDye Terminator Sequencing Kit (Applied Biosystems, Warrington, UK) and an ABI PRISM 310 genetic analyzer (Applied Biosystems) according to the respective manufacturer's instructions. The sequences obtained were then analyzed by BLAST using the NCBI databases (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

Results

Standard species-specific MALDI-TOF-MS spectra were acquired for eight Malassezia species as shown in Figure 1, and a new database for the genus Malassezia was created from these spectra. A total of 54 clinical isolates were analyzed by MALDI-TOF-MS and classified according to the newly created database. The authenticity of the identification results was assessed by rDNA sequencing method. Table 2 shows the number of clinical isolates in each score value. By MALDI-TOF-MS, 15 strains were identified as M. restricta, six as Malassezia globosa, four as Malassezia furfur, three as Malassezia pachydermatis, 18 as Malassezia sympodialis, one as Malassezia obtusa and seven as Malassezia slooffiae. These results were in good agreement with those of 28S rDNA sequence analysis (100% accuracy rate). There was no case of false identification, although some LogScores were short of the manufacturer's recommendation for species identification. The values were all above 1.6 when valid peaks were obtained. Malassezia japonica was not isolated from the clinical samples in this study. Other medically important yeasts, such as Candida albicans, Candida glabrata, Candida tropicalis, Candida parapsilosis, Trichosporon asahii and Cryptococcus neoformans, can be properly identified by MALDI-TOF-MS,[1, 3] and their identification quality were not influenced by the addition of a new Malassezia database (data not shown).

Table 2. Numbers of clinical isolates in each identification score value
MALDI-TOF-MS identification resultNo. of strainsTOF-MS LogScore valuesDNA sequencing result
<1.7≥1.7, <2.0≥2.0
  1. According to the manufacturer's instructions, a score of ≥2.0 indicates species identification, ≥1.7 indicates genus identification and <1.7 indicates no reliable identification. In this study, MALDI-TOF-MS identification results were concordant with DNA sequencing results when scores were >1.6. MALDI-TOF-MS, matrix-assisted laser desorption and ionization time-of-flight mass spectrometry.

Malassezia restricta 15069 M. restricta
Malassezia globosa 6024 M. globosa
Malassezia furfur 4004 M. furfur
Malassezia pachydermatis 3021 M. pachydermatis
Malassezia sympodialis 182133 M. sympodialis
Malassezia obtusa 1010 M. obtusa
Malassezia slooffiae 7025 M. slooffiae
Total5422626 
Figure 1.

Representative matrix-assisted laser desorption and ionization time-of-flight mass spectrometry spectra of eight standard Malassezia species. Strains used for the spectra shown were NBRC103918 (M. restricta), IFM51946 (M. globosa), CBS1878 (M. furfur), CBS1879 (M. pachydermatis), CBS7222 (M. sympodialis), CBS7876 (M. obtusa), CBS7956 (M. slooffiae) and M9966 (M. japonica).

Malassezia detection rates from each sample group and each sampling site are shown in Table 3. Malassezia species were detected in 44% of the healthy-looking skin (HS) group and 68% in the ISD group. The quantity and diversity of Malassezia species were greater in the ISD group than the HS group. There were cases in which more than one species was isolated from each sample strip. Malassezia were isolated from forehead and cheek in about one-third of the subjects in the HS group, and nearly half of the subjects in the ISD group. In addition, the detection rate of Malassezia from the upper arm was considerably higher in the ISD group than the HS group (24% vs 4%, respectively; Fisher's least significant difference, < 0.01). Malassezia clinical isolates detected from 1-month-old infants with and without ISD are tabulated in Table 4. The species identification shown here was based on MALDI-TOF-MS results. Strains with LogScores of less than 2.0 were also included with their identification verification from DNA sequencing. In both groups, the two most commonly detected species were M. sympodialis (20% and 40% in HS and ISD groups, respectively) and M. restricta (20% and 28% in HS and ISD groups, respectively).

Table 3. Number of cases in each quantity level of Malassezia grown on CHROMagar Malassezia growth medium
Skin groupSampling siteScant*ModerateHeavyTotal (%)
  1. *1–10 CFU; 11–100 CFU; >100 CFU. CFU, colony-forming units; HS, healthy-looking skin group; ISD, infantile seborrheic dermatitis group.

HS (= 25)Forehead8008 (32)
Cheek8008 (32)
Upper arm1001 (4)
Any site101011 (44)
ISD (= 25)Forehead92112 (48)
Cheek101011 (44)
Upper arm6006 (24)
Any site133117 (68)
Table 4. Malassezia clinical isolates detected from 1-month-old infants with healthy-looking skin (HS) and with infantile seborrheic dermatitis (ISD)
Species detectedForehead (%)Cheek (%)Upper arm (%)Any site (%)
  1. –, not detected.

HS: = 25
 Malassezia restricta 5 (20)2 (8)5 (20)
 Malassezia globosa 1 (4)1 (4)
 Malassezia furfur 1 (4)1 (4)2 (8)
 Malassezia pachydermatis
 Malassezia sympodialis 2 (8)4 (16)5 (20)
 Malassezia obtusa
 Malassezia slooffiae 2 (8)1 (4)1 (4)3 (12)
 Malassezia japonica
ISD: = 25
 M. restricta 4 (16)1 (4)2 (8)7 (28)
 M. globosa 1 (4)3 (12)1 (4)5 (20)
 M. furfur 2 (8)2 (8)
 M. pachydermatis 1 (4)1 (4)1 (4)1 (4)
 M. sympodialis 3 (12)7 (28)2 (8)10 (40)
 M. obtusa 1 (4)1 (4)
 M. slooffiae 2 (8)1 (4)1 (4)4 (16)
 M. japonica

Modified Leeming and Notman culture medium was also used for subculture in this study; differences in growth media from the database may have resulted in somewhat lower LogScores, but did not affect the identification results of MALDI-TOF-MS (data not shown).

Discussion

MALDI-TOF-MS identification of the genus Malassezia

Matrix-assisted laser desorption and ionization time-of-flight mass spectrometry has proven to be useful in identifying many clinically important microorganisms,[14-20] but it has not been possible to identify Malassezia species because of the lack of a database for this organism. The genus Malassezia has been classified into 14 species to date, and only two species (M. furfur and M. pachydermatis) and one strain from each species have been registered in the commercial database provided by Bruker Daltonics (http://www.bruker.jp/daltonics/). In the present study, we created a MALDI-TOF-MS database of two dermatologically important Malassezia species, M. restricta and M. globosa,[6-8] and six other Malassezia species, namely, M. furfur, M. pachydermatis, M. sympodialis, M. obtusa, M. slooffiae and M. japonica, as reference species; these species were chosen because they were isolated from human skin or associated with Malassezia-related diseases in humans more frequently than other Malassezia species.[5]

All Malassezia clinical isolates were successfully identified by MALDI-TOF-MS with the newly created database in this study, but with varying LogScore values (Table 2). The low LogScores in some strains may have been attributable to diversity within each species. Phylogenetic analysis with DNA sequencing of internal transcribed spacer 1 of ribosomal DNA has demonstrated this intraspecies diversity of Malassezia.[12] Two strains with scores lower than genus identification level (<1.7) still gave valid identification results as “M. sympodialis”. There was only one standard strain of this species used for the reference database in this study. Increasing the number of reference spectra for each species should improve the effectiveness of MALDI-TOF-MS identification as well as the LogScores. In addition, as we used default settings in the tests, improvements are expected by adjustment of some TOF-MS parameters. Furthermore, as scores above 1.6 consistently gave accurate identification to the species level in our analyses, lowering the cut-off value to 1.6 or more for species identification seemed acceptable for the genus Malassezia. The running costs and the time required for analysis per sample of our method are better than those of conventional and other molecular methods.[21]

Malassezia microbiota of 1-month-old infant with and without ISD

Malassezia species were detected in 44% of subjects in the HS group and in 68% of those in the ISD group (Table 3). Malassezia species comprise the majority of skin fungal microbiota in healthy individuals.[22] M. sympodialis is usually the most commonly detected species in culture-based analyses,[23] as was also the case in the present study. However, it has been reported that M. restricta and M. globosa are major components of normal Malassezia microbiota when Malassezia genes are detected directly in cutaneous samples by real-time PCR assay.[24] There are considerable differences in growth rate and requirements for growth conditions among species, which contribute to these discrepancies. Therefore, non-culture methods more accurately represent the distribution of each species, but the process is much more laborious than culture methods, and the isolates are not available for further analyses. In the present study, we used a culture-based method due to its technical ease and also to obtain samples for TOF-MS analysis. Although the instrument is still expensive, TOF-MS has already been introduced in clinical microbiology laboratories around Europe, and is becoming an essential technology in culture-based microbial identification.[25]

There have been only a limited number of reports on cutaneous Malassezia microbiota of young populations;[26-28] most of these were culture-based analyses, and geographical differences have also been suggested. The consensus seems to be that skin colonization by Malassezia begins immediately after birth, and neonates acquire their fungal microbiota from their mothers or caregivers. In a recent non-culture-based investigation in Japan, Malassezia DNA was detected from 100% of neonate samples on day 1 after birth, and the fungal microbiota of neonates shifted to the adult type by day 30.[29] Although the detection rate was inevitably reduced in our culture-based study, the diversity of species seemed to be well recovered especially in the ISD group with CHROMagar Malassezia growth medium.

Malassezia restricta has been reported to be the major microbial factor associated with SD in adult patients, and interspecies genetic differences between healthy subject and patients with SD have also been described.[8] In the present study, the isolation rate of M. restricta was slightly increased in the ISD group compared to the HS group (28% vs 20%, respectively; Table 4); as there were also increases in detection rate of other species, we could not determine the relationship between this species and ISD from this rate alone. One noticeable difference between the two groups was that Malassezia detection rate on the upper arm was considerably higher in the ISD group than the HS group (24% vs 4%, respectively; Table 3). This suggested that even the non-lesional skin and non-seborrheic parts of the body were more seborrheic in infants with ISD.

Malassezia species have been reported as etiological agents of systemic infections in some cases, especially in neonates treated by lipid-containing infusion.[30] It was also assumed that the source of the strains causing fungemia in children was the skin of parents or health-care workers. For example, M. pachydermatis is known to inhabit animal skin, and its transmission through the hands of dog owners has been reported to have caused infection and colonization in a neonatal intensive care unit.[31] In the present study, M. pachydermatis was detected in all three parts of the body investigated in one infant; the family was presumed to have owned a companion animal, although we could not confirm this at the time of sample collection. With its ability to provide hierarchical cluster analysis, MALDI-TOF-MS should be a suitable tool for this type of epidemiological analysis.

In conclusion, we developed a rapid and easy method to detect pathogens of cutaneous Malassezia-related disease. With MALDI-TOF-MS, Malassezia species could be identified with simple preparation steps and in a short time once the colonies had grown on the culture medium. This system provided a species identification system that is rapid, accurate and applicable even to strains with atypical morphological features. In addition, the accuracy of identification would be further improved by upgrading using a well-authenticated database.

Acknowledgments:

Some of the Malassezia standard strains were provided by the National BioResource Project in Japan (http://www.bnrp.jp/), Medical Mycology Research Center, Chiba University (http://www.pf.chiba-u.ac.jp/) and Professor Takashi Sugita, Meiji Pharmaceutical University. The authors thank Mr Yasuhiko Maekawa, Ms Yumiko Matsuyama, Mr Jutaro Kawaguchi (Bruker Daltonics), and Professor Akira Kikuchi, Dr Eishin Ogawa (Department of Pediatrics, Teikyo University School of Medicine), Dr Goto Kazuo, Mr Mitsuru Matsumura (Department of Clinical Laboratory Medicine, School of Medical Technology, Teikyo University), Ms Sayoko Kawakami, Ms Miwa Asahara and Ms Shinobu Ishigaki (Department of Central Clinical Laboratory, Teikyo University Hospital) for their technical support. This study was supported in part by Health and Labor Science Research Grants (Research on Emerging and Re-emerging Infectious Diseases) H25-shinko-ippan-006, from the Ministry of Health, Labor and Welfare of Japan (K. M.).

Conflict of interest

KM has received research grants or lecture fees from the following companies: Maruho, Pola Pharma, MSD, Janssen Pharmaceutical, Kanto Chemical, Pfizer, Eiken Chemical, Nhon Nohyaku, Becton Dickinson, Dai-Nippon Sumitomo Pharmaceutical, Galderma and Japan Space Forum. The authors alone are responsible for the content and writing of the paper and declare no conflicts of interest.

Ancillary