Comparative Analysis of Cutaneous Fungi in Atopic Dermatitis Patients and Healthy Individuals

Abstract

Background

Atopic dermatitis (AD) is a chronic relapsing inflammatory skin disease triggered by diverse factors. Microbes are one of the crucial risk factors for AD development or exacerbation. However, the effect of a fungal burden on AD has been overlooked compared to bacteria.

Objective

This study aimed to comparatively analyze cutaneous fungal distribution between AD patients and healthy individuals by polymerase chain reaction (PCR)-based analysis.

Methods

Skin samples of AD outpatients and healthy individuals collected at the Chung-Ang University were analyzed. Representative AD-associated fungal genera, Candida, dermatophytes, and Malassezia , were analyzed using specific primer and amplification methods. Amplicons were sequenced, and the fungal distribution of both groups were compared.

Results

Totally, 211 patients and 23 healthy individuals were studied. Of the 211 patients, 10.90% (23/211) had Candida species, whereas 0% (0/23) healthy individuals showed its presence. The most frequently detected species in patients was Candida albicans (5.21%) followed by Candida parapsilosis (3.79%). For dermatophytes, 1.42% (3/211) of patients showed positive results, whereas 0% (0/23) healthy individuals showed positive results. Malassezia species were identified in 20.85% (44/211) and 8.70% (2/23) in patients and healthy individuals, respectively. Malassezia restricta was the most frequently identified species in the AD patient group, and the only species found in the healthy control group.

Conclusion

The distribution of Candida spp., dermatophytes, and Malassezia spp. are altered with AD development.

INTRODUCTION

Atopic dermatitis (AD) is a chronic skin disease, which affects up to 20% of children and 2% of adults and the prevalence has increased 2 to 3 folds during the last 30 years¹,². AD is a representative complex disease because diverse factors can lead to AD, including genetic, environmental factors, and microorganisms³,⁴. Recent research on the effects of microbes on the skin and the underlying mechanisms has revealed that while the skin is a beneficial niche for a variety of microorganisms that exhibit mutualism or commensalism, the microbes could become pathogenic depending on the host immune condition. In the case of AD, the weakened skin barrier of patients is more vulnerable to microbial penetration. For example, it is known that Staphylococcus aureus, a representative AD-associated bacterium known to exacerbate the disease by producing superantigens, colonizes 200 folds more in the skin of patients with AD than healthy individuals⁵,⁶. Concerning AD, comprehensive research on cutaneous fungal microbiota is few compared to that of bacterial microbiota because of lower interest and limited methodological approach. Although sufficient research has not been constructed yet, many suggestions are emphasizing the fungal burden on disease development. For example, AD patients who responded poorly to conventional treatment showed improvement when treated with antifungal drugs⁴,⁷. Moreover, it is known that the cutaneous fungal diversity and richness is altered in AD patients compared with healthy individuals⁸,⁹. Although the fungus is not a direct causative agent, dysbiosis of the fungal community appears to contribute to AD onset or exacerbation. This makes it important to study the association between the fungal community and AD, which can provide a novel perspective on understanding microbes as contributor to AD and identifying appropriate solution to a disease.

In this study, we aimed to comparatively analyze the cutaneous distribution of Candida, dermatophytes, and Malassezia between the lesional skin of AD patients and healthy individuals. Since culture-based analysis does not accurately reflect fungal community due to cultural limitations, we conducted a polymerase chain reaction (PCR)-based analysis.

MATERIALS AND METHODS

Subjects and skin sample collection

Skin samples were collected from 211 outpatients with AD and 23 healthy controls between January 2011 and February 2013 at Chung-Ang University Hospital (Seoul, Republic of Korea). A total of 234 skin swab samples were used in this study. The inclusion criteria for AD patients were as follows: 1) patients diagnosed with AD by a dermatologist in Chung-Ang University Hospital; 2) patients who had AD for more than 10 years; 3) patients who had typical AD skin lesion on the antecubital fossa. The inclusion criteria for healthy control was subjects without AD or any other active skin diseases confirmed by thorough history taking and physical examination by a dermatologist in Chung-Ang University Hospital. For both groups, exclusion criteria were as follows: 1) participants who had any active skin diseases, exclusive of AD; 2) participants with a history of receiving oral or topical antifungal agents (used within 4 weeks of study enrolment); 3) participants a history of serious cardiovascular, respiratory, endocrine system, and central nervous system diseases, or who have been diagnosed with a mental illness that can significantly affect this clinical trial; 4) participants who are judged unsuitable for participation in clinical trials by other investigators. The study protocol was approved by the Institutional Review Board (IRB) of Chung-Ang University Hospital (C2010003, 298), and informed consent was obtained according to the IRB’s policy. Because of the lack of previous data, the sample size was determined based on feasibility considerations rather than power analysis. A total of 234 skin swab samples were collected from the antecubital area from both groups using the skin swab method. The detailed swab method was as follows: The cotton swabs premoistened in sterilized saline were repeated rubbed at least 10 times against the lesional sites of the antecubital fossae of AD patients and healthy individuals. The cotton swabs used in skin swab were stored with 500 µl of lysis buffer (100 mM NaCl, 10 mM Tris-HCl [pH 8.0], 1 mM EDTA [pH 8.0], 1% sodium dodecyl sulfate, 2% Triton X-100). Subsequently, the DNAs are extracted and used for amplification.

Real-time qPCR screening using pan-fungal primers

Real-time qPCR screening was performed for comparative analysis of fungal distribution between AD patients and healthy individuals. To perform amplification, we used pan-fungal primers (Table 1)¹⁰,¹¹,¹², which are universally selected for targeting the fungal ITS1 region. Real-time qPCR was conducted with iTaq Universal SYBR Green PCR master mix (Bio-Rad, Hercules, CA, USA) according to the manufacturer’s instructions on a CFX96 real-time PCR detection system. The amplicons were purified using a PCR purification kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions. The purified amplicons were sequenced using BigDye^® Terminator V3.1 cycle sequencing kit (Applied Biosystems, Foster City, CA, USA) with forward and reverse primers initially used in PCR amplification. To ascertain species identity, the characterized sequence was analyzed using BLASTn to align sequence similarity with intended target genes in NCBI (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

Table 1. Primers used in the real-time qPCR assay.

Name	Sequence (5’-3’)	Characterization
ITS1 F	TGC GTT CTT CAT CGA TGC GA	Pan-fungal primer (Forward)
ITS4 R	TAA GCG CAA GTC ATC AGC TTG CGT	Pan-fungal primer (Reverse)
ITS1 F	TAA GCG CAA GTC ATC AGC TTG CGT T	Pan-Candida primer (Forward)
5.8S1R	TGC GTT CTT CAT CGA TGC GA	Pan-Candida primer (Reverse)
ITS1 F2	SSC CCC ATT CTT GTC TAC MTY AC	Trichophyton detection (Forward)
ITS R2	AAC GCT CAG ACT GAC AGC TCT TC	Trichophyton detection (Reverse)
MC F	CCT AAG CGG TGG GTG GTT ACT	Microsporum detection (Forward)
MC R	TGA AAG AAC ATA CCG TCT GAG CG	Microsporum detection (Reverse)

The primer sequences shown in this table were previously published¹⁰,¹¹,¹². ITS: internal transcribed spacer, MC F: Microsporum canis forward, MC R: M. canis reverse.

Real-time qPCR screening using genus-specific primers

We carried out further analysis with genus-specific primers as compensation for the possibly poor sensitivity. Pan-candida primer pair, designed for amplifying a variable segment of 28S ribosomal gene of Candida spp. was used for Candida spp. detection. For dermatophytes detection, a genus-specific primer for Trichophyton and Microsporum was designed, which are listed in Table 1. The amplicons were analyzed in the same way as in pan-fungal primer amplification.

Nested PCR and direct sequencing for Malassezia spp. detection

Cutaneous Malassezia communities were comparatively analyzed between AD patients and healthy individuals by conventional nested PCR due to the low concentration of DNA. In the first-round amplification, the fungal ITS1 region was amplified by using a universal primer pair (Table 2)¹³, and 1 µl of the first-round PCR product was added as template DNA in the second-round amplification. We used 7 Malassezia species-specific primers in the second-round amplification. After amplification, the PCR product was identified on 2% agarose gel, following purification using a PCR purification kit (Qiagen). Purified PCR products were sequenced using BigDye^® Terminator V3.1 cycle sequencing kit (Applied Biosystems) with primers used in second-round PCR amplification. For more accurate results, the characterized sequence was analyzed using BLASTn to align sequence similarity with intended target genes in NCBI (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

Table 2. Primers used in the nested PCR assay.

Name		Sequence (5’-3’)	Characterization
First-round amplification
	ITS1	TCC GTA GGT GAA CCT GCG G	ITS target primer (Forward)
	ITS4	TCC TCC GCT TAT TGA TAT GC	ITS target primer (Reverse)
Second-round amplification
	M. restricta F	CTT GGT TGG ACC GTC ACT GG	Malassezia restricta detection (Forward)
	M. restricta R	AGG CGG ATG CAA AGT GTC TC	M. restricta detection (Reverse)
	M. globosa F	CAA TAA GTG TGT CTC TGC GG	Malassezia globosa detection (Forward)
	5.8S R	TTC GCT GCG TTC TTC ATC GA	M. globosa detection (Reverse)
	M. sympodialis F	CGG ACG CAA ACA CGT CTC TG	Malassezia sympodialis detection (Forward)
	5.8S R	TTC GCT GCG TTC TTC ATC GA	M. sympodialis detection (Reverse)
	M. furfur F	CTA CTC GCG TAC AAC GTC TCT G	Malassezia furfur detection (Forward)
	5.8S R	TTC GCT GCG TTC TTC ATC GA	M. furfur detection (Reverse)
	M. pachydermatis F	CTG CCA TAC GGA TGC GCA AG	Malassezia pachydermatis detection (Forward)
	5.8S R	TTC GCT GCG TTC TTC ATC GA	M. pachydermatis detection (Reverse)
	M. obtusa F	ACC CGT GTG CAC ACT GTT GAG	Malassezia obtusa detection (Forward)
	5.8S R	TTC GCT GCG TTC TTC ATC GA	M. obtusa detection (Reverse)
	M. slooffiae F	ACG CAC GCT AAC ACA ACG TG	Malassezia slooffiae detection (Forward)
	5.8S R	TTC GCT GCG TTC TTC ATC GA	M. slooffiae detection (Reverse)

The primer sequences shown in this table were previously published¹³.

Statistical analysis

All samples from all participants were used in statistical analysis. The chi-square test was used for statistical analysis, and the significance levels were set at p<0.05. Datasets were entered into Microsoft Excel 2016 (Microsoft, Redmond, WA, USA) and analyzed using R version 3.6.1 (R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Distribution of Candida

The abundance and diversity of Candida spp. in the skin were higher in AD patients (10.90%; 23/211) than healthy controls (0%) (Fig. 1). We also confirmed that the pan-fungal primer was much more sensitive than the pan-candida primer for Candida detection (data not shown). The most frequently detected species in the AD group was Candida albicans (5.21%; 11/211), a representative clinically isolated species. Candida parapsilosis was second-most frequent among AD patients (3.79%; 8/211). Four more samples from AD patients tested positive for Candida, but we could not characterize the exact species because of the low homology score despite further sequencing analysis including of the D1/D2 domain (Table 3).

Fig. 1. Detection frequency of Candida spp., dermatophytes, and Malassezia spp. in healthy controls and patients with atopic dermatitis (AD). Candida spp. were detected in 10.90% AD samples and 0% samples from healthy controls. Dermatophytes were detected in 1.42% AD samples and 0% samples from healthy controls. Malassezia spp. were detected most frequently in 20.85% AD samples and 8.70% samples from healthy controls.

Table 3. Distribution of Candida, dermatophytes, and Malassezia in atopic dermatitis patients and healthy controls.

Distribution	Species	Atopic dermatitis (n=211)	Healthy control (n=23)
Candida spp.	Total	23 (10.90)	-
	Candida albicans	11 (5.21)	-
	Candida parapsilosis	8 (3.79)	-
	Candida spp.	4 (1.90)	-
Dermatophytes	Total	3 (1.42)	-
	Trichophyton rubrum	2 (0.95)	-
	Microsporum canis	1 (0.47)	-
Malassezia spp.	Total	44 (20.85)	2 (8.70)
	Malassezia restricta	32 (15.17)	2 (8.70)
	Malassezia globosa	8 (3.79)	-
	Malassezia furfur	3 (1.42)	-
	Malassezia sympodialis	1 (0.47)	-
	Malassezia pachydermatis	-	-
	Malassezia obtusa	-	-
	Malassezia slooffiae	-	-

Values are presented as number (%). Cutaneous distribution of Candida, dematophytes, Malassezia from atopic dermatitis patients and healthy individuals is shown in this table. Cutaneous fungal distribution between two groups were not statistically significant (Candida spp.: not applicable, dermatophyte: not applicable, Malassezia spp.: p=0.163575).

Distribution of dermatophytes

Three samples from AD patients tested positive for dermatophytes, whereas none in the group of healthy individuals tested positive (Table 3). Unlike for Candida, the genus-specific primer was more effective to detect dermatophytes than a pan-fungal primer. In the AD patient group, Trichophyton rubrum was detected in two samples, and Microsporum canis was detected in one (Table 3).

Distribution of Malassezia

We comparatively analyzed 7 Malassezia species between AD patients and healthy controls. Overall, 20.85% (44/211) AD patients were identified to have Malassezia spp. In the case of healthy controls, only 8.70% (2/23) had Malassezia (Fig. 1). Both the species abundance and diversity of Malassezia were higher in AD patients than in healthy individuals, but this was not statistically significant (p=0.163575). Malassezia restricta was the most frequently detected species in AD patients (15.17%; 32/211) and simultaneously the only found species in healthy controls (Table 3). In AD patients, Malassezia globosa was detected second-most frequently (3.79%; 8/211), followed by Malassezia furfur and Malassezia sympodialis (1.42% and 0.47%, respectively) (Table 3). Malassezia pachydermatis, Malassezia obtusa, or Malassezia slooffiae was not identified in both groups.

DISCUSSION

Recent research has focused on the role of microbes in AD pathogenesis. While bacteria have been studied extensively, the effects of fungi have been overlooked⁴. Fungi are not a direct causative agent of AD because they are present in the normal skin of healthy individuals, but they appear to contribute to the onset and exacerbation of AD¹⁴,¹⁵,¹⁶. A change in the composition of the fungal community in lesional or non-lesional skin of AD patients has been demonstrated in local and global trials⁸,⁹,¹⁷. This change could be because of the altered physiological skin environment as AD progresses. For example, AD patients showed epidermal barrier dysfunction resulting from filaggrin mutation¹⁸,¹⁹. The weakened skin barrier of AD patients is more sensitive to microbial infections and allergenic reactions. The intact healthy skin produces antimicrobial peptides such as defensins and maintains pH weakly acidic thereby protecting it from external environments³,²⁰. However, the skin of AD patients is deficient in this respect, resulting in alteration of skin microbial flora²¹,²².

We comparatively analyzed three representative AD-associated fungi between AD patients and healthy individuals by conducting a PCR-based analysis. Candida, a commensal yeast that mainly colonizes mucosal surfaces, is thought to worsen AD via the secretion of allergens and antigens²³. Candida has been cultured more frequently from the gastrointestinal tract of AD patients than healthy individuals, but there is little information about its skin colonization²³,³⁴. In the present study, we confirmed that the skin distribution of Candida spp. is higher in AD patients than in healthy individuals. As a representative resident flora, Candida does not directly trigger AD progression, but impaired immune systems of AD patients might be susceptible to Candida infections. Dermatophytes are also known to contribute to AD; however, the association is still largely unknown. Jones et al.²⁵ found that chronic dermatophyte infections are more common in AD patients, and cases of chronic dermatophyte infection are more severe and difficult to treat. Although AD patients with dermatophyte infections may show improvement when treated with antifungal drugs, a more detailed link between dermatophytes infection and AD progression remains to be established²⁶. Malassezia spp. are the main eukaryotes composing the microbial flora in normal human skin²⁷. In healthy individuals, colonization of Malassezia spp. accounts for 53% to 80% of all fungi at different locations²⁸,²⁹. Malassezia is also associated with several human skin diseases and disorders especially, Malassezia restricta is a predominant species in patients with seborrheic dermatitis and AD³⁰,³¹. M. restricta and M. globosa are thought to play a major role in AD development because these two Malassezia species are abundantly present in almost all AD patients, thus supporting our findings. The prevalence of Malassezia is commonly studied in sebum-rich areas such as the scalp, chest, and back because of their lipophilic nature¹⁵,³². However, in this study, all swab samples were collected from the antecubital area because antecubital fossa is the most common site of AD and an easy area to collect samples. In addition, skin microbiome shows phylogenic diversity according to body parts. Therefore, we conducted the study by limiting the sample collection area to the antecubital area in both the patient group and the control group.

A hallmark of human skin microbiota communities is high diversity and high interpersonal variation. However, the skin microbiota of a healthy adult remains stable over time, despite environmental perturbations. Adults stably maintain the composition of their skin microbial communities as assessed for at least 2 years³³. So, considering that the participants in this study were all adults, it is believed that the intra-individual fungal microbiota was not significantly affected by the environment. Costello et al.³⁴ found that skin sites including the palms, fingers, and forearm had greater phylogenetic diversity than communities in the gut, external auditory canal, or oral cavity. Therefore, we conducted the study by limiting the sample collection area to the antecubital area in both the patient group and the control group.

There are some limitations to be considered when interpreting our findings. The number of specimens from healthy controls was relatively smaller than that from AD patients. In addition, the lack of clinical information including gender, age, and disease severity made it impossible to analyze the results reflecting it. Despite these limitations, our study has two strengths. First, our analysis included 234 many clinical samples. Second, we designed diverse primers and amplification methods according to analysis purposes.

In conclusion, we compared the distribution of Candida, dermatophytes, and Malassezia in the skin of AD patients and healthy individuals. The distribution of fungi in all three genera was altered in AD patients. In our study, both the species abundance and diversity of Malassezia were higher in AD patients than in healthy individuals. It is speculated that the impaired skin barrier in AD allows colonization with more different Malassezia than healthy skin. Vice versa, the altered mycobiome may cause activation of the skin immune system leading to inflammation and eczema. Considering the composition of normal skin microflora, fungi are not a direct causative agent of AD, but an imbalance in their composition appears to be associated with AD development or exacerbation. Further studies focusing on fungal immune response or allergenic mechanisms will be required to better understand the role of fungi in AD progression.

DATA SHARING STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.