doi: 10.1111/1346-8138.

14952 Journal of Dermatology 2019;

: 1–4

CONCISE COMMUNICATION
Skin microbiome differences relate to the grade of acne
vulgaris
Chun-xi LI,1 Zhi-xuan YOU,2 Yan-xia LIN,1 Hai-yue LIU,3,* Jin SU1,*
1
Department of Respiratory and Critical Care Medicine, Chronic Airways Diseases Laboratory, Nanfang Hospital, Southern Medical
University, 2Zhixin High School, 3Division of Laboratory Medicine, Zhujiang Hospital, Southern Medical University, Guangzhou, China

ABSTRACT
The skin microbiome plays important roles in the pathogenesis and development of acne. We aimed to investigate the
facial skin microbiome of acne and microbiome differences related to different grades of acne. Skin swabs from nine
healthy controls and 67 acne patients were collected, and the skin microbiomes were analyzed using 16S rRNA gene
sequencing. Compared with healthy controls, acne patients harbored significantly altered skin microbiomes. The skin
microbiomes of patients with grade 1–3 acne were similar, but patients with grade 4 acne showed a significantly differ-
ent skin microbiome compared with grade 1–3 acne, including increased alpha diversity and increased proportions of
four Gram-negative bacteria (Faecalibacterium, Klebsiella, Odoribacter and Bacteroides). In conclusion, acne patients
harbored an altered skin microbiome, and more significant dysbiosis was found in patients with grade 4 acne (severe
acne). Our findings may provide evidence for the pathogenic mechanisms of acne and microbial-based strategies to
avoid and treat acne, especially grade 4 acne.

Key words: 16S rRNA, acne grading, acne vulgaris, Gram-negative, skin microbiome.

INTRODUCTION
The human skin is colonized by a wide variety of microbes such
as bacteria, fungi and viruses. The bacteria on the skin can be
divided into resident and temporary bacteria; traditional culture-
based methods show that Cutibacterium and Staphylococcus
are the most common resident skin bacteria, while Gram-nega-
tive bacteria are regarded as temporary bacteria with limited dis-
tribution on the skin.1,2 Culture-independent technologies have
revealed close relationships between the skin microbiome and
chronic inflammatory dermatoses, such as atopic dermatitis,
psoriasis and acne vulgaris (acne).3 16S rRNA gene sequencing
showed that the skin microbiome is closely correlated with acne,
and the dominant bacterial communities on the skin of acne
patients were Cutibacterium, Staphylococcus, Corynebacterium,
Streptococcus and Micrococcus.4
Acne is a chronic inflammatory skin disease, which affects
approximately 80% of adolescents and young adults.5 Multiple fac-
tors, including inflammation, excess sebum, hormonal stimulation
and microbial overgrowth, especially overgrowth of Cutibacterium
acnes, are considered to be the main pathogenic agents of acne.5
Disease severity, which is dependent on the lesion type, quantity
and size, is an important clinical feature for acne treatment and the
evaluation of its curative effect, and the acne grading system is a
standard used for the evaluation of acne severity.6 However, very
few studies have investigated the skin microbiome differences
among patients with different grades of acne.
CASE REPORT
Sixty-seven outpatients with facial acne and nine healthy con-
trols with no skin problems were recruited from Guangzhou,
China, from July to September 2018. Table 1 shows the clinical
characteristics of study subjects.
The exclusion criteria were as follows: individuals with
another dermatosis, such as atopic dermatitis, eczema, psoria-
sis, ichthyosis or cutaneous fungal infection, or who had used
antibiotics, hormones, retinoic drugs or benzoyl peroxide within
1 month prior to recruitment. All subjects gave written informed
consent in accordance with the Declaration of Helsinki. This
study was approved by the ethics committee of Nanfang
Hospital, Southern Medical University.
The acne grading was determined by the nature of the lesion
on the face according to the 2016 European S3 Acne Guideline:
(1) comedonal acne; (2) mild–moderate papulopustular acne; (3)

Correspondence: Jin Su, Ph.D., Department of Respiratory and Critical Care Medicine, Chronic Airways Diseases Laboratory, Nanfang Hospital,
Southern Medical University, 1838 North Guangzhou Avenue, Baiyun District, Guangzhou 510515, Guangdong Province, China. Email:
drsujin@126.com and Hai-yue Liu, Ph.D., Division of Laboratory Medicine, Zhujiang Hospital, Southern Medical University, 253 Industrial Ave-
nue Center, Haizhu District, Guangzhou 510282, Guangdong Province, China. Email: 457906998@qq.com
*These authors contributed equally to this study.
Received 19 February 2019; accepted 15 May 2019.

© 2019 Japanese Dermatological Association 1

C.-x. Li et al.
Table 1. Clinical characteristics of the subjects
Acne vulgaris (n = 67)
Healthy (n = 9) Grade 1 (n = 34)
Sex, n (male/female) 6/3 15/19
Age (years) 23.7  1.12 23.3  1.09
Duration (years) 0.89  2.66 2.68  3.64
The data are presented as n (v2-test) or as the mean  standard deviation.
†
P < 0.01.
severe papulopustular acne or moderate nodular acne; and (4)
severe nodular acne or conglobate acne.6
Samples were obtained by the swab method (the skin surface
of cheeks of each subject was rubbed with a sterile swab for at
least 30 s) using 0.9% saline on the facial skin of subjects who
had not used any skin creams or cosmetics and were immediately
stored at 80°C for subsequent DNA extraction.
DNA extraction and 16S rRNA gene sequencing
DNA was extracted from skin samples using the genomic DNA
Nucleic Acid Extraction Kit (Bioeasy Technology, Shenshen,
China). The V4 region of the 16S rRNA gene was amplified using
the upstream primer forward 50 -GTGTGCCAGCMGCCGCGG-
TAA-30 and the downstream primer reverse 50 -CCGGAC-
TACHVGGTWTCTAAT-30 and then sequenced using an Illumina
HiSeq2000 Sequencer (Illumina, San Diego, CA, USA).7 Quantita-
tive Observation of Microbial Ecology (version 1.9.1; QIIME) soft-
ware was used for subsequent analyses. The sequence data are
available in the European Nucleotide Archive under accession
number PRJEB13307.
Statistical analysis
The observed operational taxonomic unit index (richness), Shan-
non index (richness and evenness) and PD_whole_tree index
(phylogenetic diversity) were used to analyze alpha diversity
(within-sample microbial diversity). Beta diversity (between-sam-
ple microbial diversity) was calculated by principal coordinate
analysis based on unweighted uniFrac distances. The linear dis-
criminant analysis effect size (LEfSe) method was used with a
threshold cut-off value of 2.0 to identify biomarkers that differed
between the groups. The Wilcoxon rank sum test was performed
for comparisons of acne grades. P < 0.05 was considered statis-
tically significant.
Skin microbiomes of acne patients and healthy
controls
The skin microbiomes of both the acne patients and the healthy
controls were dominated by four bacterial phyla (Actinobacteria,
Firmicutes, Proteobacteria and Bacteroidetes) and five genera
(Cutibacterium, Staphylococcus, Streptococcus, Corynebac-
terium and Lactobacillus) with no significant differences; how-
ever, the relative abundance of Oscillospira, Enhydrobacter and
Bacteroides was increased in acne patients (Fig. S1). In addition,
the alpha diversity of acne patients was greater than that of
healthy individuals, and beta diversity also showed differences
between the two groups (Fig. S2).
2 © 2019 Japanese Dermatological Association
Grade 2 (n = 12) Grade 3 (n = 9) Grade 4 (n = 12) P
5/7 1/8 9/3 0.217
23.4  1.24 22.1  2.09 23.8  2.41 0.476
6.75  4.39 5.44  3.47 8.25  3.14 0.002†
Relationship between skin microbiome and acne
grading
The comparison of skin microbiomes among patients with dif-
ferent acne grading showed that the microbial composition
and diversity were similar among the patients with grade 1–3
acne, but patients with grade 4 acne harbored a significantly
different skin microbiome compared with those with grade 1–3
acne, with a significantly increased alpha diversity and remark-
ably different beta diversity (Figs 1a,S3). LEfSe analysis
revealed that the following seven bacterial genera were more
abundant in grade 4 acne: Enhydrobacter, Bacteroides, Faecal-
ibacterium, Klebsiella, Oscillospira, Ruminococcus and Escheri-
chia (Fig. S4). Importantly, we found that four bacterial genera,
including Faecalibacterium, Klebsiella, Odoribacter and Bac-
teroides, were dramatically increased only in grade 4 acne
patients, while no significant differences were observed among
patients with grade 1–3 acne and almost no detectable data
were found in healthy controls (Fig. 1b). Overall, these results
revealed that the skin microbiome is closely correlated with
acne and the acne grading.
DISCUSSION
This is the first study to explore the relationship between the
skin microbiome and grades of acne using 16S rRNA gene
sequencing. We found that the skin microbiome was closely
associated with acne and the grades of acne, providing poten-
tial evidence for a microbial-mediated mechanism of disease
exacerbation and microbial-based treatment of acne.
Compared with healthy controls, three bacterial genera
(Oscillospira, Enhydrobacter and Bacteroides) were increased
in acne patients, suggesting that these genera possibly include
pathogenic bacteria associated with acne. Previous study
using 16S rRNA gene sequencing suggested that Cutibac-
terium was closely related to increasing disease severity, con-
versely, the abundance of Corynebacterium was inversely
correlated with acne severity.4 Nevertheless, very few studies
have reported skin microbiome differences among patients
with different acne grading. Our results revealed that the skin
microbial structures of patients with grade 1–3 acne, including
the bacterial alpha diversity, beta diversity and distribution of
certain bacterial genera, were similar. However, grade 4 acne
patients showed a significantly different microbiome character-
ized by higher alpha diversity and increased amounts of
Faecalibacterium, Klebsiella, Odoribacter and Bacteroides
compared with grade 1–3 acne patients. Therefore, we
 Skin microbiome and acne grading

Figure 1. (a) Comparison of the skin bacterial community structure between healthy controls and patients with different grades of
acne. Bacterial genera with a relative abundance of more than 1% are presented, and all other genera are classified as “Others”.
(b) Skin microbiome differences among patients with different acne grades. The relative abundances (%) of the facial skin bacterial
genera with significant differences among grades (Wilcoxon rank-sum test) are presented.

considered the four bacterial genera to be possible biomarkers
of grade 4 acne (severe acne), and their overgrowth may be
related to the aggravation of acne. However, most acne-asso-
ciated bacteria are Gram-positive bacteria (such as Cutibac-
terium and Staphylococcus) which trigger acne inflammation
by activating monocytes through Toll-like receptor 2 to pro-
duce tumor necrosis factor-a, and interleukin-12 and -8.8 In
contrast, the four bacterial genera found in our study were all
Gram-negative bacilli, which belong to transient skin bacteria
and were rarely found on the skin.9 Although our data sug-
gested the potential importance of these Gram-negative bacte-
ria, very few studies have investigated the role of these
bacteria in acne due to low and inconsistent yields of them
from skin by culture.10
The pathogenicity of Gram-negative bacteria is mainly attribu-
ted to the secretion of membrane vesicles, which are important to
bacterial physiology and pathogenesis.11 The skin is a physical
barrier between the body and the environment, preventing the
colonization of pathogens.12 Gram-negative bacteria can over-
grow on the skin by destroying the skin barrier. In addition,
Gram-negative bacteria are common bacteria causing skin
infections and play an important role in skin infections,13 for
example, papules, pustules and folliculitis of the facial skin.
Mechanisms including seborrhea, microbial dysbiosis and
impaired immune defense of the host contributed to Gram-nega-
tive folliculitis.14 The four Gram-negative bacteria found in our
results may have more vital significance in severe acne than
Gram-positive bacteria, such as promoting acne inflammation,
impairing immune defense, impacting tissue repair and inhibiting
the growth of resident skin bacteria. Furthermore, it is noteworthy
that the antibiotic resistance of Gram-negative bacteria is higher
than that of Gram-positive bacteria,15 suggesting that the treat-
ments of mild to moderate acne may not be suitable for severe
acne.
ACKNOWLEDGMENTS: This research was supported by
the Open Project of the State Key Laboratory of Respiratory Disease
(no. SKLRD2016OP014). We thank American Journal Experts for per-
forming English-language editing.
CONFLICT OF INTEREST: None declared.

© 2019 Japanese Dermatological Association 3

C.-x. Li et al.
REFERENCES
1 Numata S, Akamatsu H, Akaza N, Yagami A, Nakata S, Matsunaga
K. Analysis of facial skin-resident microbiota in Japanese acne
patients. Dermatology 2014; 228: 86–92.
2 Holland KT, Bojar RA. Cosmetics: what is their influence on the skin
microflora? Am J Clin Dermatol 2002; 3: 445–449.
3 Weyrich LS, Dixit S, Farrer AG, Cooper AJ, Cooper AJ. The skin
microbiome: associations between altered microbial communities
and disease. Australas J Dermatol 2015; 56: 268–274.
4 Kelha€ la
€ HL, Aho VTE, Fyhrquist N et al. Isotretinoin and lymecycline
treatments modify the skin microbiota in acne. Exp Dermatol 2018;
27: 30–36.
5 Bergler-Czop B. The aetiopathogenesis of acne vulgaris - what’s
new? Int J Cosmet Sci 2014; 36: 187–194.
6 Nast A, Rosumeck S, Erdmann R, Alsharif U, Dressler C, Werner
RN. Methods report on the development of the European evidence-
based (S3) guideline for the treatment of acne - update 2016. J Eur
Acad Dermatol Venereol 2016; 30: e1–e28.
7 Su J, Liu H Y, Tan X L et al. Sputum bacterial and fungal dynamics
during exacerbations of severe COPD. PLoS One 2015; 10:
e0130736.
8 Kim J. Review of the innate immune response in acne vulgaris: acti-
vation of Toll-like receptor 2 in acne triggers inflammatory cytokine
responses. Dermatology 2005; 211: 193–198.
9 Mackowiak PA. The normal microbial flora. N Engl J Med 1982;
307: 317–320.
10 Myles IA, Reckhow JD, Williams KW, Sastalla I, Frank KM, Datta
SK. A method for culturing Gram-negative skin microbiota. BMC
Microbiol 2016; 16: 60.
11 Roier S, Zingl FG, Cakar F, Schild S. Bacterial outer membrane
vesicle biogenesis: a new mechanism and its implications. Microb
Cell 2016; 3: 257–259.
12 Andonova M, Urumova V. Immune surveillance mechanisms of the
skin against the stealth infection strategy of Pseudomonas aeruginosa
—Review. Comp Immunol Microbiol Infect Dis 2013; 36: 433–448.
4 © 2019 Japanese Dermatological Association
13 Ruef C. Complicated skin and soft-tissue infections–consider gram-
negative pathogens. Infection 2008; 36: 295.
14 Neubert U, Jansen T, Plewig G. Bacteriologic and immunologic
aspects of gram-negative folliculitis: a study of 46 patients. Int J
Dermatol 1999; 38: 270–274.
15 Arzanlou M, Chai WC, Venter H. Intrinsic, adaptive and acquired
antimicrobial resistance in Gram-negative bacteria. Essays Biochem
2017; 61: 49–59.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the online
version of this article:
Figure S1. Skin microbiome differences between healthy con-
trols and patients with acne.
Figure S2. (a) Box plot showing the alpha diversity of the skin
bacterial community collected from healthy controls and acne
patients. Observed operational taxonomic unit (OTU) index,
Shannon index and PD_whole_tree were used as richness,
evenness and richness metrics and phylogenetic, respectively.
P-values were calculated using the Wilcoxon rank sum test to
compare healthy controls and acne patients. (b) Principal coor-
dinate analysis (PCoA) plot (beta diversity) based on the
unweighted uniFrac distance of the skin bacterial community
between the healthy controls and acne patients.
Figure S3. (a) Alpha diversity and (b) beta diversity of the facial
skin microbiome of the healthy controls and patients with dif-
ferent grades of acne.
Figure S4. Skin microbiome differences between the grade 4
acne patients and healthy controls.