ORIGINAL ARTICLE

Propionibacterium acnes Strain Populations in the

Human Skin Microbiome Associated with Acne
Sorel Fitz-Gibbon1,8, Shuta Tomida1,8, Bor-Han Chiu1, Lin Nguyen1, Christine Du1,2, Minghsun Liu3,
David Elashoff2, Marie C. Erfe4, Anya Loncaric2, Jenny Kim2,5, Robert L. Modlin2,3, Jeff F. Miller3,
Erica Sodergren6, Noah Craft4, George M. Weinstock6 and Huiying Li1,7

The human skin microbiome has important roles in skin health and disease. However, bacterial population
structure and diversity at the strain level is poorly understood. We compared the skin microbiome at the strain
level and genome level of Propionibacterium acnes, a dominant skin commensal, between 49 acne patients and
52 healthy individuals by sampling the pilosebaceous units on their noses. Metagenomic analysis demonstrated
that although the relative abundances of P. acnes were similar, the strain population structures were significantly
different in the two cohorts. Certain strains were highly associated with acne, and other strains were enriched in
healthy skin. By sequencing 66 previously unreported P. acnes strains and comparing 71 P. acnes genomes, we
identified potential genetic determinants of various P. acnes strains in association with acne or health. Our
analysis suggests that acquired DNA sequences and bacterial immune elements may have roles in determining
virulence properties of P. acnes strains, and some could be future targets for therapeutic interventions. This study
demonstrates a previously unreported paradigm of commensal strain populations that could explain the
pathogenesis of human diseases. It underscores the importance of strain-level analysis of the human microbiome
to define the role of commensals in health and disease.
Journal of Investigative Dermatology advance online publication, 28 February 2013; doi:10.1038/jid.2013.21

INTRODUCTION Chase-Topping et al., 2008). Acne vulgaris (commonly called

The diversity of the human microbiota at the strain level and acne) is one of the most common skin diseases with a
its association with human health and disease are largely prevalence in up to 85% of teenagers and 11% of adults
unknown. However, many studies have shown that microbe- (White, 1998). Although the etiology and pathogenesis of acne
related human diseases are often caused by certain strains of a are still unclear, microbial involvement is considered to be
species, rather than the entire species being pathogenic. one of the main mechanisms contributing to the development
Examples include methicillin-resistant Staphylococcus aureus of acne (Cunliffe, 2002; Bojar and Holland, 2004). In
(Chambers and Deleo, 2009; Chen et al., 2010; Hansra and particular, Propionibacterium acnes has been hypothesized
Shinkai, 2011) and Escherichia coli O157 (Tarr et al., 2005; to be an important pathogenic factor (Webster, 1995).
Antibiotic therapy targeting P. acnes has been a mainstay
1
Department of Molecular and Medical Pharmacology, Crump Institute for
treatment for more than 30 years (Leyden, 2001). However,
Molecular Imaging, David Geffen School of Medicine, UCLA, Los Angeles, despite decades of study, it is still not clear how P. acnes
California, USA; 2Department of Medicine, David Geffen School of Medicine, contributes to acne pathogenesis while being a major
UCLA, Los Angeles, California, USA; 3Department of Microbiology, commensal of the normal skin flora (Gao et al., 2007;
Immunology and Molecular Genetics, David Geffen School of Medicine,
UCLA, Los Angeles, California, USA; 4Center for Immunotherapeutics Bek-Thomsen et al., 2008; Cogen et al., 2008; Fierer et al.,
Research, Los Angeles Biomedical Research Institute at Harbor-UCLA Medical 2008; Costello et al., 2009; Grice et al., 2009; Dominguez-
Center, David Geffen School of Medicine, UCLA, Los Angeles, California, Bello et al., 2010). Whether P. acnes protects the human skin
USA; 5Department of Dermatology, Veterans Affairs Greater Los Angeles
Healthcare System, Los Angeles, California, USA; 6The Genome Institute at
as a commensal bacterium or functions as a pathogenic factor
Washington University, St Louis, Missouri, USA and 7UCLA-DOE Institute for in acne, or both, remains to be elucidated.
Genomics and Proteomics, Los Angeles, California, USA Here we compared the skin microbiome at the strain level
8
These authors contributed equally to this work. and genome level in 49 acne patients and 52 normal
Correspondence: Huiying Li, 4339 CNSI, 570 Westwood Plaza, Building 114, individuals using a combination of metagenomics and gen-
UCLA, Los Angeles, California 90095-1770, USA. ome sequencing. First, for each sample, 16S ribosomal DNA
E-mail: huiying@mednet.ucla.edu
(rDNA) was amplified, B400 clones were sequenced, and an
Abbreviations: CRISPR, Clustered Regularly Interspaced Short Palindromic
average of 311 nearly full-length 16S rDNA sequences were
Repeat; HMP, Human Microbiome Project; rDNA, ribosomal DNA; RT,
ribotype analyzed. The population structure of P. acnes strains was
Received 16 August 2012; revised 19 November 2012; accepted 3 December determined in each sample. Second, each P. acnes strain
2012; accepted article preview online 21 January 2013 was assigned an ‘‘acne index’’ by calculating its prevalence in

& 2013 The Society for Investigative Dermatology www.jidonline.org 1

 S Fitz-Gibbon et al.
The Skin Microbiome Associated with Acne

acne patients based on the 16S rDNA metagenomic data. The sequences, 27,358 matched to P. acnes with greater than 99%
P. acnes strains associated with the acne patient group were identity. Our data demonstrated that P. acnes dominates the
identified, as well as the strains enriched in the individuals microbiota of pilosebaceous units, accounting for 87% of the
with normal skin. This metagenomic approach is fundamen- clones (Figure 1). Other commonly found species in pilose-
tally different from prior approaches in determining disease baceous units included Staphylococcus epidermidis, Propio-
associations; it is more powerful and less biased than nibacterium humerusii, and Propionibacterium granulosum,
traditional methods by bypassing the biases and selection in each representing 1–2.3% of the total clones. A total of 536
strain isolation and culturing. To our knowledge, this study has species-level operational taxonomic units belonging to 42
the largest number of individual skin microbiomes reported at genera and six phyla were identified in the samples
the strain level to date. Finally, we sequenced 66 previously (Supplementary Table S1 online).
unreported P. acnes strains and compared 71 P. acnes To bypass the potential biases due to PCR amplification and
genomes covering the major lineages of P. acnes found in due to uneven numbers of 16S rDNA gene copies among
the skin microbiota. By combining a metagenomic study of different species, we performed a metagenomic shotgun
the skin microbiome and genome sequencing of this major sequencing of the total DNA pooled from the pilosebaceous
skin commensal, this study provides insight into potential unit samples of 22 additional normal individuals. Microbial
bacterial genetic determinants in acne pathogenesis and species were identified by mapping metagenomic sequences
emphasizes the importance of strain-level analysis of the to reference genomes. The results confirmed that P. acnes was
human microbiome to understand the role of commensals in the most abundant species (89%) (Figure 1). This is consistent
health and disease. with the results obtained from 16S rDNA sequencing (87%).

RESULTS Different P. acnes strain populations in acne

P. acnes dominates the pilosebaceous unit There was no statistically significant difference in the relative
We characterized the microbiome in pilosebaceous units abundance of P. acnes when comparing acne patients and
(‘‘pores’’) on the nose, collected from 49 acne patients normal individuals. We next examined whether there were
and 52 individuals with normal skin. Nearly full-length 16S differences at the strain level of P. acnes by extensively
rDNA sequences were obtained using the Sanger method, analyzing the P. acnes 16S rDNA sequences. We define each
which allowed us to analyze P. acnes at the strain level. After unique 16S rDNA sequence as a 16S rDNA allele type, called
quality filtering, our final data set contained 31,461 16S rDNA a ribotype (RT). The most abundant P. acnes sequence was
sequences ranging from position 29 to position 1,483. Of the defined as ribotype 1 (RT1); all other defined ribotypes have

Pooled normal skin

samples by metagenomic
Samples from acne patients Samples from normal individuals shotgun sequencing

100%

90%

80%
Relative abundance of species

70%

60%

50%

40%

30%

20%

10%

0%

Propionibacterium acnes Staphylococcus epidermidis Propionibacterium humerusii

Propionibacterium granulosum Prevotella oris Unclassified corynebacterineae
Staphylococcus capitis Escherichia coli Other

Figure 1. Propionibacterium acnes was dominant in pilosebaceous units in both acne patients and individuals with normal skin. By 16S ribosomal DNA (rDNA)
sequencing, P. acnes sequences accounted for 87% of all the clones. Species with a relative abundance of 40.35% are listed in order of relative abundance.
Species distribution from a metagenomic shotgun sequencing of pooled samples from normal individuals confirmed the high abundance of P. acnes in
pilosebaceous units, as shown in the far right column.

2 Journal of Investigative Dermatology

 S Fitz-Gibbon et al.
The Skin Microbiome Associated with Acne

Table 1. Top 10 most abundant ribotypes found in pilosebaceous units

Nucleotide Percentage Percentage of
changes Number Number of clones from clones from
Ribotype (RT) from RT1 of subjects of clones acne patients1 normal individuals2 P-value3

RT1 — 90 5,536 48% 52% 0.84

RT2 T854C 48 1,213 51% 49% 0.36
RT3 T1007C 60 2,104 40% 60% 0.092
RT4 G1058C, A1201C 23 275 84% 16% 0.049
RT5 G1058C 15 205 99% 1% 0.00050
RT6 T854C, C1336T 11 262 1% 99% 0.025
RT7 G529A 10 188 99% 1% 0.12
RT8 G1004A, T1007C 5 239 100% 0% 0.024
RT9 G1268A 4 68 99% 1% 0.29
RT10 T554C, G1058C 5 61 100% 0% 0.024
1
The percentage was calculated after the number of clones of each ribotype was normalized by the total number of clones in acne patients (acne index).
2
The percentage was calculated after the number of clones of each ribotype was normalized by the total number of clones in normal individuals.
3
Mann–Whitney–Wilcoxon rank-sum test.

Z99% sequence identity to RT1. Similar to the distributions antibiotics harbored strains already resistant to antibiotics
seen at higher taxonomical levels (Bik et al., 2010), at the (Dreno et al., 2001; Coates et al., 2002). On the other hand,
strain level a few ribotypes were highly abundant in the one ribotype, RT6, although detected in only 11 subjects, was
samples with a significant number of rare ribotypes strongly associated with normal skin (P ¼ 0.025, Wilcoxon
(Supplementary Figure S1 online). After careful examination test; Table 1). Its relative abundance in our normal group
of the sequence chromatograms and manual correction of the was similar to that found in the healthy cohort data from the
sequences, a total of 11,009 ribotypes were assigned to the Human Microbiome Project (HMP) (see Supplementary
P. acnes 16S rDNA sequences. Most of the minor ribotypes Information and Supplementary Figure S2 online). The per-
were singletons. On average, each individual harbored 3±2 centage of positive subjects (11/52) was similar as well. Three
P. acnes ribotypes with three or more clones. On the basis of of the 14 HMP subjects had RT6 found in the anterior nares,
the genome sequences described below, all the sequenced and one additional subject had RT6 in the left retroauricular
P. acnes strains have three identical copies of 16S rDNA genes crease.
(note in Supplementary Information online). This allowed us to On the basis of the distributions of the top 10 ribotypes,
compare the P. acnes strain populations in individuals based statistical analysis using several different tests showed signifi-
on the 16S rDNA sequences. The top 10 major ribotypes with cant differences in P. acnes population structure between acne
more than 60 clones and found in multiple subjects are shown and normal skin (Supplementary Table S3 online). This is
in Table 1. consistent with a principal coordinate analysis, in which acne
Analysis of the top 10 ribotypes showed both disease- samples and normal skin samples were separated mostly
specific and health-specific associations. The three most by principal coordinates 1 and 2 (Supplementary Figure S3
abundant ribotypes (RT1, RT2, and RT3) were fairly evenly online), explaining 44% and 20% of the variation,
distributed among acne and normal individuals. However, the respectively.
next seven major ribotypes were significantly skewed in their To examine whether different individuals share similar
distributions (Table 1). Ribotypes 4, 5, 7, 8, 9, and 10 were P. acnes population structures, we clustered the samples on
found predominantly in acne patients, with four of these six the basis of the relative abundances of the top 10 ribotypes.
ribotypes being statistically significantly enriched in acne Five main microbiome types were observed at the P. acnes
(Po0.05, Wilcoxon test). Ribotypes 4, 5, and 10 contain a strain level (microbiome types I to V). Types IV and V, which
nucleotide substitution G1058C in the 16S rDNA sequences, are dominated by P. acnes RT4 and RT5, respectively, were
which has previously been shown to confer increased resis- mainly found in acne patients (Figure 2 and Supplementary
tance to tetracycline (Ross et al., 1998, 2001). However, only Figure S4 online). The same five main microbiome types were
a small percentage of the subjects in our study harboring these observed in the HMP data and the data from Grice et al.
ribotypes had been treated with antibiotics (Supplementary (2009) (see Supplementary Information and Supplementary
Table S2 online); therefore, enrichment of these three Figure S5 online).
ribotypes in the acne group was not correlated with antibiotic
treatment. This is consistent with previous studies, which Genome sequence analysis of 71 P. acnes strains
showed that previous use of antibiotics was not always All of the top 10 most abundant ribotypes differ from RT1 by
associated with the presence of antibiotic-resistant strains only one or two nucleotide changes in the 16S rDNA
and that some patients who were not previously treated with sequence (Table 1). To determine whether such small changes

www.jidonline.org 3
 S Fitz-Gibbon et al.
The Skin Microbiome Associated with Acne

I II IV V III II I III
100%

90%
Relative abundance of the top 10 ribotypes

80%

RT10
70%
RT9
RT8
60%
RT7
RT6
50%
RT5
RT4
40%
RT3
30% RT2
RT1
20%

10%

0%
Acne (n =48) Normal (n =51)

I Microbiome type I II Microbiome type II III Microbiome type III

IV Microbiome type IV V Microbiome type V Minor microbiome types

Figure 2. Distribution of the top 10 most abundant Propionibacterium acnes ribotypes in acne patients and individuals with normal skin. Each column
represents the percentage of the top 10 ribotypes identified in each subject. The average P. acnes clone number per subject was 262, and the average clone
number of the top 10 ribotypes was 100. Five major microbiome types at the P. acnes strain level were observed in the data. Types IV and V were mostly found in
acne patients. Two samples (one from acne and one from normal skin) with o50 P. acnes 16S ribosomal DNA (rDNA) sequences are not displayed.

in the 16S rDNA sequence reflect the lineages and enriched in normal skin. Three distinct regions, loci 1, 2,
evolutionary history at the genome level, we selected 66 and 3, were found almost exclusively in strains that belong to
P. acnes isolates representing major ribotypes 1, 2, 3, 4, 5, 6, clade IA-2 in the phylogenetic tree. Clade IA-2 consists of
and 8, as well as two minor ribotypes, 16 and 532, for genome mainly RT4 and RT5 (Figure 3 and Supplementary Figure S7
sequencing. The genomes of these 66 isolates were fully online). Loci 1 and 2 are located on the chromosome. Locus 1
sequenced and assembled to high-quality drafts or complete contains prophage-related genes and appears to be a genomic
genomes with Z50X coverage. Five other P. acnes genomes, island. Locus 2 has plasmid integration sites and thus could be
KPA171202 (Bruggemann et al., 2004), J165, J139, SK137, derived from a plasmid sequence. Locus 3 appears to be on a
and SK187, were publicly available and were included in our large mobile genetic element, likely a plasmid. The plasmid is
analysis. We constructed a phylogenetic tree based on 96,887 B55 Kb long and has inverted terminal repeats according to
unique single-nucleotide polymorphism positions in the core our finished genome HL096PA1 (Supplementary Information
genome obtained from these 71 P. acnes genomes. Most of online). The sequence data suggest that the plasmid is linear
the genomes with the same ribotypes clustered together. and possibly originated from a phage (Hinnebusch and Tilly,
The tree suggests that the 16S rDNA ribotypes do represent 1993). All but 1 of the 15 genomes of RT4 and RT5 have at
the relationship of the lineages to a large extent, and that16S least 60% of the genes of the plasmid represented, and all of
rDNA sequence is a useful molecular marker to distinguish them have regions homologous to the inverted terminal
major P. acnes lineages (Figure 3 and Supplementary repeats in the plasmid, suggesting that they harbor the same
Figure S6 online). or a similar linear plasmid (Figure 3). The copy number of the
plasmid in the genomes ranges from 1 to 3 based on genome
Genetic elements detected in P. acnes sequencing coverage, which was confirmed by quantitative
We further performed comparative genome analysis among all PCR (Supplementary Figures S8 and S9 online).
71 genomes grouped by ribotypes. Our analysis revealed The fact that acne-enriched RT4 and RT5 strains carry a
potential genetic elements by which acne-associated strains linear plasmid and two unique loci of genomic islands
could contribute to acne pathogenesis and the elements suggests that these plasmid and chromosomal regions may
by which health-associated strains could contribute to have a role in acne pathogenesis. In fact, the linear plasmid
maintaining skin health. Specifically, we describe here the encodes a tight adhesion (Tad) locus, which has been sugges-
unique genome regions of RT4 and RT5, which had a strong ted to have a role in virulence in other organisms (Kachlany
association with acne, and RT6, which was found to be et al., 2000; Schreiner et al., 2003). The complete Tad locus is

4 Journal of Investigative Dermatology

 S Fitz-Gibbon et al.
The Skin Microbiome Associated with Acne

Clade Acne index Locus 1 Locus 2 Locus 3 (plasmid)

IA-1

IA-2

IB-1

IB-2

IB-3

II

Genes 1–300 on the chromosome Genes 1–74 on the plasmid

RT1 RT2 (T854C) RT3 (T1007C) RT4 (G1058C, A1201C) RT5 (G1058C)
RT6 (T854C, C1336T) RT8 (G1004A, T1007C) RT16 (C1001T, T1007C) RT532 (G88A)

Acne index for ribotype Normal Acne

0 20 40 60 80 100 %

Figure 3. Genome comparison of 71 Propionibacterium acnes strains showed that the genomes of ribotypes 4 and 5 (RT4 and RT5) are distinct from those of
others. Two chromosomal regions, loci 1 and 2, are unique to clade IA-2 and one other genome HL086PA1. Clade IA-2 consists of mainly RT4 and RT5 that
were highly enriched in acne. The presence of a plasmid (locus 3) is also characteristic of RT4 and RT5. Each row represents a P. acnes genome colored
according to the ribotypes. Rows are ordered by the phylogeny calculated based on the single-nucleotide polymorphisms (SNPs) in the P. acnes core genome.
Only the topology is shown. The clades were named based on their recA types (IA, IB, and II). Columns represent predicted open reading frames (ORFs) in the
genomes and are ordered by ORF positions along the finished genome HL096PA1, which encodes a 55-Kb plasmid. Only the first 300 ORFs on the chromosome
(on the left) and all the ORFs on the plasmid (on the right) are shown. The colored plasmid regions represent genes on contigs that match exclusively to the
HL096PA1 plasmid region. The genes that fall on contigs that clearly extend beyond the plasmid region are likely to be chromosomally located and are colored in
gray. Acne index for the ribotypes was calculated based on the percentage of clones of each ribotype found in acne as shown in column 5 in Table 1.

found in all but 1 of the 15 genomes of RT4 and RT5, and is (Horvath and Barrangou, 2010; Makarova et al., 2011). The
only occasionally found in other ribotypes. In addition, in CRISPR locus encoded in P. acnes consists of a series of cas
locus 2, a Sag gene cluster is encoded, which has been shown genes—cas3, cse1, cse2, cse4, cas5e, cse3, cas1, and cas2—
to contribute to hemolytic activity in pathogens (Nizet et al., which are homologous to the CRISPR locus reported in
2000; Fuller et al., 2002; Humar et al., 2002). Supplementary E. coli (Supplementary Figure S10 online) and the CRISPR4
Table S4 online summarizes the genes that are mostly unique locus in Streptococcus thermophilus (Horvath and Barrangou,
to RT4 and RT5, several of which have essential roles in 2010).
virulence in other organisms. We speculate that some of these CRISPR arrays are composed of a cluster of identical
genes encoded in RT4 and RT5 may increase virulence, repetitive sequences separated by spacer sequences of similar
promote stronger adherence to the human host, or induce a length but with different nucleotide sequences. It has been
pathogenic host immune response. found that spacer sequences are identical, or with one or two
In genome comparison analysis, we found that all the mismatches, to phage or plasmid DNA sequences. A total of
genomes of RT2 and RT6 encode CRISPRs (Clustered 39 spacer sequences were found in 8 P. acnes strains, 25 of
Regularly Interspaced Short Palindromic Repeats). Among which were unique, as shown in Table 2. As expected, most of
the sequenced genomes, RT2 and RT6 are the only ribotypes the identifiable spacers target to known P. acnes phage
encoding CRISPRs. CRISPRs have been shown to confer sequences. However, among the unique CRISPR spacer
protective ‘‘immunity’’ against viruses, phages, and plasmids sequences, one matched locus 2 on the chromosome and

www.jidonline.org 5
 S Fitz-Gibbon et al.
The Skin Microbiome Associated with Acne

Table 2. CRISPR spacer sequences found in the genomes of RT2 and RT6
Spacer Match
Ribotype Strain number Spacer sequence BLAST result found

RT2 HL001PA1 1 CATGGCCTGCACACCAGGCGCTTTTAGCACCT No hits

2 CATGGCCTGCACACCAGGCGCTTTTAGCACCT No hits
3 CATGGCCTGCACACCAGGCGCTTTTAGCACCT No hits
4 GGCGTATGACGAGTTGTGGTCGGCGTTTCCTC P. acnes phage PA6 gp15 (minor tail protein)
5 CGGTGTTAACGGCTTGCCTGGCTTGGATGGAG No hits
RT2 HL060PA1 1 CGCCTACCGTCAGCTGACTCACGCCTCCGCGTT No hits
2 TCACACCAGTCATCAGCGTCATAGTCCTCTCGG No hits
RT2 HL082PA2 1 GGCTCAGCCCTGCCCGATGCCTACGCCAAATGG C. leptum DSM 753 CLOLEP_00129 Locus 3
(cell wall–associated hydrolases (invasion-
associated proteins))
2 TCACACCAGTCATCAGCGTCATAGTCCTCTCGG No hits
RT2 HL103PA1 1 CACCGGGCCCATCCCGGTCGGCCTCCTGAAAGG C. leptum DSM 753 CLOLEP_00135 Locus 3
RT2 HL106PA1 1 GATCGAGTTGGCTGAGTCGAAGGTGTTGCGGTT P. acnes phage PA6 gp16 (conserved protein)
P. acnes phage PAD20 gp16
2 CTGCTCATCGCTCAGCTCCTGCGCCTCATCACA No hits
3 CTGCGCCAACAGCCGCATCTGATCCGAATACGG P. acnes phage PA6 gp3 (phage portal protein)
4 CGCAGCAATCTCAGAAGGCCACAACAAGTTCGT P. acnes phage PA6 gp7 (conserved protein)
P. acnes phage PAD20 gp7
P. acnes phage PAS50 gp7
5 CAAATCACCCAAGCCCAACACGCCGCCACCACC No hits
6 TGTCACCGATTCAATGTATCTATGAGTGGTGTA No hits
7 TTGGGTGGGTGAGGTCGGGTCGTCAGTCATGAG No hits
8 GTCGATGTCGAGATTGGCCTGGGGGTCCATGTC C. leptum DSM 753 CLOLEP_00142 Locus 3
9 ACGTCGTGAACGTACCCCTTGACGGAGACGGCA No hits
RT2 J139 1 CGAGGGCTACCACGTGGTCGATTTGGACTGTCG C. leptum DSM 753 CLOLEP_00167 Locus 2
P. acnes SK137 HMPREF0675_3193 (domain of
unknown function)
2 CAGGCGCTCCACTCCCTCGCCCTGGCCACCAAC No hits
RT6 HL110PA3 1 CTATGTGGACAGTGTTGGTTACTGTGGGGGGAA P. acnes phage PA6 intergenic region between
HL110PA4 gp45 and gp46
2 GCACTGCACCGATATCGTCGTGGCTGTCACTTG No hits
3 CCCAGACAACCTCGACAACCTGTTCAGGGGATG P. acnes phage PAS50 gp25
4 CATGGCTAGCCCGGATTTTTGGCTGCCTGAGCG P. acnes phage PA6 gp34 (multidrug resistance
protein-like transporters)
P. acnes phage PAD20 gp34 (DNA helicase)
5 CGGCCTGCGGCAGATTTTTGTTGCGTTGAATCC P. acnes phage PA6 gp14 (tape measure protein)
P. acnes phage PAD20 gp14 (tape measure protein)
P. acnes phage PAS50 gp14 (tape measure protein)
6 CGGGCAGAGGATGTGTTGCTCGTTCCTGGATGG P. acnes phage PA6 gp32 (CHC2 zinc-finger)
P. acnes phage PAD20 gp32 (DNA primase)
P. acnes phage PAS50 gp32 (DNA primase)
7 GTTACGCTGGAACCCCCAATGAACACGCGAGAA P. acnes phage PAD42 major head protein gene
P. acnes phage PAD20 major head protein gene
P. acnes phage PAD9 major head protein gene
P. acnes phage PAS40 major head protein gene
P. acnes phage PAS12 major head protein gene
P. acnes phage PAS10 major head protein gene
P. acnes phage PAD21 major head protein gene
P. acnes phage PAS2 major head protein gene
P. acnes phage PA6 gp6 (phage capsid family)
P. acnes phage PAS50 gp6 major head protein gene
8 CGAGGGCTACCACGTGGTCGATTTGGACTGTCG C. leptum DSM 753 CLOLEP_00167 Locus 2
P. acnes SK137 HMPREF0675_3193 (domain of
unknown function)
9 CAGGCGCTCCACTCCCTCGCCCTGGCCACCAAC No hits
Abbreviations: BLAST, Basic Local Alignment Search Tool; CRISPR, Clustered Regularly Interspaced Short Palindromic Repeat; C. leptum, Clostridium leptum;
P. acnes, Propionibacterium acnes; RT, ribotype.

three matched the plasmid region (locus 3) in P. acnes genomes of RT2 and RT6 may be capable of protecting against
genomes of mainly RT4 and RT5. This suggests that these loci the invasion of the plasmids or other foreign DNA through the
may have been acquired by RT4 and RT5 strains, whereas the CRISPR mechanism.

6 Journal of Investigative Dermatology

 S Fitz-Gibbon et al.
The Skin Microbiome Associated with Acne

DISCUSSION id=phs000263.v1.p1). The diagnosis of acne was made by board-

Our study of the human skin microbiome associated with acne certified dermatologists. The presence of acne was graded on a
provides a previously unreported portrait of the microbiota of scale of 0 to 5 relating closely to the Global Acne Severity Scale
pilosebaceous units at the bacterial strain level. As P. acnes is (Dreno et al., 2011). Grades were recorded for both the face and
the major skin commensal bacterium found in both acne and the nose separately, where 0 represents normal skin and 5
healthy skin, this strain-level analysis is important to help represents the most severe inflammatory cystic acne. In acne
understand the role of P. acnes in acne pathogenesis and in patients, the grades of the face ranged from 1 to 5, with an
skin health. We demonstrate a strong association between average of 2.1, and the grades of the nose ranged from 0 to 2, with
strains of RT4 and RT5 with acne and a strong association an average of 0.3. The presence of scarring was also noted.
between strains of RT6 and healthy skin, each with unique Subjects with normal skin were determined by board-certified
genetic elements. Other P. acnes strains, including ribotypes dermatologists and were defined as people who had no
7, 8, 9, and 10, or interactions among different strains, may acneiform lesions on the face, chest, or back. They were also
also contribute to the development of the disease. In addition, excluded if they had other skin problems that the investigators felt
host factors, such as hormone level, sebum production, and would affect sampling or the microbial population on the skin.
physical changes in the pilosebaceous unit, may also have a Among the 101 subjects, 59 were female (31 acne patients and 28
role in acne pathogenesis. Further studies aimed at identifying normal subjects) and 42 were male (18 acne patients and 24
the specific functions of these strains, host factors in the normal subjects). The average age of the acne cohort was 22.2, and
development of acne, as well as the associations of micro- the average age of the normal cohort was 29.6. There was no
biome characteristics with the subtypes of acne (comedonal, significant difference in ethnicity between the acne and normal
pustular, inflammatory, cystic, etc.) with larger cohort sizes, populations. The subjects responded to a written questionnaire,
may improve our understanding of the molecular mechanisms administered by a physician or a well-trained study coordinator
of the disease. These studies may also help develop a targeted who went over each question with the subjects. Most of the
therapeutic approach to treat this extremely common and subjects had not been treated for acne in the past or were not
sometimes disfiguring skin disease. being treated when samples were collected (Supplementary Table
Our metagenomic approach in revealing the association of S2). Only 9 out of 78 subjects, who provided treatment informa-
P. acnes strains with the disease or health is, to our knowl- tion, were being treated for acne when samples were taken. Among
edge, previously unreported, and is more powerful than the nine subjects, two were being treated with antibiotics, five were
previous studies using traditional methods (Lomholt and being treated with topical retinoids, one was being treated with
Kilian, 2010; McDowell et al., 2011). Because the skin both antibiotics and retinoids, and one did not list the treatment.
microbiota of each individual and each skin site may harbor We also asked subjects for acne treatment history in the past
‘‘good,’’ ‘‘neutral,’’ and ‘‘bad’’ strains at the same time, which (anytime in their life). Out of 73 subjects who provided treatment
may have different growth rates under in vitro culturing history, 18 had been treated for acne in the past. Among them,
conditions, culturing a few isolates from a disease lesion or seven had been treated with antibiotics, eight had been treated with
healthy skin site may not provide an accurate and unbiased retinoids, two had been treated with both antibiotics and retinoids,
measurement of the association of the strains with the disease and one did not list the treatment. All subjects provided written
or health. The sampling technique and disease associations in informed consent. All protocols and consent forms were approved
this study do not depend on sampling locations, on the by the institutional review boards of both the University of
presence of lesions in the sampling field, or on inherently California, Los Angeles and the Los Angeles Biomedical Research
biased culture techniques. Although sampling lesional skin Institute. The study was conducted in adherence to the Declaration
intentionally may yield interesting results, these results would of Helsinki Principles.
not be capable of defining the disease associations that
unbiased sampling can. The metagenomic approach used in Samples
this study to identify underlying strain differences in acne Skin microcomedone (white head or black head) samples were taken
might also be applied to the study of other disease/health from the nose of the subjects using Bioré Deep Cleansing Pore Strips
associations with commensal or pathogenic bacteria. (Kao Brands Company, Cincinnati, OH) by following the instructions
Ultimately, these studies could lead to targeted therapeutics of the manufacturer. Clean gloves were used for each sampling. After
to restore natural commensal population structures, and could being removed from the nose, the strip was immediately placed into a
help determine whether therapeutic modulations of the 50-ml sterile tube and kept on ice or at 4 1C. The cells were lysed
microbiota can return the host to a state of health. within 4 hours in most of the cases.

MATERIALS AND METHODS Metagenomic DNA extraction, 16S rDNA amplification, cloning,
Subjects and sequencing
Subjects with acne and subjects with normal skin were recruited from Individual microcomedones were isolated from the adhesive nose
various clinics in Southern California, including private practice, strip using sterile forceps. Genomic DNA was extracted using the
managed care, and public hospital settings, as well as outside of QIAamp DNA Micro Kit (Qiagen, Valencia, CA). 16S rDNA was
dermatology clinics, to best represent the diversity of populations and amplified and cloned according to the protocol by HMP, which is
history of medical care. The subject data are available at dbGaP described in detail in Supplementary Information online. Nearly full-
(http://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_ length sequences were obtained by the Sanger method.

www.jidonline.org 7
 S Fitz-Gibbon et al.
The Skin Microbiome Associated with Acne

16S rDNA sequence analysis REFERENCES

Base calling and quality was determined with Phred (Ewing and Bek-Thomsen M, Lomholt HB, Kilian M (2008) Acne is not associated with
Green, 1998; Ewing et al., 1998). Bidirectional reads were assembled yet-uncultured bacteria. J Clin Microbiol 46:3355–60
and aligned to a core set of NAST-formatted sequences (rRNA16S. Bik EM, Long CD, Armitage GC et al. (2010) Bacterial diversity in the oral
gold) using AmosCmp16Spipeline and NAST-ier (Haas et al., 2011). cavity of 10 healthy individuals. ISME J 4:962–74
Suspected chimeras were identified using ChimeraSlayer and Bojar RA, Holland KT (2004) Acne and Propionibacterium acnes.
Clin Dermatol 22:375–9
WigeoN (Haas et al., 2011). 16S rDNA sequences were extensively
manually examined. Chromatograms were visually inspected at all Borodovsky M, McIninch J (1993) Recognition of genes in DNA sequence with
ambiguities. Biosystems 30:161–71
bases, with a Phred quality score of o30. Appropriate corrections
Bruggemann H, Henne A, Hoster F et al. (2004) The complete genome
were applied. QIIME (Caporaso et al., 2010) was used to cluster the sequence of Propionibacterium acnes, a commensal of human skin.
sequences into operational taxonomic units. Science 305:671–3
Caporaso JG, Kuczynski J, Stombaugh J et al. (2010) QIIME allows analysis of
P. acnes isolation and genotyping high-throughput community sequencing data. Nat Methods 7:335–6
Colonies with the macroscopic characteristics of P. acnes were Chambers HF, Deleo FR (2009) Waves of resistance: Staphylococcus aureus in
picked from each sample plate and were passed twice. The ribotype the antibiotic era. Nat Rev Microbiol 7:629–41
of each isolate was determined by PCR amplification and sequencing Chase-Topping M, Gally D, Low C et al. (2008) Super-shedding and the link
of the full length of the 16S rDNA gene by the Sanger method. between human infection and livestock carriage of Escherichia coli O157.
Nat Rev Microbiol 6:904–12

Whole-genome shotgun sequencing, assembly, and annotation Chen CJ, Su LH, Lin TY et al. (2010) Molecular analysis of repeated methicillin-
resistant Staphylococcus aureus infections in children. PLoS One
Genome HL096PA1 was sequenced using Roche/454 FLX and was 5:e14431
assembled using a combination of PHRAP/CONSED (Gordon et al., Coates P, Vyakrnam S, Eady EA et al. (2002) Prevalence of antibiotic-resistant
1998) and GSMAPPER (Roche, Branford, CT) with extensive manual propionibacteria on the skin of acne patients: 10-year surveillance data
editing in CONSED. The remaining 65 genomes were sequenced and snapshot distribution study. Br J Dermatol 146:840–8
using Illumina/Solexa GAIIx (Illumina, San Diego, CA). Sequence data Cogen AL, Nizet V, Gallo RL (2008) Skin microbiota: a source of disease or
sets were processed by quality trimming and were assembled using defence? Br J Dermatol 158:442–55
Velvet (Zerbino and Birney, 2008). Coding sequences were predicted Costello EK, Lauber CL, Hamady M et al. (2009) Bacterial community variation
in human body habitats across space and time. Science 326:1694–7
using GeneMark (Borodovsky and McIninch, 1993) and GLIMMER.
Cunliffe WJ (2002) Looking back to the future - acne. Dermatology 204:
The final gene set was processed through a suite of protein
167–72
categorization tools consisting of Interpro, psort-b, and KEGG. A
Dominguez-Bello MG, Costello EK, Contreras M et al. (2010) Delivery mode
more detailed protocol can be found at http://hmpdacc.org/doc/sops/ shapes the acquisition and structure of the initial microbiota across
reference_genomes/annotation/WUGC_SOP_DACC.pdf. multiple body habitats in newborns. Proc Natl Acad Sci USA
107:11971–5
Comparative genome analysis Dreno B, Poli F, Pawin H et al. (2011) Development and evaluation of a
A total of 71 P. acnes genome sequences were compared using Global Acne Severity Scale (GEA Scale) suitable for France and Europe.
J Eur Acad Dermatol Venereol 25:43–8
Nucmer (Kurtz et al., 2004). Phylogenetic analysis was performed
Dreno B, Reynaud A, Moyse D et al. (2001) Erythromycin-resistance of
using MEGA5 (Tamura et al., 2007). CRISPRFinder (Grissa et al.,
cutaneous bacterial flora in acne. Eur J Dermatol 11:549–53
2007) was used to identify the CRISPR repeat-spacer sequences.
Ewing B, Green P (1998) Base-calling of automated sequencer traces using
The data reported in this paper are listed in Supplementary phred. II. Error probabilities. Genome Res 8:186–94
Information online and archived at GenBank. Ewing B, Hillier L, Wendl MC et al. (1998) Base-calling of automated sequencer
traces using phred. I. Accuracy assessment. Genome Res 8:175–85
CONFLICT OF INTEREST Fierer N, Hamady M, Lauber CL et al. (2008) The influence of sex, handedness,
The authors state no conflict of interest.
and washing on the diversity of hand surface bacteria. Proc Natl Acad Sci
USA 105:17994–9
ACKNOWLEDGMENTS Fuller JD, Camus AC, Duncan CL et al. (2002) Identification of a streptolysin
We thank G Kasimatis, B Shi, EE Curd, R Yan, M Wong, and J Liu for comments S-associated gene cluster and its role in the pathogenesis of Streptococcus
and technical support. We thank C Lee for performing statistical analyses in the iniae disease. Infect Immun 70:5730–9
initial phase. We also thank Z Guo and CS Miller for critical reading of the
manuscript. This research was funded as one of the Demonstration Projects by Gao Z, Tseng CH, Pei Z et al. (2007) Molecular analysis of human forearm
the NIH Human Microbiome Project (HMP). It was supported by grant superficial skin bacterial biota. Proc Natl Acad Sci USA 104:2927–32
UH2AR057503 from NIH/NIAMS and grant U54HG004968 from NIH. Gordon D, Abajian C, Green P (1998) Consed: a graphical tool for sequence
finishing. Genome Res 8:195–202
Author contributions Grice EA, Kong HH, Conlan S et al. (2009) Topographical and temporal
SF-G and ST analyzed the data; B-HC, LN, and CD performed experiments; diversity of the human skin microbiome. Science 324:1190–2
DE performed some of the initial statistical analyses; MCE, AL, JK, RLM, and Grissa I, Vergnaud G, Pourcel C (2007) CRISPRFinder: a web tool to identify
NC collected samples; ES and GMW directed sequencing, genome assembly, clustered regularly interspaced short palindromic repeats. Nucleic Acids
and annotation; HL, ML, RLM, and JFM conceived the demonstration project in Res 35:W52–7
the initial phase; HL designed and directed the project, analyzed the data, and
wrote the paper; and ST, SF-G, and NC co-wrote the paper. Haas BJ, Gevers D, Earl AM et al. (2011) Chimeric 16S rRNA sequence
formation and detection in Sanger and 454-pyrosequenced PCR ampli-
cons. Genome Res 21:494–504
SUPPLEMENTARY MATERIAL
Hansra NK, Shinkai K (2011) Cutaneous community-acquired and hospital-
Supplementary material is linked to the online version of the paper at http:// acquired methicillin-resistant Staphylococcus aureus. Dermatol Ther
www.nature.com/jid 24:263–72

8 Journal of Investigative Dermatology

 S Fitz-Gibbon et al.
The Skin Microbiome Associated with Acne

Hinnebusch J, Tilly K (1993) Linear plasmids and chromosomes in bacteria. Nizet V, Beall B, Bast DJ et al. (2000) Genetic locus for streptolysin S
Mol Microbiol 10:917–22 production by group A streptococcus. Infect Immun 68:4245–54
Horvath P, Barrangou R (2010) CRISPR/Cas, the immune system of bacteria Ross JI, Eady EA, Cove JH et al. (1998) 16S rRNA mutation associated with
and archaea. Science 327:167–70 tetracycline resistance in a gram-positive bacterium. Antimicrob Agents
Humar D, Datta V, Bast DJ et al. (2002) Streptolysin S and necrotising Chemother 42:1702–5
infections produced by group G streptococcus. Lancet 359:124–9 Ross JI, Snelling AM, Eady EA et al. (2001) Phenotypic and genotypic
Kachlany SC, Planet PJ, Bhattacharjee MK et al. (2000) Nonspecific adherence characterization of antibiotic-resistant Propionibacterium acnes isolated
by Actinobacillus actinomycetemcomitans requires genes widespread in from acne patients attending dermatology clinics in Europe, the U.S.A.,
bacteria and archaea. J Bacteriol 182:6169–76 Japan and Australia. Br J Dermatol 144:339–46

Kurtz S, Phillippy A, Delcher AL et al. (2004) Versatile and open software for Schreiner HC, Sinatra K, Kaplan JB et al. (2003) Tight-adherence genes of
comparing large genomes. Genome Biol 5:R12 Actinobacillus actinomycetemcomitans are required for virulence in a rat
model. Proc Natl Acad Sci USA 100:7295–300
Leyden JJ (2001) The evolving role of Propionibacterium acnes in acne. Semin
Cutan Med Surg 20:139–43 Tamura K, Dudley J, Nei M et al. (2007) MEGA4: Molecular Evolutionary
Genetics Analysis (MEGA) software version 4.0. Mol Biol Evol 24:1596–9
Lomholt HB, Kilian M (2010) Population genetic analysis of Propionibacterium
acnes identifies a subpopulation and epidemic clones associated with Tarr PI, Gordon CA, Chandler WL (2005) Shiga-toxin-producing Escherichia
acne. PLoS One 5:e12277 coli and haemolytic uraemic syndrome. Lancet 365:1073–86

Makarova KS, Haft DH, Barrangou R et al. (2011) Evolution and classification Webster GF (1995) Inflammation in acne vulgaris. J Am Acad Dermatol
of the CRISPR-Cas systems. Nat Rev Microbiol 9:467–77 33:247–53

McDowell A, Gao A, Barnard E et al. (2011) A novel multilocus sequence White GM (1998) Recent findings in the epidemiologic evidence, classifica-
typing scheme for the opportunistic pathogen Propionibacterium acnes tion, and subtypes of acne vulgaris. J Am Acad Dermatol 39:S34–7
and characterization of type I cell surface-associated antigens. Zerbino DR, Birney E (2008) Velvet: algorithms for de novo short read
Microbiology 157:1990–2003 assembly using de Bruijn graphs. Genome Res 18:821–9

www.jidonline.org 9