Journal of Medical Microbiology (2014), 63, 721728

DOI 10.1099/jmm.0.067611-0

Relationship between the severity of acne vulgaris

and antimicrobial resistance of bacteria isolated
from acne lesions in a hospital in Japan
Keisuke Nakase,1 Hidemasa Nakaminami,1 Yuko Takenaka,2
Nobukazu Hayashi,3 Makoto Kawashima2 and Norihisa Noguchi1
Correspondence
Norihisa Noguchi
noguchin@toyaku.ac.jp

Department of Microbiology, School of Pharmacy, Tokyo University of Pharmacy and Life Sciences,
1432-1 Horinouchi, Hachioji, Tokyo 192-0392, Japan

Department of Dermatology, Tokyo Womens Medical University, 8-1 Kawadachou, Shinjuku-ku,

Tokyo 162-8666, Japan

Department of Dermatology, Toranomon Hospital, 2-2-2 Toranomon, Minato-ku, Tokyo 105-8470,

Japan

Received 30 August 2013

Accepted 11 February 2014

Propionibacterium acnes and Staphylococcus epidermidis are normal skin inhabitants that are
frequently isolated from lesions caused by acne, and these micro-organisms are considered to
contribute to the inflammation of acne. In the present study, we examined the antimicrobial
susceptibilities and resistance mechanisms of P. acnes and S. epidermidis isolated from patients
with acne vulgaris in a university hospital in Japan from 2009 to 2010. Additionally, we analysed
the relationship between the antimicrobial resistance of P. acnes and the severity of acne vulgaris.
Some P. acnes strains (18.8 %; 13/69) were resistant to clindamycin. All strains had a mutation in
the 23S rRNA gene, except for one strain that expressed erm(X) encoding a 23S rRNA
methylase. Tetracycline-resistant P. acnes strains were found to represent 4.3 % (3/69) of the
strains, and this resistance was caused by a mutation in the 16S rRNA gene. Furthermore, three
strains with reduced susceptibility to nadifloxacin (MIC516 mg ml1) were detected. When
analysing the correlation between the antimicrobial resistance of P. acnes and S. epidermidis,
more than 80 % of the patients who carried clindamycin-resistant P. acnes also carried
clindamycin-resistant S. epidermidis. However, no epidemic strain that exhibited antimicrobial
resistance was detected in the P. acnes strains when analysed by PFGE. Therefore, our results
suggest that the antimicrobial resistance of P. acnes is closely related to antimicrobial therapy.
Additionally, those P. acnes strains tended to be frequently found in severe acne patients rather
than in mild acne patients. Consequently, the data support a relationship between using
antimicrobial agents and the emergence of antimicrobial resistance.

INTRODUCTION
Propionibacterium acnes and Staphylococcus epidermidis,
which are normal inhabitants of the skin, were isolated
from acne vulgaris inflammatory sites of patients (Kim
et al., 2008; Nishijima et al., 2000). P. acnes is regarded as a
cause of the pathogenesis and exacerbation of acne
vulgaris, and antimicrobial therapy for P. acnes is provided
for acne vulgaris cases (Iinuma et al., 2009). Recently,
global reports have shown that the antimicrobial susceptibility varies for P. acnes isolated from patients with acne
vulgaris. In Europe and the United States, antimicrobialresistant P. acnes strains have been isolated frequently and
have become a large problem (Leyden, 2004; Ross et al.,
Abbreviations: MLSB, macrolides-lincosamides-streptogramin B; MRSE,
meticillin-resistant Staphylococcus epidermidis.

067611 G 2014 The Authors

2003; Thiboutot et al., 2009). A previous study that

investigated the antimicrobial susceptibilities of P. acnes
and S. epidermidis isolated from acne patients in a university hospital in Japan from 1996 to 1997 revealed few
resistant strains of P. acnes, and that S. epidermidis showed
a high susceptibility to antimicrobials (Nishijima et al.,
2000). In addition to the classical oral minocycline doxycycline, which has been approved for the treatment of acne,
the oral macrolides roxithromycin and clarithromycin,
topical nadifloxacin and topical clindamycin were approved
in 1991, 1993 and 2002, respectively, in Japan. These oral
and topical antimicrobial agents for acne therapy are
commonly used, and the number of clindamycin-resistant
strains is increasing (Nakase et al., 2012). Lincosamideresistant strains are known as macrolide-lincosamidestreptogramin B (MLSB)-resistant and show cross-resistance

Downloaded from www.microbiologyresearch.org by

IP: 93.91.26.97
On: Thu, 29 Oct 2015 13:56:47

Printed in Great Britain

721

K. Nakase and others

to macrolides and streptogramin B by binding to an

identical region of the 23S rRNA (Ross et al., 2002). Two
mechanisms of resistance of MLSB in P. acnes are known.
One mechanism involves an rRNA methylase, which is
encoded by erm(X), that modifies the 23S rRNA of the 50S
subunit and contributes to the high-level of MLSB resistance.
The other mechanism involves mutations in the peptidyltransferase region of the 23S rRNA, which results in different
MLSB resistance characteristics according to the point
mutation (Ross et al., 2003). Fluoroquinolone-resistant P.
acnes has not been reported, but cases of tetracyclineresistant P. acnes caused by mutation of the 16S rRNA have
been reported (Oprica et al., 2005; Ross et al., 2002).
Although S. epidermidis functions on the skin barrier,
multi-drug-resistant strains carry multiple resistance determinants, e.g. meticillin-resistant S. epidermidis (MRSE),
which has an acquired meticillin resistance gene (mecA)
(Hisata et al., 2005; Otto, 2009). S. epidermidis causes an
opportunistic infection and has the potential to become a
reservoir of antimicrobial resistance genes (Bisno, 1984). A
previous study investigated the antimicrobial susceptibilities of P. acnes and S. epidermidis isolated from patients
with acne vulgaris in a university hospital in Japan from
1996 to 1997 (Nishijima et al., 2000). At that time, no
antimicrobial-resistant strains were found for P. acnes, and
S. epidermidis showed a high susceptibility to antimicrobials. In the present study, we examined the antimicrobial
susceptibilities and resistance mechanisms of P. acnes and
S. epidermidis isolated from patients with acne vulgaris in a
university hospital in Japan from 2009 to 2010.

METHODS
Bacterial strains and growth conditions. Sixty-nine P. acnes and
58 S. epidermidis strains were collected from 91 patients (ratio of
males to females, 18 : 73; mean age, 25.5 years) with acne vulgaris at
the Tokyo Womens Medical University in Japan from 2009 to 2010.
P. acnes ATCC 11828, S. epidermidis ATCC 14990, Staphylococcus
aureus ATCC 29213 and S. aureus N315 were used as quality control
strains for antimicrobial susceptibility testing. P. acnes ATCC 11828
was used as a wild-type strain for 23S rRNA and 16S rRNA (GenBank
accession no. CP003084). The samples were spread onto mannitol
salt agar (MSA: Oxoid) and modified GAM agar (GAM; Nissui
Pharmaceutical). The MSA was then incubated under aerobic
conditions at 35 uC for 24 h, and the GAM was incubated under
anaerobic conditions at 35 uC for 72 h. The bacterial strains that grew
on MSA and GAM were identified using API Staph (bioMerieux) and
API 20A (bioMerieux), respectively. Typing of P. acnes was performed
using the method described by Shannon et al. (2006).

manufacturers. Amoxicillin, oxacillin sodium, cefaclor, cefdinir,

levofloxacin hydrochloride, erythromycin, clarithromycin, roxithromycin, clindamycin hydrochloride, doxycycline hyclate, minocycline
hydrochloride and fusidic acid sodium were purchased from SigmaAldrich. Ciprofloxacin hydrochloride, gentamicin sulphate and chloramphenicol were purchased from Wako Pure Chemical Industries.
The breakpoints for the antimicrobial agents were defined using the
MIC values for the quality control strains and the CLSI criteria (CLSI,
2011, 2012; Nakaminami et al., 2008; Nakase et al., 2012).
Determination of antimicrobial resistance mechanisms in P.
acnes. The detection of erm(X) using PCR was performed as

previously described (Ishida et al., 2008; Nakase et al., 2012), and the
sequences of the amplified erm(X) genes were verified by DNA
sequencing. The presence of a 23S rRNA mutation was determined
using DNA sequencing as previously described (Nakase et al., 2012).
The presence of a mutation in the 16S rRNA gene that causes
tetracycline resistance was also determined using DNA sequencing.
The primers 8UA (59-AGAGTTTGATCMTGGCTCAG-39) and 1485B
(59-TACGGTTACCTTGTTACGAC-39) were used to amplify the 16S
rRNA gene (Noguchi et al., 2011). The primers acnes 16S_seq-F (59GAGCGAACAGGCTTAGATAC-39) and acnes 16S_seq-R (59-TCGGGTGTTACCAACTTTCA-39) were designed and used as sequencing
primers.
Detection of antimicrobial resistance genes in S. epidermidis.

The detection of antimicrobial resistance genes [mecA, ermA, ermB,

ermC, msrA/B, mphC, tetK, tetM, aac(69)aph(29), aph(39), aad(49,40)
and ant(49)] in S. epidermidis was performed as previously described
(de Vries et al., 2009; Gomez-Sanz et al., 2011; Lina et al., 1999;
Nakaminami et al., 2008). SCCmec typing was performed according
to the method of Boye et al. (2007).
PFGE analysis for P. acnes. PFGE analysis of P. acnes was

performed using a modification of the method described by Oprica

et al. (2004) that utilized the SpeI restriction enzyme. The samples
were lysed using 12.5 g lysozyme l21 (Wako) for 6 h and loaded onto
a 1 % (w/v) pulsed-field-certified agarose gel (Bio-Rad) with the
following running conditions: initial switch time, 2.16 s; final switch
time, 8.53 s; run time, 24 h; angle, 120u; gradient, 6.0 V cm21;
temperature, 14 uC; and ramping factor, linear. The resulting PFGE
patterns were analysed using Bio Numerics version 6.01 (Applied
Maths), and dendrograms were created using the unweighted pair
group method of averages (UPGMA) calculated according to the
method of Dice (band tolerance, 1.5 %; optimization, 1.5 %)
(Noguchi et al., 2011).
Statistical analysis. Differences in the severity grade of acne vulgaris

and patient age were compared using Students t-test. A chi-squared

test was used to compare the severity grade of acne vulgaris and the
percentages of MLSB resistance. The odds ratios for these data were
calculated using the statistical analysis software JMP (SAS Institute)
with a 95 % confidence interval.

RESULTS

Susceptibility testing. The antimicrobial susceptibility of P. acnes

strains was determined by a dilution procedure using brucella agar

(Becton Dickinson) containing 5 % (v/v) lysed defibrinated sheeps
blood (Nippon Bio-Test Laboratories) (CLSI, 2011, 2012). The MIC
was determined after incubation under anaerobic conditions at 35 uC
for 48 h. The antimicrobial susceptibility of S. epidermidis strains was
determined using the agar dilution procedure with MuellerHinton
agar (Oxoid) (CLSI, 2009). The MIC was determined after incubation
under aerobic conditions at 35 uC for 24 h. Cefditoren sodium,
faropenem sodium and nadifloxacin were kindly provided by their
722

Patient histories
Of the 91 patients, 18 were male (1435 years; mean age,
22.4 years), and 73 were female (1847 years; mean age,
26.2 years).
The severity of acne vulgaris in patients was determined
according to the Japanese Dermatological Association acne
vulgaris treatment guideline, and two (2.2 %) and 14

Downloaded from www.microbiologyresearch.org by

IP: 93.91.26.97
On: Thu, 29 Oct 2015 13:56:47

Journal of Medical Microbiology 63

Antimicrobial-resistant P. acnes in Japan

patients (15.4 %) were classified as having very severe and

severe acne vulgaris, respectively (Fig. 1a) (Hayashi et al.,
2008). A comparative analysis showed that the mean age of
the patients with severe acne vulgaris (23.3 years) was
significantly lower than that of the patients with mild acne
vulgaris (26.9 years, P50.042) (Fig. 1b). The duration of
acne vulgaris ranged from three days to 22 years. P. acnes
and S. epidermidis were isolated from 69 patients (75.8 %)
and 58 patients (63.7 %), respectively. Within a year, 59 of
the patients (64.8 %) were treated with antimicrobial
agents. Clindamycin, fluoroquinolones, tetracyclines and
macrolides were used on 39 (66.1 %), 22 (37.3 %), 18
(30.5 %) and 15 (25.4 %) of the patients, respectively.

Analysis of antimicrobial resistance mechanisms

Antimicrobial susceptibility

In S. epidermidis, various antimicrobial resistance genes

were detected (Table 3). When SCCmec typing was
performed in 19 strains that carried mecA, types I, IV
and V were detected in seven (36.8 %), 10 (52.6 %) and two
(10.5 %) of the strains, respectively (Table 4). At least one
of the MLSB resistance genes was found in 35 strains
(60.3 %). The tetracycline resistance genes tetK and tetM
were detected in nine (15.5 %) and four (6.9 %) strains,
respectively. The aminoglycosides resistance gene aac(69)
aph(29) (aacAaphD) was detected in 15 strains (25.9 %).

We determined the MLSB resistance mechanisms using 23S

rRNA gene sequencing and the detection of erm(X) for 16
MLSB-resistant P. acnes strains (Table 2). Furthermore, we
determined the tetracycline resistance mechanism using
16S rRNA gene sequencing. The erm(X) gene was detected
in one strain, and this strain showed high-level MLSB
resistance. The A2058G mutation in the 23S rRNA was
detected in 11 strains (68.8 %), and this mutation was
predominant in MLSB-resistant P. acnes. Despite resistance
to clarithromycin, no MLSB resistance determinant was
detected in strain 68b. The G1058C mutation in the 16S
rRNA was detected in three strains that had reduced
doxycycline susceptibility.

Antimicrobial susceptibility testing was performed on 69 P.

acnes and 58 S. epidermidis isolates that were collected from
91 patients with acne vulgaris (Table 1). All P. acnes strains
were susceptible to b-lactams, and reduced susceptibilities
(MIC516 mg ml21) to nadifloxacin (fluoroquinolone) and
doxycycline (tetracycline) were found in three strains
(4.3 %). The MIC90 of the macrolides and clindamycin was
64 mg ml21, and 16 strains (23.2 %) were resistant to
those agents.
In S. epidermidis, b-lactams, macrolides, clindamycin and
gentamicin showed high MIC90 values, and strains were
found that were resistant to all of the antimicrobial
agents. Especially, oxacillin-, levofloxacin-, gentamicinand MLSB-resistant strains were found in 16 (27.6 %), 21
(36.2 %), 11 (19.0 %) and 34 (58.6 %) strains, respectively.

PFGE analysis for P. acnes

To analyse the relatedness of the strains, the molecular
epidemiological characteristics of P. acnes isolated from
patients with acne vulgaris were determined using PFGE
analysis (Fig. 2). Additionally, the susceptibility of the

*
(b) 30

**

50

25

80
Very severe
60
Severe
40
Moderate

40

20
30
15
20
10
10

20
Mild
0

Mean age of patients

Percentage of patients

Unknown

0
Mild

Previous
report

This study

Moderate

Severe

Percentage of patients with MLSB-R strain

(a) 100

Severity grade

Fig. 1. Patient background analysis. (a) Comparison of the severity grades between this study (n591) and a previous study
(n5343) (Nishijima et al., 2000). (b) Relationship between severity grades and the mean age of patients and the possession
rate of MLSB-resistant (MLSB-R) strains. The severe grade contains patients with severe and very severe acne vulgaris.
*Significant difference in the mean age of patients with mild and severe cases of acne vulgaris, using Students t-test
(P50.042). **Significant difference in the percentage of patients with a MLSB-resistant strain in moderate and severe cases of
acne vulgaris, using a chi-squared test (P50.038). The others were not significantly different.
http://jmm.sgmjournals.org

Downloaded from www.microbiologyresearch.org by

IP: 93.91.26.97
On: Thu, 29 Oct 2015 13:56:47

723

K. Nakase and others

Table 1. Antimicrobial susceptibility of P. acnes and S. epidermidis isolated from patients with acne vulgaris
Antimicrobial agent

Amoxicillin
Oxacillind
Cefaclor
Cefditoren
Faropenem
Levofloxacin
Nadifloxacin
Erythromycin
Clarithromycin
Spiramycin
Clindamycin
Gentamicin
Doxycycline
Minocycline
Fusidic acid

P. acnes (n569)

S. epidermidis (n558)

MIC range (mg ml1)

MIC90*

Resistance (%)D

0.060.25

0.52
0.060.25
0.06
0.2532
0.2516
0.06256
0.06256
0.06256
0.06256
0.2532
0.1316
0.068
0.068

0.13

2
0.13
0.06
1
0.5
256
256
64
128
32
1
0.5
8

0
0
0
4.3

23.2
23.2
21.7
18.8

4.3
0
0

MIC range (mg ml1)

MIC90*

Resistance (%)D

4
16
32
4
0.5
4
2
256
256
256
256
16
8
0.5
0.06

0.0632
0.06256
0.06256
0.13128
0.06256
0.13128
0.0632
0.06256
0.06256
0.5256
0.06256
0.06128
0.0616
0.0616
0.068

25.9
27.6
3.4
5.2
1.7
36.2
10.3
58.6
58.6
50.0
51.7
19.0
8.6
5.2
1.7

*MIC90 values indicate the MICs (mg ml21) that inhibit the growth of 90 % of strains.
DResistance breakpoints of nadifloxacin and gentamicin were not established in P. acnes. Resistance breakpoints of the following antimicrobial
agents were defined according to a previous study: spiramycin, 8 mg ml21; minocycline, 16 mg ml21; fusidic acid, 32 mg ml21 for P. acnes.
dSusceptibility to oxacillin was not measured in P. acnes.

strains to MLSB and the severity grade of acne vulgaris in

the patients were compared. We did not obtain a clear
PFGE result for one of the strains; thus, the data were
omitted from the present study.

Several strains showed identical PFGE patterns. However,

no common background was observed among patients
carrying P. acnes with identical PFGE patterns. The strains
were classified into two groups (A and B) according to the

Table 2. Analysis of resistance determinants of MLSB and tetracyclines in P. acnes

Strain
no.

6
9
11
13b
24b
28b
29
34b
37b
43b
45
62b
68b
71b
77b
90b

MIC (mg ml1)*

Year

2009
2009
2009
2009
2009
2009
2009
2009
2009
2009
2009
2009
2009
2009
2009
2010

Erythromycin

Spiramycin

Clindamycin

Doxycycline

(14-Mac)

(16-Mac)

(Linco)

(Tetracyclines)

256
256
256
256
256
256
256
256
256
256
256
256
256
256
256
256

32
32
32
256
256
128
128
128
256
64
64
32
1
64
256
64

64
128
128
256
64
128
128
128
8
128
32
32
0.25
128
0.5
0.5

0.13
16
16
0.25
0.13
0.13
0.13
16
0.25
0.25
1
0.13
0.13
0.13
2
0.25

erm(X)

23S rRNA
mutation

16S rRNA
mutation

A2058G
A2058G
A2058G
WT
A2058G
A2058G
A2058G
A2058G
A2059G
A2058G
A2058G
A2058G
WT
A2058G
A2059G
A2059G

WT
G1058C
G1058C
WT
WT
WT
WT
G1058C
WT
WT
WT
WT
WT
WT
WT
WT

*14-Mac, 14-membered macrolides; 16-Mac, 16-membered macrolides; Linco, lincosamides.

724

Downloaded from www.microbiologyresearch.org by

IP: 93.91.26.97
On: Thu, 29 Oct 2015 13:56:47

Journal of Medical Microbiology 63

Antimicrobial-resistant P. acnes in Japan

Table 3. Detection of antimicrobial resistance genes in S.

epidermidis (n558)
Gene

Percentage of strains (n)

mecA
ermA
ermB
ermC
msrA/B
mphC
tetK
tetM
aacAaphD
aph (39)
aad (49,40)
ant (49)

32.7 (19)
3.4 (2)
0
51.7 (30)
8.6 (5)
5.2 (3)
15.5 (9)
6.9 (4)
25.9 (15)
20.9 (12)
5.2 (3)
5.2 (3)

PFGE pattern. Type IA and IB strains were classified into

group A, and type II strains were classified into group B.
Whereas all strains with the A2058G mutation in the 23S
rRNA were placed in PFGE group A, all strains with the
A2059G mutation in the 23S rRNA were placed in PFGE
group B. Furthermore, the number of strains in PFGE
group A (13 strains) that were isolated from patients with
severe acne vulgaris was greater than that in PFGE group B
(one strain). The severity of acne vulgaris and isolation rate
of MLSB-resistant strains was compared (Fig. 1b). The
isolation rates of the MLSB-resistant strains were 40.0 %
(6/15), 12.9 % (4/31) and 18.8 % (3/16) in the patients with
severe, moderate and mild acne vulgaris, respectively.

DISCUSSION
P. acnes and S. epidermidis have been primarily isolated
from acne vulgaris patients; however, we found few reports
discussing sampling of those bacteria from acne patients.
The pathogenesis of acne vulgaris is also affected by the
balance of hormones, which tends to be worse during
adolescence when the secretion of hormones increases.
Thus, pathogenesis involves bacterial and host biological
effects (Dipiro et al., 2005).
Because the subjects of the study were outpatients with
acne vulgaris who were in the university hospital, this study
had a large number of intractable cases of patients who had
already been treated with an antimicrobial agent. The
isolation rate of MLSB-resistant P. acnes in severe acne
Table 4. SCCmec type of S. epidermidis (n519)
SCCmec
Type
Type
Type
Type

I
II
IV
V

http://jmm.sgmjournals.org

Percentage of strains (n)

36.8 (7)
0
52.6 (10)
10.5 (2)

patients (40.0 %) was higher than that of mild acne patients

(18.8 %), although no significant difference was found
(P50.193). No significant difference was found between
the ages of patients (25.4, n516) carrying MLSB-resistant
P. acnes and those (24.7, n553) carrying susceptible P.
acnes (P50.34). The mean age of the severe acne patients
was significantly lower than that of mild acne patients in
the present study. These data were different from those of a
previous report (Schafer et al., 2013); Japanese acne
patients who feel the stress of acne vulgaris are believed
to usually visit a doctor in their adolescent years. In the
present study, no significant differences were found
between severity grades and treatment histories of acne
vulgaris or between treatment history with macrolides and
the isolation rate of MLSB-resistant P. acnes.
Reduced antimicrobial susceptibility of P. acnes isolated
from patients with acne vulgaris has been found throughout the world (El-Mahdy et al., 2010; Mendoza et al.,
2013). Nishijima et al. (2000) found 30.8 % MLSB-resistant
S. epidermidis strains but no P. acnes strains. In the present
study, MLSB-resistant strains were 23.2 % and 58.6 % P.
acnes and S. epidermidis, respectively. Therefore, our data
showed that the rate of acquiring antimicrobial resistance
has increased in bacteria isolated from acne lesions.
In Japan, P. acnes that carries erm(X), which is involved in
high-level MLSB resistance, was first detected in 2008
(Nakase et al., 2012). The 23S dimethylase encoded by
erm(X) causes dimethylation of the peptidyltransferase
region of domain V of the 23S rRNA, which decreases drug
affinity. The prevalence of the erm(X) gene causes serious
clinical concerns because P. acnes strains that carry erm(X)
exhibit high-level resistance to oral macrolides and topical
clindamycin. The transposon Tn5432, which contains
erm(X), was first found in Corynebacterium; and the only
other bacterial species that Tn5432 has been found in is P.
acnes (Ross et al., 2002). Although erm(X) was detected in
three strains, two of the strains were omitted from the
present study because the strains were identified as
Propionibacterium granulosum and Propionibacterium avidum by 16S rRNA gene sequencing (data not shown).
Although P. acnes strains that carry erm(X) have been
reported elsewhere, the percentage of those strains among
all MLSB-resistant strains was low. The percentage of P.
acnes strains that carry erm(X) may be further decreased by
the accurate identification of bacterial species using DNA
sequencing. We presumed that although erm(X) results in
high-level MLSB resistance, this resistance determinant may
be difficult to maintain on the chromosome of P. acnes.
Our data showed that most MLSB-resistant P. acnes strains
had a 23S rRNA mutation, and the A2058G mutation was
the most frequent for MLSB resistance. Additionally, we
found one MLSB-resistant strain without the resistance
determinant. Further investigation is necessary to understand the unknown mechanism of MLSB resistance.
Nadifloxacin is widely used in the world. However, no
nadifloxacin-resistant strains have been found in P. acnes,

Downloaded from www.microbiologyresearch.org by

IP: 93.91.26.97
On: Thu, 29 Oct 2015 13:56:47

725

K. Nakase and others

Similarity (%)

PFGE pattern

MLSB

Severity

*
**
**

Type

PFGE group

50 60 70 80 90 100

**
**
**

*
*
*
*
*

IA
**
**

A
*

**
**
**

**

*
*

*
IB
*
*

***
****

****

II

Fig. 2. PFGE analysis for P. acnes. For MLSB-resistant strains: *strain carrying erm(X); **strains with the A2058G mutation of
the 23S rRNA gene; ***strain without MLSB resistance determinants; ****strains with the A2059G mutation of the 23S rRNA
gene. For severity determination: *strains isolated from patients with severe and very severe acne vulgaris.

and strains with reduced susceptibility (MIC516 mg ml21)

were first found in this study. Topical drugs can administer
a high concentration of the antimicrobial onto skin
surfaces because the drugs are directly applied. If topical
nadifloxacin (1 %) is applied by one finger-tip unit to the
skin at 5 cm2, the amount of nadifloxacin is expected to be
approximately 600 mg cm22 (Long et al., 1998). While
topical clindamycin is also expected to achieve similar
concentrations, it is possible that an insufficient therapeutic effect will be obtained in the presence of high-level
clindamycin-resistant P. acnes strains that carry erm(X).
Therefore, topical nadifloxacin is considered to have a
sufficient bactericidal effect against P. acnes with reduced
fluoroquinolone susceptibility. Mutations in DNA gyrase
726

and/or DNA topoisomerase are known to be major

mechanisms of fluoroquinolone resistance; however, those
mechanisms have not been observed in P. acnes. Further
analysis is necessary to discover the fluoroquinolone
resistance factors in P. acnes.
Resistance to tetracyclines is caused by increasing the
activity of an efflux pump, ribosomal protection and
mutation of drug binding points (Ross et al., 2003). The
present study found the first strains with the G1058C
mutation of the 16S rRNA gene in P. acnes in Japan. When
the histories of acne treatments were verified, 77.8 % (7/9)
of the patients carrying P. acnes with reduced tetracycline
susceptibility and 81.3 % (13/16) of the patients carrying

Downloaded from www.microbiologyresearch.org by

IP: 93.91.26.97
On: Thu, 29 Oct 2015 13:56:47

Journal of Medical Microbiology 63

Antimicrobial-resistant P. acnes in Japan

MLSB-resistant P. acnes had used the respective antimicrobial

agents (data not shown). Additionally, more than 80 % of the
patients who carried MLSB-resistant P. acnes also carried
MLSB-resistant S. epidermidis (data not shown). Our data
strongly suggest that the 16S rRNA mutation for tetracycline
resistance and the 23S rRNA mutation for macrolide
resistance in P. acnes were caused by the use of antimicrobial
agents. Therefore, minocycline is considered to have a more
effective therapeutic effect than oral macrolides for acne
vulgaris because no P. acnes strains were resistant to this drug
in this study or in previous reports.
PFGE analysis suggested that the mutation points of the
23S rRNA in P. acnes were different according to the
genetic background. Furthermore, nearly all strains that
were isolated from patients with severe acne vulgaris
belonged to the same PFGE group. Additionally, a clear
relationship was found between the severity grade of acne
vulgaris and the MLSB resistance rate of P. acnes. Our data
showed that the antimicrobial resistance of P. acnes was
closely related to the severity of acne vulgaris.
Understanding the trend of antimicrobial susceptibilities of
S. epidermidis isolated from patients with acne vulgaris is
valuable because few studies have been reported. In addition
to Staphylococcus aureus, S. epidermidis is one of the normal
inhabitants of the skin and easily acquires and transmits the
various antimicrobial resistance genes (Bloemendaal et al.,
2010). In our study, SCCmec type II, which is primarily
detected in hospital-acquired strains, was not found in S.
epidermidis strains carrying mecA. MLSB- and aminoglycoside-resistance genes were detected in 60.2 % and 29.2 % of
strains, respectively. The prevalence of those antimicrobial
resistance genes in S. epidermidis may be caused by the use of
clindamycin and gentamicin in treating skin and soft tissue
staphylococci infections. Although the pathogenicity of S.
epidermidis for acne vulgaris remains unclear, S. epidermidis
could become a reservoir of various antimicrobial resistance
genes on the skin.
In conclusion, the existence of antimicrobial-resistant
bacteria should be considered when selecting antimicrobial
therapy for inflammatory acne, and antimicrobials should
not be used for the maintenance therapy of acne vulgaris.
Additionally, following an approval of adapalene, benzoyl
peroxide, which is an alternative antimicrobial agent, may
have an important role in the prevention of antimicrobialresistant P. acnes.

Bloemendaal, A. L., Brouwer, E. C. & Fluit, A. C. (2010). Methicillin

resistance transfer from Staphylocccus epidermidis to methicillinsusceptible Staphylococcus aureus in a patient during antibiotic
therapy. PLoS ONE 5, e11841.
Boye, K., Bartels, M. D., Andersen, I. S., Mller, J. A. & Westh, H. (2007).

A new multiplex PCR for easy screening of methicillin-resistant

Staphylococcus aureus SCCmec types I-V. Clin Microbiol Infect 13, 725
727.
CLSI (2009). Methods for Dilution Antimicrobial Susceptibility Test for

Bacteria That Grow Aerobically; Approved standard M07A8.

CLSI (2011). Perfomance Standard for Antimicrobial Susceptibility
Testing; approved standard M100S21.
CLSI (2012). National Committee for Clinical Laboratory Standards.

Methods of Antimicrobial Susceptibility Testing of Anaerobic Bacteria,

8th edn, pp. M11A8. Approved Standard.
de Vries, L. E., Christensen, H., Skov, R. L., Aarestrup, F. M. &
Agers, Y. (2009). Diversity of the tetracycline resistance gene tet(M)

and identification of Tn916- and Tn5801-like (Tn6014) transposons

in Staphylococcus aureus from humans and animals. J Antimicrob
Chemother 64, 490500.
El-Mahdy, T. S., Abdalla, S., El-Domany, R., Mohamed, M. S., Ross,
J. I. & Snelling, A. M. (2010). Detection of a new erm(X)-mediated

antibiotic resistance in Egyptian cutaneous propionibacteria. Anaerobe

16, 376379.
Gomez-Sanz, E., Torres, C., Lozano, C., Saenz, Y. & Zarazaga, M.
(2011). Detection and characterization of methicillin-resistant

Staphylococcus pseudintermedius in healthy dogs in La Rioja, Spain.

Comp Immunol Microbiol Infect Dis 34, 447453.
Hayashi, N., Akamatsu, H., Iwatsuki, K., Kurokawa, I., Kono, T.,
Tanioka, M., Hidaka, R., Furukawa, F., Yamasaki, O. & other authors
(2008). Guidelines for treatment of acne vulgaris. Jpn J Dermatol 118,

18931923 (in Japanese).

Hisata, K., Kuwahara-Arai, K., Yamanoto, M., Ito, T., Nakatomi, Y.,
Cui, L., Baba, T., Terasawa, M., Sotozono, C. & other authors (2005).

Dissemination of methicillin-resistant staphylococci among healthy

Japanese children. J Clin Microbiol 43, 33643372.
Iinuma, K., Sato, T., Akimoto, N., Noguchi, N., Sasatsu, M., Nishijima,
S., Kurokawa, I. & Ito, A. (2009). Involvement of Propionibacterium

acnes in the augmentation of lipogenesis in hamster sebaceous glands

in vivo and in vitro. J Invest Dermatol 129, 21132119.
Ishida, N., Nakaminami, H., Noguchi, N., Kurokawa, I., Nishijima, S. &
Sasatsu, M. (2008). Antimicrobial susceptibilities of Propionibacterium

acnes isolated from patients with acne vulgaris. Microbiol Immunol 52,
621624.
Kim, S. S., Kim, J. Y., Lee, N. H. & Hyun, C. G. (2008). Antibacterial

and anti-inflammatory effects of Jeju medicinal plants against acneinducing bacteria. J Gen Appl Microbiol 54, 101106.
Leyden, J. J. (2004). Antibiotic resistance in the topical treatment of

acne vulgaris. Cutis 73 (Suppl), 610.

Lina, G., Quaglia, A., Reverdy, M. E., Leclercq, R., Vandenesch, F. &
Etienne, J. (1999). Distribution of genes encoding resistance to

ACKNOWLEDGEMENTS
This work was supported by the Matching Fund Subsidy for the
Private Schools of Japan. We would like to thank Misa Arima and
Arisa Kawarai for help with the bacteriological experiments.

macrolides, lincosamides, and streptogramins among staphylococci.

Antimicrob Agents Chemother 43, 10621066.
Long, C. C., Mills, C. M. & Finlay, A. Y. (1998). A practical guide to

topical therapy in children. Br J Dermatol 138, 293296.

Mendoza, N., Hernandez, P. O., Tyring, S. K., Haitz, K. A. & Motta, A.
(2013). Antimicrobial susceptibility of Propionibacterium acnes isolates

REFERENCES

from acne patients in Colombia. Int J Dermatol 52, 688692.

Bisno, A. L. (1984). Cutaneous infections: microbiologic and epide-

Nakaminami, H., Noguchi, N., Ikeda, M., Hasui, M., Sato, M.,
Yamamoto, S., Yoshida, T., Asano, T., Senoue, M. & Sasatsu, M.
(2008). Molecular epidemiology and antimicrobial susceptibilities of 273

miologic considerations. Am J Med 76 (5A), 172179.

http://jmm.sgmjournals.org

Downloaded from www.microbiologyresearch.org by

IP: 93.91.26.97
On: Thu, 29 Oct 2015 13:56:47

727

K. Nakase and others

exfoliative toxin-encoding-gene-positive Staphylococcus aureus isolates
from patients with impetigo in Japan. J Med Microbiol 57, 12511258.
Nakase, K., Nakaminami, H., Noguchi, N., Nishijima, S. & Sasatsu,
M. (2012). First report of high levels of clindamycin-resistant

Propionibacterium acnes carrying erm(X) in Japanese patients with

acne vulgaris. J Dermatol 39, 794796.
Nishijima, S., Kurokawa, I., Katoh, N. & Watanabe, K. (2000). The

bacteriology of acne vulgaris and antimicrobial susceptibility of

Propionibacterium acnes and Staphylococcus epidermidis isolated from
acne lesions. J Dermatol 27, 318323.
Noguchi, N., Nakaminami, H., Nakase, K. & Sasatsu, M. (2011).

Characterization of enterococcus strains contained in probiotic

products. Biol Pharm Bull 34, 14691473.
Oprica, C., Emtestam, L., Lapins, J., Borglund, E., Nyberg, F.,
Stenlund, K., Lundeberg, L., Sillerstrom, E. & Nord, C. E. (2004).

Antibiotic-resistant Propionibacterium acnes on the skin of patients

with moderate to severe acne in Stockholm. Anaerobe 10, 155
164.
Oprica, C., Lofmark, S., Lund, B., Edlund, C., Emtestam, L. & Nord, C.
E. (2005). Genetic basis of resistance in Propionibacterium acnes

strains isolated from diverse types of infection in different European

countries. Anaerobe 11, 137143.
Otto, M. (2009). Staphylococcus epidermidisthe accidental pathogen.

Nat Rev Microbiol 7, 555567.

728

Ross, J. I., Eady, E. A., Carnegie, E. & Cove, J. H. (2002). Detection of

transposon Tn5432-mediated macrolide-lincosamide-streptogramin

B (MLSB) resistance in cutaneous propionibacteria from six
European cities. J Antimicrob Chemother 49, 165168.
Ross, J. I., Snelling, A. M., Carnegie, E., Coates, P., Cunliffe, W. J.,
Bettoli, V., Tosti, G., Katsambas, A., Galvan Perez Del Pulgar, J. I. &
other authors (2003). Antibiotic-resistant acne: lessons from Europe.

Br J Dermatol 148, 467478.

Schafer, F., Fich, F., Lam, M., Garate, C., Wozniak, A. & Garcia, P. (2013).

Antimicrobial susceptibility and genetic characteristics of Propionibacterium acnes isolated from patients with acne. Int J Dermatol 52, 418425.
Shannon, B. A., Cohen, R. J. & Garrett, K. L. (2006). Polymerase chain

reaction-based identification of Propionibacterium acnes types isolated

from the male urinary tract: evaluation of adolescents, normal adults
and men with prostatic pathology. BJU Int 98, 388392.
Thiboutot, D., Gollnick, H., Bettoli, V., Dreno, B., Kang, S., Leyden,
J. J., Shalita, A. R., Lozada, V. T., Berson, D. & other authors (2009).

New insights into the management of acne: an update from the

Global Alliance to Improve Outcomes in Acne group. J Am Acad
Dermatol 60 (Suppl), S1S50.
West, D. P., West, L. E., Letizia, M. & Micali, G. (2005). Acne vulgaris.

In Pharmacotherapy: A Pathophysiologic Approach, 6th edn. pp. 1755

1768. Edited by J. T. Dipiro, R. L. Talbert, G. C. Yee, G. R. Matzke,
B. G. Wells & L. M. Posey. New York: McGraw-Hill.

Downloaded from www.microbiologyresearch.org by

IP: 93.91.26.97
On: Thu, 29 Oct 2015 13:56:47

Journal of Medical Microbiology 63