Download A newly discovered Anaerococcus strain responsible for axillary

RESEARCH ARTICLE
A newly discovered Anaerococcus strain responsible for axillary
odor and a new axillary odor inhibitor, pentagalloyl glucose
Takayoshi Fujii, Junko Shinozaki, Takayuki Kajiura, Keiji Iwasaki & Ryosuke Fudou
Frontier Research Laboratories, Institute for Innovation, Ajinomoto Co. INC., Kanagawa, Japan
Correspondence: Takayoshi Fujii, Frontier
Research Laboratories, Institute for
Innovation, Ajinomoto Co. INC., 1-1 Suzukicho, Kawasaki-ku, Kawasaki-shi 210-8681,
Japan. Tel.: +81 44 244 7181;
fax: +81 44 244 4757;
e-mail: [email protected]
Received 9 January 2014; revised 24 March
2014; accepted 19 April 2014. Final version
published online 14 May 2014.
DOI: 10.1111/1574-6941.12347
MICROBIOLOGY ECOLOGY
Editor: Alfons Stams
Keywords
skin microbiota; axillary odor; 3-hydroxy-3metyl-hexanoic acid; Anaerococcus;
pentagalloylglucose.
Abstract
Skin surface bacteria contribute to body odor, especially axillary odor. We
aimed to investigate anaerobic bacteria that had not been previously studied
for axillary odor formation. A new anaerobic Anaerococcus sp. A20, that
releases 3-hydroxy-3-metyl-hexanoic acid (HMHA, main component of axillary
odor) from its glutamyl conjugate, was discovered from axillary isolates. This
strain showed strong resistance to the antimicrobial agents, triclosan and 4-isopropyl-3-methylphenol; therefore, we screened plant extracts that inhibit the
A20 strain. We discovered that pentagalloyl glucose (PGG) extracted from the
Chinese Gall plant exhibited both antibacterial and inhibitory activities against
HMHA release by the A20 strain. As the excellent antibacterial activity and
inhibitory effect of PGG against HMHA release were seen in vitro, we conducted an open study to evaluate the deodorant effects of PGG on axillary
odor. The sensory tests on odor strength showed that application of the PGG
solution could reduce axillary odors in vivo. Although there was a small change
in axillary microbiota, the microbial count of A20 significantly reduced. These
results strongly indicate PGG as a new innovative deodorant material that only
affects odor-releasing bacteria in the axillary microbiota.
Introduction
The human body harbors a complex microbiota. Depending on the location of skin, the structure of skin microbiota is clearly different because of the different types of
sweat glands such as the eccrine, apocrine, and sebaceous
glands. One of the important functions of microbiota on
the skin surface is to protect against infections (Coagen
et al., 2008). The main components of the skin microbiota, Staphylococcus epidermidis and Propionibacterium acnes, interact with each other to maintain a slightly acidic
skin pH, thereby protecting the skin from infectious pathogenic bacteria (Coagen et al., 2008).
However, skin surface bacteria also contribute to body
odor; the role of bacteria in axillary odor has been
reported since the 1950s (Shelley et al., 1953). Zeng
et al. (1991) showed that the main component of axillary odor was 3-methyl-2-hexenoic acid (3M2H), derived
from symbiotic Corynebacterium species living on the
secretory products of the many apocrine sweat glands
present in the axillary region. Natsch et al. (2003)
showed that 3-hydroxy-3-methyl-hexanoic acid (HMHA)
ª 2014 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
is another major contributor to axillary odor and has a
structure similar to that of 3M2H. This fatty acid derivative is released from precursor, which is conjugates of
HMHA with L-glutamine, by the action of bacterial Naacylglutamine aminoacylase (Natsch et al., 2003, 2006).
In addition, Taylor et al. (2003) found that there was a
clear correlation between the Corynebacterium count and
odor strength in the axillary region. Similar to mediumchain (C6-C10) volatile fatty acids (VFA), such as
3M2H and HMHA, shorter-chain (C2-C5) VFAs also
make a major contribution to axillary odor. Apart from
Corynebacterium, other common skin bacteria, such as
Staphylococcus and Microbacterium, also contribute to
the production of short-chain VFAs (James et al.,
2004a, b).
A considerable amount of research has been performed
on the involvement of aerobic bacteria, such as Corynebacterium and Staphylococcus, in axillary odor production,
but research on anaerobic bacteria, such as Propionibacterium, is lacking. Propionibacterium metabolizes lactic acid
and glycerol to produce VFAs such as acetic and propionic acids. As these VFAs make up some of the compoFEMS Microbiol Ecol 89 (2014) 198–207
199
Discovery of new axillary bacteria and new odor inhibitor
nents of axillary odor, Propionibacterium is considered a
potential contributor. Taylor et al. (2003) used conventional culture methods to determine the relationship
between the Propionibacterium count and odor strength
in the axillary region but was unable to establish any correlation. In contrast, Hasegawa et al. (2004) reported that
more sulfurous odor was derived from sweat incubated
under anaerobic conditions rather than under aerobic
conditions, suggesting that the anaerobic bacteria in sweat
contribute to axillary odor. Analysis with a recently developed next-generation sequencing method (without culturing) has confirmed that the skin microbiota contains
Propionibacterium as well as other anaerobic bacteria
(Costello et al., 2009). These findings led us to focus on
investigating the relationship between anaerobic bacteria
and axillary odor by isolating these bacteria from the
axilla in healthy humans and evaluating the HMHArelease ability of each strain. In addition, as part of our
continuing search for an antimicrobial agent against the
newly discovered bacterium that is responsible for axillary
odor, we report a new deodorant substance that did not
damage skin microbiota but only affected bacteria responsible for axillary odor.
Materials and methods
Materials
Triclosan and 4-isopropyl-3-methylphenol (IPMP) were
purchased from Wako Pure Chemical Industries, Ltd
(Osaka, Japan). Pentagalloyl glucose (PGG) was prepared
by hydrolysis of partially purified tannic acid from the
Chinese Gall plant (BREWTAN, SA Ajinomoto Omnichem NV, Louvain-la-Neuve, Belgium) through recrystallization from 2% methyl alcohol aqueous solution. The
purity was > 85% by HPLC analysis. Bacterial strains
were listed in each result table.
Bacterial isolate from the axilla and culture
conditions
For isolation of anaerobic bacteria from axilla, swabscrubbed samples from 12 healthy people (seven men and
five women) were obtained. A sterile cotton swab was
dipped for 30 s in 1 mL phosphate-buffered saline (PBS),
pH 7.0 and used to scrub 21 cm2 of the armpit. Afterward, the cotton swab was suspended for 30 s in PBS.
The sample solutions were spread on BL agar (Nissui
Pharmaceutical Co., Ltd, Osaka, Japan) containing 5%
horse whole blood (Nippon Bio-Supp., Center, Tokyo,
Japan) and were cultured at 37 °C under anaerobic conditions. After overnight incubation, single isolates were
obtained and separated based on the shape of colonies.
FEMS Microbiol Ecol 89 (2014) 198–207
Nine isolates were used to evaluate the ability to release
HMHA from the glutamyl conjugate.
HMHA release from glutamyl conjugate by
isolated bacteria
To evaluate the activity of isolated bacteria, an overnight
culture was harvested by centrifugation and resuspended
to 2.0 9 1010 cells mL1 in a semi-synthetic medium
[per L: 3 g monopotassium phosphate (KH2PO4), 1.9 g
dipotassium phosphate (K2HPO4), 0.2 g yeast extract,
0.2 g magnesium sulfate heptahydrate (MgSO4 9 7H2O),
1.4 g sodium chloride (NaCl), 1 g ammonium chloride
(NH4Cl), 10 mg manganese (II) chloride (MnCl2), 1 mg
iron (III) chloride (FeCl3), and 1 mg calcium chloride
(CaCl2)]. The samples were added 0.5-mm glass beads
(Yasui Kikai, Osaka, Japan) and mechanically disrupted
with a Multi-Beads Shocker (Yasui Kikai) at 2700 r.p.m.
for 90 s. After mechanical disruption, a final concentration of 2 mM 3-hydroxy-3-methyl-hexanoic acid-glutamine (HMHA-Gln; Sanyo Chemical Industries, Ltd,
Kyoto, Japan) was added to the samples. After 24 h of
incubation (36 °C) with shaking at 300 r.p.m., the samples were acidified with 1 M aqueous hydrochloric acid
(final concentration, 0.016 M) and extracted with an
equal volume of methyl tert-butyl ether (MTBE), and the
amount of HMHA released was determined with a GC
353B gas chromatograph (GL Science Inc., Tokyo, Japan)
with flame ionization detection (FID). An InterCapTM
Pure-Wax column (length of 30 m, inner diameter of
0.25 mm, and film thickness of 0.25 lm; GL Science
Inc.) was used, and 4 lL of the MTBE solution was
injected in the split injection mode (split ratio, 10 : 1).
The temperature of the injection port and FID was
270 °C and of the column oven was 220 °C for 10 min.
The amount of HMHA released was expressed as the
released HMHA percentage, which was obtained from the
peak area of each sample divided by the peak area of
2 mM HMHA (same concentration as the initial concentration of HMHA-Gln).
Genomic DNA extraction from the A20 strain
The A20 strain was grown on BL agar (containing 5%
horse whole blood) and was suspended in a 500 lL
extraction buffer (pH 9.0) containing 100 mM Tris-HCl
and 40 mM ethylenediaminetetraacetic acid (EDTA) with
10% sodium dodecyl sulfate (SDS). Glass beads (1.0 g,
0.5 mm; Yasui Kikai) were added, and the samples were
mixed with a Multi-Beads Shocker (Yasui Kikai) at
2700 r.p.m. for 90 s. The samples were incubated at
70 °C for 10 min. After incubation, 500 lL phenol (saturated with TE buffer; Nacalai Tesque Inc., Kyoto, Japan)
ª 2014 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
200
was added to the samples and mixed for 5 s by vortexing.
The samples were centrifuged at 20 400 g for 5 min at
room temperature. The supernatant was transferred to
new 1.5 mL centrifuge tubes, and 400 lL phenol : chloroform : isoamylalcohol (25 : 24 : 1, pH 6.7; Nacalai
Tesque Inc.) was added. The samples were centrifuged at
20 400 g for 5 min at 4 °C, and the supernatant was
transferred to new 1.5-mL centrifuge tubes. Afterward,
3 M sodium acetate (0.1 volumes) and an equal volume
of cold isopropanol were added and mixed gently. DNA
was pelleted by centrifugation at 20 400 g for 30 min at
4 °C and resuspended in a 40 lL TE buffer solution.
16S rRNA gene sequence analyses for the A20
strain
PCR was performed using DNA Engine Tetrad 2 (BioRad Laboratories, Inc., CA) in a 50 lL reaction mixture
containing 5 lL DNA, 1.25 U TaKaRa Ex TaqTM (Takara
Bio Inc., Shiga, Japan), 109 Ex TaqTM buffer, 4 lL dNTP
mixture (2.5 mM each), and 10 pmol of each universal
primer, 27f (50 -AGAGTTTGATCCTGGCTCAG-30 ) and
1492r (50 -GGTTACCTTGTTACGACTT-30 ) (Dekio et al.,
2007). DNA amplification was conducted using the following method: preheating at 95 °C for 3 min; 30 cycles
each of denaturation at 95 °C for 30 s, annealing at
50 °C for 30 s, extension at 72 °C for 1.5 min, and final
extension at 72 °C for 10 min. The quantitative real-time
PCR (qPCR) products were purified by AMpure (Beckman Coulter, Inc., CA). For the sequencing reaction, the
BigDye Terminator version 3.1 Cycle Sequencing kit (Life
Technologies, CA) was used. The primers were the same
as those used for amplification; an additional primer,
520r (50 -CCAGCMGCYGCGGTAA-30 ) was also used.
Automated sequence determination was performed using
an ABI PRISM 3130xl Genetic Analyzer (Applied Biosystems
Inc., CA), and the alignment of the resulting 16S rRNA
gene sequences was performed by MEGA 5 (Tamura et al.,
2011). The 16S rRNA gene sequence of A20 was deposited under DDBJ accession no. AB853090 in GenBank.
Sequence data for phylogenetic trees were retrieved from
GenBank/DDBJ/EMBL, aligned by CLUSTALW using MEGA 5
and checked manually.
Antimicrobial activity of triclosan, IPMP, and
PGG
The minimum inhibitory concentrations (MICs) were
determined by measuring the test bacterial growth on the
basis of absorbance. Staphylococcus, Corynebacterium, and
Propionibacterium were cultured in TTL [per L: 30 g tryptic soy broth (Becton Dickinson, Tokyo, Japan) 5 g
Tween 80, 1 g lecithin, 15 g agar], Mueller Hinton broth
ª 2014 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
T. Fujii et al.
(MH; Becton Dickinson, Tokyo, Japan), and Gifu Anaerobic Broth (GAM Broth; Nissui Pharmaceutical Co.),
respectively, overnight at 37 °C. For the A20 strain, the
colonies grown in BL agar (containing 5% horse whole
blood) were added to nutrient broth (Becton Dickinson,
Tokyo, Japan) and cultured for 7 day at 37 °C. Cultured
bacteria were harvested by centrifugation (13 000 g for
3 min) and resuspended to 1.0 9 109 cells mL1 in a culture medium. Triclosan, IPMP, and PGG were subjected
to twofold dilution with EtOH and dispensed to a 96-well
plate (each well, 2 lL). Test bacteria were added to the
plate (final bacterial count, 1.0 9 106 cells mL1), and
2 lL of EtOH was used as a control. The plate (final volume 200 lL) was incubated at 37 °C for 2 or 7 day, and
OD660 nm was measured.
Two-week in vivo test with the 0.5% PGG
solution
An open study was conducted to evaluate the deodorant
effects of PGG on axillary odor (Fig. 1). Nine male inhouse volunteers were asked to spray 1 mL solution
containing 0.5% PGG (0.5% PGG solution) into the
axilla of the dominant hand and 1 mL solution without
PGG (non-PGG solution) into the axilla of the nondominant hand once a day for two consecutive weeks
after bathing.
Evaluation of axillary odor intensity before
and after a 2-week test with the visual
analogue scale (VAS)
Axillary malodor was measured by three judges trained
with the VAS using the cotton shirts worn by test subjects. The cotton shirt (BVD Men’s T-shirts, Fujibo Holdings Inc., Tokyo, Japan) was deodorized with a steam
iron (Aquaspeed Ultracord 250) for 15 min before use.
The volunteers wore the cotton shirt for 16 h after bathing on the day prior to measurement. After the left and
Sampling point I
(Before application)
Non-deodorant
usage for 1 week
Skin microbiota
analysis of swabscrubbed samples
(T-RFLP, qPCR)
Sampling point I I
(After application)
0.5% PGG or non-PGG
Once a day for 2 weeks
VAS analysis
of underwear
shirt
Fig. 1. Deodorant test scheme using PGG, and protocol for sampling
from the axilla. T-RFLP, qPCR, and VAS.
FEMS Microbiol Ecol 89 (2014) 198–207
201
Discovery of new axillary bacteria and new odor inhibitor
right axillary regions were cut out from the cotton shirts
that had been worn by the subjects, these pieces of cloth
were placed in a Ziploc bag (16.5 9 14.9 cm, Asahi Kasei
Home Products Corporation, Tokyo, Japan), sealed in an
anaerobic jar (AnaeroPack Series, Mitsubishi Gas Chemical company, Inc., Tokyo, Japan), and incubated at 37 °C
for 16 h. Samples with two types of odor intensity were
used as standard samples for VAS analysis. For odor
intensity 5, 10 lL of 10 mM HMHA was added to a rectangular piece from a cotton shirt, while for odor intensity
0, 10 lL of 100% EtOH was added to the cloth. The
judges first smelled the cloth with odor intensity 5 and
then smelled the cloth with odor intensity 0 prior to evaluating the odor intensity of each test sample.
DNA extraction from swab-scrubbed samples
obtained from the test subjects
Swab-scrubbed samples were obtained by the above methods from the 2-week-test volunteers. The swab-scrubbed
sample solutions were centrifuged at 13 000 g for 5 min
to obtain pellets. DNA was extracted from the swabscrubbed pellets as described above.
Terminal restriction fragment length
polymorphism (T-RFLP) analysis
PCR was performed using DNA Engine Tetrad 2 (BioRad Laboratories, Inc.) in a 50 lL reaction mixture containing 5 lL dissolved DNA (100 ng), 1.25 U TaKaRa Ex
TaqTM, 109 Ex TaqTM buffer, 4 lL dNTP mixture
(2.5 mM each), and 10 pmol of each universal primer,
27f and 1492r. Primer 27f was labeled with 6-FAM (6carboxyfluorescein; Applied Biosystems Inc.) for T-RFLP
analysis. Amplification was performed using the following
method: preheating at 95 °C for 3 min; 30 cycles each of
denaturation at 95 °C for 30 s, annealing at 50 °C for
30 s and extension at 72 °C for 1.5 min; and final extension at 72 °C for 10 min. The PCR products were purified by AMpure (Beckman Coulter, Inc.) and digested
with 20 U of either HhaI or MspI (Takara Bio Inc.) in a
total volume of 10 lL at 37 °C for 12 h. The length of
the terminal restriction fragments (T-RFs) was determined based on the standard size markers GS500 ROX
and 1000 ROX (Applied Biosystems Inc.) using the ABI
PRISM TM 3130xl genetic analyzer (Applied Biosystems
Inc.) and GENESCAN analysis software (Applied Biosystems
Inc.). Dendrogram analysis was performed using T-RFLP
patterns in the BIONUMERICS software (Applied Maths,
Sint-Martens-Latem, Belgium). The distances between
samples were represented graphically by constructing a
dendrogram based on the binary coefficient-dendrogram
type (Dice-UPGMA).
FEMS Microbiol Ecol 89 (2014) 198–207
qPCR for five skin bacteria
The counts of S. epidermidis, S. aureus, Corynebacterium
xerosis, P. acnes, and A20 in the 2-week-test samples were
estimated by qPCR. Primers specific to C. xerosis were
designed for the divIVA gene, and P. acnes and the A20
strain were designed for the gyrB gene in this study. The
divIVA gene sequence data of C. xerosis (accession no.
AM286228) and gyrB gene sequence data of P. acnes
(accession no. CP001977) were obtained from GenBank/
DDBJ/EMBL. The gyrB gene sequence of the A20 strain
was cloned from the genome of this strain. To clone the
gyrB gene, the primers Anagyrf_1222 (50 -AGACCKG
GWATGTATATMGGHC-30 ) and AnagyrR_1222 (50 -KT
CTWGGTTCTACCTTYTCWC-30 ) were used. Multiple
alignments of each gene sequences were carried out using
MEGA 5 to determine the sequence similarity between the
closely related species (Fig. 2a). The primer pairs
(Table 1) were designed by the LIGHTCYCLER PRONE DESIGN
Software 2.0 (Roche Diagnostics GmbH, Mannheim, Germany). The specification was tested for 18 skin bacterial
genome DNA, and amplification was detected only in target bacteria (Table 2). qPCR was performed in a 20 lL
mixture containing 2 lL DNA sample, 6 pmol of each
primer, and 10 lL THUNDERBIRD SYBR qPCR Mix
(Toyobo Co., Ltd, Osaka, Japan). Amplification and
detection were carried out using LightCycler DX400
(Roche Diagnostics GmbH) with a profile of preheating
at 95 °C for 20 s followed by 45 cycles each of denaturation at 95 °C for 5 s, annealing at 60 °C for 15 s and
extension at 72 °C for 30 s. The counts of the five
bacterial species in the samples were estimated with an
internal standard curve prepared with three replicates of
four concentrations (1 9 103, 1 9 105, 1 9 107, and
1 9 109 CFU mL1) of each bacterium. The bacteria
genomic DNA used for the standard curve was extracted
as described above. The counts of the five bacterial species in each sample have been expressed as CFU cm2.
Results
Isolation and identification of novel anaerobic
bacteria contributing to axillary odor
Based on colony shape, nine strains of anaerobic bacteria
from the axillary region were isolated. To determine their
HMHA-releasing activity, the percentage of HMHA
release from the precursor molecule HMHA-Gln was
measured. Most species, including Propionibacterium, Dermabacter, Finegoldia, Rothia, and Bacillus, did not release
any HMHA from HMHA-Gln, but the release percentage
from the A20 strain was very high, that is 79.8%
(Table 3). The 1484 base-pairs 16S rRNA gene of A20
ª 2014 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
202
T. Fujii et al.
(a)
Anaerococcus tetradius CCUG46590 AF542234
100
Anaerococcus lactolyticus CCUG31351 AF542233
Anaerococcus prevotii CCUG41932 AF542232
Anaerococcus vaginalis CCUG31349 AF542229
99
A20 AB853090
100
Anaerococcus octavius NCTC9810 Y07841
0.05
A20 AB853090
(b)
100
99
Anaerococcus sp. 8405254 HM587319
Uncultured Anaerococcus sp. clone ML2-55 DQ847450
Anaerococcus octavius NCTC9810 Y07841
Anaerococcus prevotii CCUG41932 AF542232
100
Anaerococcus tetradius CCUG46590 AF542234
77
Anaerococcus lactolyticus CCUG31351 AF542233
98
Anaerococcus murdochii WAL17230 DQ911243
Anaerococcus vaginalis CCUG31349 AF542229
0.01
Fig. 2. NJ tree based on 16S rRNA gene (a) and gyrB gene (b) sequences showing the relationship of the A20 and other Anaerococcus strains.
The bar represents 10 nucleotide substitutions per 1000 sites. Bootstrap values (> 50%) based on 1000 replications are shown at branch nodes.
Table 1. Primer pairs used in the qPCR for five skin bacteria
Bacterium (gene)
S. epidermidis (sodA)
S. aureus (nuc)
P. acnes (gyrB)
C. xerosis (divIVA)
A20 (gyrB)
Primer
SE-F
SE-R
nuc-F
nuc-R
PA_gyrBF
PA_gyrBR
cxdiv-nef1
cxdiv-ner1
A839F
A1082R
Sequence
0
5 -TCAGCAGTTGAAGGGACAGAT-3
50 -CCAGAACAATGAATGGTTAAGG-30
50 -AGGGATGGCTATCAGTAATG-30
50 -GCTGAGCTACTTAGACTTGAAA-30
50 -CTACCGATCATCCTGATGGTC-30
50 -ACCGGCATCGTAGGAAC-30
50 -GACGAGACCCTGGCCAA-30
50 -GTCTCGGACTCCGTCTTC-30
50 -CATTTATATCTGTTGATTGACAATC-30
50 -CAATTTGAAGGACAAACTAAGGCAA-30
was sequenced and analyzed for homology using the GenBank database. The results showed 96.2% homology with
Anaerococcus octavius, a very high homology of 99.8%
with Anaerococcus sp. 8405254 (accession no. HM587319)
from human osteoarticular origin and a very high homolª 2014 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
Reference
0
Iwase et al. (2008)
Brakstad et al. (1992)
In this study
In this study
In this study
ogy of 99.6% with uncultured Anaerococcus sp. clone
ML2-55 (accession no. DQ847450) from human skin. In
addition, the A20 strain belonged to the same group as
Anaerococcus sp. 8405254 and uncultured Anaerococcus
sp. clone ML2-55 and to a different group on the neighFEMS Microbiol Ecol 89 (2014) 198–207
203
Discovery of new axillary bacteria and new odor inhibitor
Table 2. Specificity test of A20 specific primer
No.
Name
Strain No.
or isolated No.
PCR*
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Anaerococcus sp.
Anaerococcus hydrogenalis
Anaerococcus lactolyticus
Anaerococcus murdochii
Anaerococcus octavius
Anaerococcus prevotii
Anaerococcus senegalensis
Anaerococcus tetradius
Anaerococcus vaginalis
Corynebacterium xerosis
Corynebacterium coyleae
Corynebacterium pseudogenitalium
Staphylococcus epidermidis
Staphylococcus aureus
Propionibacterium acnes
Propionibacterium avidum
Propionibacterium granulosum
Propionibacterium propionicus
A20
JCM7635
JCM8140
JCM15630
DSM11663
JCM6508
DSM25366
JCM1964
JCM8138
ATCC373
M8WR5
M8WR2
ATCC14990
ATCC12600
P1
P2
P3
P4
+
ATCC, American Type Culture Collection; JCM, Japan Collection of
Microorganisms; DSM, German Collection of Microorganisms and Cell
Cultures; and other numbers were isolated strains.
*A single PCR band was defined as a positive test result (+).
Table 3. HMHA-releasing rates of anaerobic isolated bacteria and
nine Anaerococcus strains
Name
Isolated No.
or strain No.
HMHA release (%)
Propionibacteria acnes
Propionibacterium avidum
Propionibacterium granulossum
Propionibacterium propionicus
Dermabacter hominis
Finegoldia magna
Rothia dentocariosa
Bacillus firmus
Anaerococcus species
Anaerococcus hydrogenalis
Anaerococcus lactolyticus
Anaerococcus murdochii
Anaerococcus octavius
Anaerococcus prevotii
Anaerococcus senegalensis
Anaerococcus tetradius
Anaerococcus vaginalis
P1
P2
P3
P4
D1
F1
R1
B1
A20
JCM7635
JCM8140
JCM15630
DSM11663
JCM6508
DSM25366
JCM1964
JCM8138
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
79.8
0.4
0.0
0.5
1.3
0.2
0.0
0.3
0.3
JCM, Japan Collection of Microorganisms; DSM, German Collection
of Microorganisms and Cell Cultures; and other numbers were isolated strains.
bor-joining (NJ) phylogenetic tree (Fig. 2b). The released
HMHA percentage was measured in eight Anaerococcus
strains
(A. hydrogenalis
JCM7635,
A. lactolyticus
JCM8140,
A. murdochii
JCM15630,
A. octavius
DSM11663, A. senegalensis DSM25366, A. tetradius
FEMS Microbiol Ecol 89 (2014) 198–207
JCM1964, A. vaginalis JCM8138, and the A20) to confirm
whether the ability to release HMHA was common. The
analysis showed that HMHA release occurred only in the
A20 strain (Table 3).
MICs of antimicrobial agents against bacteria
present on the skin, including the newly
discovered axillary odor bacterium A20
To evaluate the effectiveness of known antimicrobial
agents against the A20 isolate, the MICs of triclosan and
IPMP were tested against A20, as well as the skin bacteria
S. epidermidis, P. acnes, C. xerosis, and the pathogenic
S. aureus (Table 4). Triclosan inhibited the growth of
S. epidermidis, P. acnes, and S. aureus to < 0.0001% and
of C. xerosis to 0.025%. IPMP reduced the growth of
S. epidermidis, S. aureus, and P. acnes to 0.025% and of
C. xerosis to 0.1%. However, triclosan and IPMP could
not inhibit the A20 strain at the maximum tested concentration of 0.1%.
We conducted screening for antimicrobial materials
against the A20 isolate from about 500 plant extracts.
PGG, a type of polyphenol extracted from the Chinese
Gall plant, was found to inhibit the A20 strain. PGG
inhibited the A20 isolate at the concentration of 0.025%;
triclosan and IPMP did not suppress the isolate when
used at the concentration of 0.1%. The inhibition activity
of PGG was also stronger than that of IPMP for S. epidermidis, S. aureus, and P. acnes (Table 4).
Comparison of the effect of inhibitory
concentrations on HMHA release in the case of
PGG and known antimicrobial agents
The inhibitory concentrations of HMHA release by the
A20 cell lysate were measured in triplicate (Fig. 3). To
compare the inhibitory activity of PGG, triclosan and
IPMP, the average IC50 values were calculated and were
found to be 0.0033%, 0.0048%, and 0.011%, respectively.
The inhibitory activity of PGG was 1.4 times stronger
than the activity of triclosan and 3.3 times stronger than
that of IPMP.
Reduction of axillary odor strength by PGG
application
As PGG had an antimicrobial effect for the A20 isolate as
well as an inhibitory effect on HMHA release in vitro, its
effect on inhibiting axillary odor in vivo was estimated. At
the end of the 2-week test, the axillae of the volunteers were
carefully observed, and no dermatological problems were
detected, leading to the conclusion that PGG did not cause
any skin problems. The VAS value for the nine subjects on
ª 2014 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
204
T. Fujii et al.
Table 4. Minimum growth inhibitory concentration of PGG, triclosan, and IPMP against five skin bacteria
Minimum inhibition concentration (final %)
Sample
S. epidermidis ATCC14990
S. aureus ATCC12600
C. xerosis ATCC373
P. acnes JCM6425
Anaerococcus sp. A20
PGG
Triclosan
IPMP
0.0004
< 0.0001
0.0250
0.0016
< 0.0001
0.0250
0.0500
0.0250
0.1000
0.0016
< 0.0001
0.0250
0.0250
> 0.1
> 0.1
S, Staphylococcus; C, Corynebacterium; P, Propionibacterium; ATCC, American Type Culture Collection; JCM, Japan Collection of Microorganisms, and the A20 was isolated strain.
1
*
, pentagalloylglucose
, triclosan
, 4-isopropyl-3-methylphenol
100
0
50
0
0
0.005
0.01
0.015
0.02
0.025
0.03
Concentration (% wt vol–1)
Fig. 3. Inhibitory activities of pentagalloylglucose, triclosan, and 4isopropyl-3- methylphenol on HMHA releasing by lysates of the A20
strain. Inhibition percentage is obtained from the peak area of
released HMHA of each sample divided by released HMHA without
sample treatment.
the control treatment side was subtracted from the VAS
value of the experimental treatment side to obtain a mean
value. This value was 0.2 before treatment and 1.1 after
treatment, which is a significant difference (Fig. 4). The
negative value of this difference indicates that the axillary
odor strength decreased after 2 weeks of treatment.
Reduction of the A20 count in the axilla by
PGG application
Two weeks after applying either the 0.5% PGG solution
or the non-PGG solution as a control, bacteria counts of
S. epidermidis, C. xerosis, P. acnes, and A20 were quantified by qPCR. In contrast to the in vitro results, the
counts of S. epidermidis, C. xerosis, and P. acnes did not
differ between the non-PGG solution and the 0.5% PGG
solution treatments (Fig. 5). However, significant reduction in the counts of the A20 was observed in 0.5% PGG
solution treatments as compared to those for the nonPGG solution (Fig. 5).
PGG does not have a major effect on axillary
microbiota
In contrast to the in vitro result, the 0.5% PGG solution
had no effect on the counts of S. epidermidis, C. xerosis,
ª 2014 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
Difference in VAS
HMHA production rate (%)
150
–1
–2
Before
After
–3
Fig. 4. The difference in VAS of axillary odor between before and
after PGG application. The difference is obtained from subtracting the
non-PGG solution side from the PGG solution side. If the difference is
< 0, it is implied that odor intensity of the PGG solution side is
reduced after application. The average values of before and after
PGG application difference in VAS were 0.2 and 1.1, respectively.
Significant difference (P < 0.05) was calculated by the two-way
repeated-measures ANOVA. *P < 0.05 vs. before. Markers represent
each subject.
and P. acnes; therefore, the effect of the PGG application
on axillary microbiota was evaluated. The microbiota
before and after application were analyzed by T-RFLP
and clustered based on T-RFLP profiles digested with
MspI and HhaI. After application, most of the samples
(except no. 6) were from the same cluster, and no clear
clusters corresponding to the presence or absence of PGG
were observed (Fig. 6). Similarities in T-RFLP profiles
before and after treatment with both 0.5% PGG solution
and non-PGG solution were determined. There was no
significant difference between the results for the 0.5%
PGG solution side and the non-PGG solution side.
Discussion
The main components of axillary odor are isovaleric acid,
3M2H, HMHA, and 3-methyl-3-mercaptohexan-1-ol
(Natsch et al., 2003, 2004, 2006; James et al., 2004b;
FEMS Microbiol Ecol 89 (2014) 198–207
205
Discovery of new axillary bacteria and new odor inhibitor
Number of bacteria (cells cm–2)
*
1.0E+06
, Non-PGG
, 0.5% PGG
1.0E+04
1.0E+02
1.0E+00
Fig. 5. Quantification of four axillary bacteria using qPCR after
application. Statistically significant differences (P < 0.05) were
calculated by the t-test (P = 0.0018). *P < 0.05. S: Staphylococcus, C:
Corynebacterium, P: Propionibacterium
Starkenmann et al., 2005; Emter & Natsch, 2008). These
compounds are produced from odorless precursors in
sweat secreted from the apocrine sweat glands through
the action of indigenous skin bacteria, mainly Corynebacterium species (Fredrich et al., 2013). These species are
particularly responsible for the production of 3M2H and
HMHA. These molecules are produced from L-glutamineconjugated precursors by the action of Na-acylglutamine
aminoacylase, an enzyme specific to the Cornyebacterium
species (Natsch et al., 2003, 2006). There are many
reports on the involvement of aerobic bacteria in axillary
odor production, but fewer reports exist on the role of
anaerobic bacteria (Leyden et al., 1981; Taylor et al.,
2003). However, with new sequence analytical methods,
such as next-generation sequencing, many anaerobic bacteria that could not be isolated previously have been discovered on the skin surface (Costello et al., 2009). This
study focused on anaerobic bacteria and identified a
strain that contributes to axillary odor production.
We obtained six anaerobic genera with nine isolates
from the axilla of healthy subjects by sampling and culture techniques. Using the method of Natsch et al. (2003)
which involved measuring HMHA release from HMHAGln, we found that the A20 strain had a strong HMHAreleasing activity (Table 3). The identity of the A20 strain
was determined using matches based on 16S rRNA gene
Fig. 6. Dendrogram of terminal restriction
fragment polymorphism profiles before and
after PGG application. The dendrogram was
constructed by the binary coefficientdendrogram (Dice-UPGMA). The numerals
show the subject number, while PG and nPG
show 0.5% PGG solution application and nonPGG solution application, respectively. A and B
show before and after application,
respectively.
FEMS Microbiol Ecol 89 (2014) 198–207
100
95
90
85
80
75
70
65
60
55
50
45
40
35
30
25
Similarity (%)
8 nPG B
8 PG B
8 nPG A
8 PG A
6 PG A
2 nPG B
2 PG B
4 nPG B
1 nPG B
1 PG B
5 nPG B
7 nPG B
7 PG B
5 PG B
6 nPG B
5 nPG A
6 PG B
5 PG A
6 nPG A
9 nPG B
9 PG B
1 PG A
7 nPG A
7 PG A
9 nPG A
9 PG A
2 nPG A
2 PG A
4 nPG A
4 PG A
1 nPG A
4 PG B
3 PG B
3 nPG B
3 nPG A
3 PG A
ª 2014 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
206
similarities, and the strain was found to have a very high
homology with Anaerococcus sp. 8405254 of osteoarticular
origin (La Scola et al., 2011) and uncultured Anaerococcus
sp. clone ML2-55 (accession no. DQ847450) of forearm
origin (Gao et al., 2007) (99.8% and 99.6% homology,
respectively). According to Stackebrandt and Ebers
(2006), a 16S rRNA gene homology > 98% correlates
with DNA–DNA hybridization indicating that the isolated
A20 strain belongs to the genus Anaerococcus. As of now,
anaerobic bacteria have not been identified to contribute
to axillary odor. However, the ability of the A20 strain to
release HMHA and the high numbers of this species in
the axilla, found through qPCR (Fig. 5), suggest that A20
strongly contributes to producing axillary odor.
Compared to other skin bacteria, the A20 strain showed
strong resistance to the existing antimicrobial agents triclosan and IPMP (Table 4). Hence, we tested various plant
extracts for antimicrobial activity against the A20 strain
and discovered that PGG has antimicrobial activity against
the A20 strain. PGG was also more effective than IPMP
against S. epidermidis, S. aureus, P. acnes, and C. xerosis in
vitro (Table 4). Our research also showed that PGG
reduced the A20 strain count when applied to the skin surface (Fig. 5). In contrast, PGG application did not affect
the counts of the other bacteria or the skin microbiota
(Fig. 5 and 6). This difference between the effects in vivo
and in vitro may be due to the difference in the growth rate
of skin bacteria. The concentration of PGG is kept constant
in vitro, whereas the concentration is gradually diluted by
sweat after application in vivo. Therefore, the antimicrobial
effect of PGG is predicted to be gradually weakened after
application in vivo. Because aerobic bacteria grow faster
than anaerobic bacteria, PGG was not expected to reduce
aerobic bacteria with fast growth in vivo due to dilution.
Accordingly, only the number of the anaerobic A20 was
significantly reduced in vivo. PGG also inhibits HMHA
release at a lower concentration than other antimicrobial
agents such as triclosan and IPMP (Fig. 3). Considering
the reports on PGG as an enzyme inhibitor (Zhang et al.,
2009), PGG is suspected to directly inhibit the activities of
HMHA-releasing enzymes.
Usually, we can feel the body odor after daily activities.
Body odors are generated by skin bacteria from the precursors adhering to the clothes after daily activities.
Therefore, the using cotton shirts were incubated for 16 h
to produce odors. The sensory tests on odor strength
revealed that, after 2 weeks of treatment with PGG, the
axillary odor significantly declined (Fig. 4). Gas chromatography measurements showed that HMHA concentrations declined in five of the nine subjects (data not
shown), implying that PGG applications inhibited axillary
odor. In addition, we interviewed the test subjects and
asked them about the change in the odor profile, for
ª 2014 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
T. Fujii et al.
example, the presence of a sulfurous odor or acidic odor
after the test. This analysis revealed a reduction of some
kind of odor after the treatment, suggesting that PGG
also reduced the axillary odor components other than
HMHA.
The above results show that PGG application does not
damage the skin microbiota and can reduce axillary odor
by inhibiting Anaerococcus. Thus, PGG will be an innovative deodorant material that shows odor suppressive
activity while still maintaining the skin microbiota.
References
Brakstad OG, Aasbakk K & Maeland JA (1992) Detection of
Staphylococcus aureus by polymerase chain reaction
amplification of the nuc gene. J Clin Microbiol 30: 1654–1660.
Coagen AL, Nizet V & Gallo RL (2008) Skin microbiota: a
source of disease or defence? Br J Dermatol 158: 442–455.
Costello EK, Lauber CL, Hamady M, Fierer N, Gordon JI &
Knight R (2009) Bacterial community variation in human
body habitats across space and time. Science 326: 1694–
1697.
Dekio I, Sakamoto M, Hayashi H, Amagai M, Suematsu M &
Bennno Y (2007) Characterization of skin microbiota in
patients with atopic dermatitis and in normal subjects using
16S rRNA gene-based comprehensive analysis. J Med
Microbiol 56: 1675–1683.
Emter R & Natsch A (2008) The sequential action of a
dipeptidase and a beta-lyase is required for the release of
the human body odorant 3-methyl-3-sulfanylhexan-1-ol
from a secreted Cys-Gly-(S) conjugate by Corynebacteria. J
Biol Chem 283: 20645–22065.
Fredrich E, Barzantny H, Brune I & Tauch A (2013) Daily
battle against body odor: towards the activity of the axillary
microbiota. Trends Microbiol 21: 305–312.
Gao Z, Tseng CH, Pei Z & Blaser MJ (2007) Molecular
analysis of human forearm superficial skin bacterial biota. P
Natl Acad Sci USA 104: 2927–2932.
Hasegawa Y, Yabuki M & Matsukane M (2004) Identification
of new odoriferous compounds in human axillary sweat.
Chem Biodivers 1: 2042–2050.
Iwase T, Hoshina S, Seki K, Shinji H, Masuda S & Mizunoe Y
(2008) Rapid identification and specific quantification of
Staphylococcus epidermidis by 50 nuclease real-time
polymerase chain reaction with a minor groove binder
probe. Diagn Microbiol Infect Dis 60: 217–219.
James AG, Casey J, Hyliands D & Mycock G (2004a) Fatty
acid metabolism by cutaneous bacteria and its role in
axillary malodour. World J Microbiol Biotechnol 20: 787–793.
James AG, Hyliands D & Johnston H (2004b) Generation of
volatile fatty acids by axillary bacteria. Int J Cosmet Sci 26:
149–156.
La Scola B, Fournier PE & Raoult D (2011) Burden of
emerging anaerobes in the MALDI-TOF and 16S rRNA gene
sequencing era. Anaerobe 17: 106–112.
FEMS Microbiol Ecol 89 (2014) 198–207
Discovery of new axillary bacteria and new odor inhibitor
Leyden JJ, McGinley KJ, Holzle E, Labows JN & Kligman AM
(1981) The microbiology of the human axilla and its
relationship to axillary odor. J Invest Dermatol 77: 413–416.
Natsch A, Gfeller H, Gygax P, Schmid J & Acuna G (2003)
A specific bacterial aminoacylase cleaves odorant
precursors secreted in the human axilla. J Biol Chem 278:
5718–5727.
Natsch A, Schmid J & Flachsmann F (2004) Identification of
odoriferous sulfanylalkanols in human axilla secretions and
their formation through cleavage of cysteine precursors by a
C-S lyase isolated from axilla bacteria. Chem Biodivers 1:
1058–1072.
Natsch A, Derrer S, Flachsmann F & Schmid J (2006) A broad
diversity of volatile carboxylic acids, released by a bacterial
aminoacylase from axilla secretions, as candidate molecules
for the determination of human-body odor type. Chem
Biodivers 3: 1–20.
Shelley WB, Hurley HJ Jr & Nichols AC (1953) Axillary
odor experimental study of the role of bacteria, apocrine
sweat, and deodorants. AMA Arch Derm Syphilol 68: 430–
446.
FEMS Microbiol Ecol 89 (2014) 198–207
207
Stackebrandt E & Ebers J (2006) Taxonomic parameters revisited:
tarnished gold standards. Microbiol Today 33: 152–155.
Starkenmann C, Niclass Y, Troccaz M & Clark AJ (2005)
Identification of the precursor of (S)-3-methyl-3-sulfany
lhexan-1-ol, the sulfury malodour of human axilla sweat.
Chem Biodivers 2: 705–716.
Tamura K, Peterson D, Peterson N, Stecher G, Nei M &
Kumar S (2011) MEGA 5: molecular evolutionary genetics
analysis using maximum likelihood, evolutionary distance,
and maximum parsimony methods. Mol Biol Evol 28: 2731–
2739.
Taylor D, Daulby A, Grimshaw S, James G, Mercer J & Vaziri
S (2003) Characterization of the microflora of the human
axilla. Int J Cosmet Sci 25: 137–145.
Zeng X-N, Leyden JJ, Lawley HJ, Sawano K, Nohara I & Preti
G (1991) Analysis of characteristic odors from human male
axillae. J Chem Ecol 17: 1469–1492.
Zhang J, Li L, Kim SH, Hagerman AE & L€
u J (2009)
Anti-cancer, anti-diabetic and other pharmacologic and
biological activities of penta-galloyl-glucose. Pharm Res 26:
2066–2080.
ª 2014 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved