MAJOR ARTICLE
Native Efflux Pumps Contribute Resistance to
Antimicrobials of Skin and the Ability of
Staphylococcus aureus to Colonize Skin
Que Chi Truong-Bolduc, Regis A. Villet, Zoe A. Estabrooks, and David C. Hooper
Division of Infectious Diseases, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts
Background. Staphylococcus aureus colonizes skin in the presence of antimicrobial fatty acids and polyamines.
The chromosomally encoded Tet38 efflux transporter confers resistance to tetracycline and fitness in abscesses, but
its natural substrates and those of the Nor quinolone efflux pumps are unknown.
Methods. Susceptibility of tet38 and other pump mutants to and pump gene induction by fatty acids and polyamines were compared. Transport of fatty acids by Tet38 was determined in membrane vesicles. Survival on skin
was tested in an adapted mouse skin infection model.
Results. The tet38 expression caused a 5- to 8-fold increase in resistance to palmitoleic and undecanoic acids
but not polyamines. Subinhibitory concentrations of these fatty acids induced 4-fold increases in tet38 transcripts
and competitively inhibited transport of Hoechst 33 342 dye in Tet38 membrane vesicles. Colonization of skin in
BALB/c mice was decreased 5-fold in a Δtet38 mutant, which was complemented by plasmid-encoded tet38. Although polyamine minimum inhibitory concentrations (MICs) decreased 4-fold in a norC::cat mutant and increased 8-fold with norC overexpression, spermidine did not induce expression of norC and other pump genes, and
norC::cat exhibited wild-type colonization.
Conclusion. Antibacterial fatty acids may be natural substrates of Tet38, which contributes to resistance and the
ability of S. aureus to colonize skin.
Keywords.
Tet38; NorC; Fatty acid; S. aureus; Spermidine; Tetracycline.
Lipids are important for energy storage, membrane
structure, and cell signaling [1, 2]. Free fatty acids may
be found in staphylococcal abscesses and are usually
present on the skin surface where they may serve to
limit skin colonization by bacteria such as Staphylococcus aureus and Staphylcococcus epidermidis [3–5].
Human sebum contains 16% free fatty acids, and sapienic acid is a sebum fatty acid that is unique to humans
[6, 7]. The equivalent of sapienic acid in mouse sebum
is palmitoleic acid, a monounsaturated fatty acid
present in all tissues, including the skin surface. Lauric
acid is a minor sebum component, occupying 1%–2%
Received 27 July 2013; accepted 12 November 2013; electronically published 26
November 2013.
Correspondence: Que Chi Truong-Bolduc, PhD, Infectious Disease Division, Massachusetts General Hospital, 55 Fruit St, Boston MA 02114-2696 (qtruongbolduc@
partners.org).
The Journal of Infectious Diseases 2014;209:1485–93
© The Author 2013. Published by Oxford University Press on behalf of the Infectious
Diseases Society of America. All rights reserved. For Permissions, please e-mail:
[email protected].
DOI: 10.1093/infdis/jit660
of total sebum free fatty acids. These fatty acids have
antimicrobial properties and can cause a reduction in
the growth of S. aureus and S. epidermidis [8, 9].
In human tissue, skin surface, and nasal fluid, the
main bactericidal long chain unsaturated fatty acids are
linoleic (C18:2), oleic (C18:1), palmitoleic (C16:1), and
long chain saturated fatty acids are palmitic (C16:0),
and stearic (C18:0) acids [10]. In addition, unsaturated
fatty acids, such as undecanoic and undecenoic acids
and their monacyl glycerols, are also active against
S. aureus and Bacillus cereus [6, 11].
Antimicrobial polyamines are also present on skin
[12, 13]. Polyamines are aliphatic compounds and exert
pleiotropic effects on cellular physiology of all organisms. S. aureus lacks polyamine biosynthetic genes and
cannot produce spermidine, spermine, or their precursors. Exogenous polyamines inhibit S. aureus growth
and are bactericidal at concentrations that are nontoxic
to host cells [14].
S. aureus has developed a number of strategies to survive on skin in the presence of a range of antimicrobial
Role of Efflux Pumps in S. aureus Survival
•
JID 2014:209 (1 May)
•
1485
compounds [15]. Under iron-limited conditions, the surface
protein IsdA contributes to resistance to fatty acids and antimicrobial peptides due to its ability to reduce binding or sequestration of these antimicrobial compounds [16]. Wall teichoic
acids also appear to prevent penetration of hydrophobic antimicrobial fatty acids across the cell wall and reduce binding to
the cytoplasmic membrane [17].
Most S. aureus strains are highly susceptible to polyamines,
such as spermine and spermidine, by an uncertain mechanism,
but some strains, such as the dominant community-associated
strain USA300, are resistant to skin polyamines due to acquisition of the speG gene found on the mobile ACME element [18].
A number of efflux pumps of S. aureus have been identified
to confer resistance to antimicrobial agents. The norB (resistance to quinolones), norD, and tet38 (resistance to tetracycline),
Table 1.
are known to be selectively increased in subcutaneous abscesses,
which contain antimicrobial fatty acids [4], and these pumps
contribute to the ability of S. aureus to survive in the abscess
environment [19, 20]. Linoleic acid has also been shown to
induce expression of efflux transporter genes tet38, mdeA, and
mmpL [21]. We thus evaluated the effects of five known efflux
pumps, Tet38, NorA, NorB, NorC, and NorD on resistance to
antimicrobial fatty acids and polyamines. We studied the effects
of these compounds on pump gene expression, the ability of
Tet38 to transport fatty acids, and their ability to promote survival on skin. NorB and NorD had little effect, but expression of
tet38 was induced by fatty acids, and Tet38 conferred resistance
to and transported select free fatty acids and contributed to
S. aureus survival on mouse skin. Although NorC contributed to
polyamine resistance, it did not affect skin colonization.
Bacterial Strains and Plasmids Used in This Study
Strains or Plasmids
Genotypes or Relevant Characteristic(s)
Reference or Source
Staphylococcus aureus
RN6390
8325-4 wild type
[22]
KL820
MT23142
RN4222 norA::cat, norA(Δ)
norA overexpressor, (norA++)
[23]
[24]
QT5
8325-4 norB::cat, norB(Δ)
[25]
QT7
QT9
8325-4 Δtet38, tet38(Δ)
8325-4 norC::cat, norC(Δ)
[25]
[26]
QT10
8325-4 norD::cat, norD(Δ)
This study
MW2
RN6390 (pLI50-tet38)
CA-MRSA (USA400 lineage)
tet38 overexpressor, tet38 (++)
[20]
[25]
RN6390 (pQT8)
norB overexpressor, norB (++)
[25]
RN6390 (pQT9)
RN6390 (pQT10)
norC overexpressor, norC (++)
norD overexpressor, norD (++)
[26]
This study
MW2 (ΔnorB)
MW2 norB::cat
[20]
MW2 (Δtet38)
MW2 (ΔnorC)
MW2 tet38::cat
MW2 norC::cat
[20]
This study
MW2 (ΔnorD)
MW2 norD (Δ)
[19]
MW2 (Δtet38geh::pSKtet38)
MW2 (pQT8)
MW2 tet38 with pSKtet38 insert in geh
norB overexpressor, norB (++)
[20]
This study
MW2 (pQT9)
norC overexpressor, norC (++)
This study
MW2 (pLI50-tet38)
MW2 (pQT10)
tet38 overexpressor, tet38 (++)
norD overexpressor, norD (++)
This study
This study
BL21(DE3)
BL21(DE3) (pTrcHis2-norA)
−
E. coli B F−dcm ompT hsdS(r−
BmB) gal λ(DE3)
Overexpress NorA
Stratagene
[27]
BL21(DE3) (pTrcHis2-tet38)
Overexpress Tet38
This study
Escherichia Coli
Plasmids
pGEM3-zf(+)
2.9-kb E. coli cloning vector, ApR
Promega
pLI50
Shuttle cloning vector (ApR CmR)
[28]
pSK950
10.5-kb plasmid carrying the attP site of phage L54a, replicon
of pE194, TcR, EmR (S. aureus)
[29]
pTrcHis2
cloning and His-tag expressing vector in E. coli
Invitrogen
pQT8
pSK950-norB
[25]
1486
•
JID 2014:209 (1 May)
•
Truong-Bolduc et al
MATERIALS AND METHODS
were prepared using a French press at 1500 lb/in2, as previously
described [27].
Bacterial Strains and Growth Conditions
S. aureus was cultivated in trypticase soy broth (TSB; Difco,
Sparks, Md.), and Escherichia coli in Luria-Bertani (LB) broth.
Bacteria grew at 37°C in glass flasks, with a flask-to-media ratio
of 2.5 (100 mL LB/250-mL flask) and shaking at 220 rpm
under normal aeration. Ampicillin (100 µg/mL) was added
when necessary (Table 1). Strains QT10 and MW2 (ΔnorC)
were created as previously described using phage Φ85 with
MW2 (ΔnorD) and QT9 as donors.
Antibiotics, Chemical Compounds, and MICs
Tetracycline, lysostaphin, Hoechst 33 342, spermidine, spermine, putrescine, linoleic, oleic, lauric, palmitoleic, palmitic,
and undecanoic acids were from Sigma Chemical Co., St. Louis,
MO. Ampicillin and isopropyl-β-D-thiogalactopyranoside
(IPTG) were from Fisher Scientific, Pittsburgh, PA. MICs of
fatty acids, polyamines, and other compounds were determined
by LB broth microdilution, as described elsewhere [30].
Induction of tet38 and norC Expression by Skin Antimicrobials
S. aureus RN6390 and MW2 cultures were grown at 37°C to
OD600 0.5. Fatty acids or polyamines were added to the culture
followed by incubation for 1 hour. Real-time quantitative
reverse-transcription polymerase chain reactions (qRT-PCRs)
were carried out as described elsewhere [31]. Primers designed
from the tet38 gene (Tet38-F, 5′-ATCGTAGTATTTACGTT
GCCATTCC-3′; Tet38-R, 5′-GTTGCCACTAGAATTAAGCC
AACAA-3′), and the norC gene (NorC-F, 5′-TGGGTTGGAG
TGGATTTTC-3′; NorC-R, 5′-ACAATTAGCCCTGCAACG
TC-3′), were synthesized at the Tufts University, Boston, MA.
All samples were analyzed in triplicate and normalized against
the gmk housekeeping gene internal control.
Plasmid Construction
Plasmid pTrcHis2-tet38 was constructed as previously described by Yu et al [27].
tet38 was amplified from S. aureus RN6390 (Tet38-BamHI,
5′-ATTATGAGGATCCATGTTGAATAT-3′, and Tet38-EcoRI,
5′-TATCTATGAATTCTTATTTTTCAG-3′, underlined regions
indicate the inserted BamHI and EcoRI sites). The PCR product
was ligated to the BamHI and EcoRI sites of pTrcHis2 (Invitrogen Inc, Carlsbad, CA). The construct was transformed into
E. coli BL21, followed by DNA sequencing.
Preparation of Everted Membrane Vesicles
E. coli BL21 with plasmid pTrcHis2-tet38 or pTrcHis2
[control] was cultured in LB broth with ampicillin (100 µg/mL)
and grown at 37°C. At OD600 = 0.7, IPTG was added to 0.5 mM.
Cells were harvested after 4 hours and resuspended in 50 mM
potassium phosphate buffer (pH 7.2). Everted membrane vesicles
Tet38 Efflux Pump Transport in Everted Membrane Vesicles
Lactate-dependent decreases in fluorescence of Hoechst 33 342,
a dye that exhibits membrane-specific fluorescence, were used
to measure membrane transport as described elsewhere [27].
Assay mixtures contained 40 µg/mL of everted vesicle protein
in 50 mM HEPES pH 7.2, 8.5 mM NaCl, and 2 mM MgSO4 in
a final volume of 2 mL. Hoechst 33 342 produced a linear fluorescence response between 50 and 200 nM, as described elsewhere [27]. Double reciprocal plots of concentrations of
Hoechst vs the rates of fluorescence change were generated to
calculate the Vmax and Km of Hoechst 33 342 transport, and the
kinetics of transport with tetracycline and fatty acid were measured as competition for Hoechst 33 342 transport. Competition patterns were determined by the concurrence of the y and
x intercepts from least-squares regression plots of 1/V and 1/S,
where V is the initial velocity of transport and S is the concentration Hoechst 33 342 substrate. Values of Ki were calculated
as previously described [27]. Tetracycline fluorescence at the
amounts used (70 and 35 nM) did not affect Hoechst fluorescence. Fatty acids and their solvents did not produce measureable quenching of Hoechst 33 342 fluorescence in the assays.
Mouse Skin Colonization Model
To evaluate skin colonization, we modified a superficial skin infection model [32]. A saline suspension containing 107 cells of
S. aureus strain MW2, MW2Δtet38, or MW2norC::cat mutants
was added to 4 mm filter paper discs, placed onto the shaved
flank of 8- to 14-week old S. aureus-free C57BL/6 male mice
(Charles River Laboratories, Wilmington, MA). A salineloaded control disc was applied to the opposite flank. Both sites
were covered with 1.0-cm2 pieces of plastic sheet and secured
with Transpore tape and Nexcare waterproof tape (3 M). After
24 hours, the skin beneath the discs was abraded with a scalpel
blade and resuspended in 1 mL saline for plating. Mouse experiments were approved by the MGH IACUC (Institutional
Animal Care and Use Committee).
RESULTS
Effects of Efflux Pump Expression on Resistance to Fatty Acids
and Polyamines in S. aureus
We determined the effects of S. aureus efflux pumps on the
antibacterial activity of 6 fatty acids by comparing parental,
knockout, and plasmid-overexpression constructs for Tet38,
NorA, NorB, NorC, and NorD efflux pumps [19, 20, 23]
(Table 2). All fatty acids had similar activity (MIC 10–12.5 µg/mL)
against RN6390, with the exception of lauric acid, which was 4to 5-fold less active. Differences between strains differing in
efflux pump expression were seen principally with the Tet38
Role of Efflux Pumps in S. aureus Survival
•
JID 2014:209 (1 May)
•
1487
Table 2. Effect of Fatty Acids, Polyamines, Tetracycline, and Hoechst 33 342 on Staphylococcus aureus Mutants and Overexpressor
MDR Pumps
MIC (µg/mL)
Fatty Acids
Polyamines
S. aureus
Unde
(C11)
Lauric
(C12)
Palmitol.
(C16:1)
Palmitic
(C16:0)
Oleic
(C18:1)
Linoleic
(C18:2)
Sped
RN6390
tet38 (++)
12.5
100
50
200
10
50
10
10
10
50
10
50
2
2
Spe
Putr
1
1
1
1
Tetra
Hoe
.125
8
.5
2
tet38 (Δ)
6.25
50
5
10
10
10
2
1
1
.06
.25
norB (++)
norB (Δ)
50
6.25
100
50
10
10
10
10
20
10
20
10
2
2
1
1
1
1
.125
.125
.5
.5
norA (++)
12.5
50
10
10
10
10
2
1
1
.125
.5
norA (Δ)
norC (++)
12.5
12.5
50
50
10
10
10
10
10
10
10
10
2
16
1
8
1
8
.125
.125
.5
.5
norC (Δ)
12.5
50
10
10
10
10
.125
.5
norD (++)
norD (Δ)
12.5
12.5
50
50
10
10
10
10
10
10
10
10
2
2
1
1
1
1
.125
.125
.5
.5
MW2
12.5
50
10
10
10
10
2
1
1
tet38 (++)
tet38 (Δ)
50
6.25
100
50
10
5
10
10
20
10
20
10
2
2
1
1
1
1
.06
.25
norB (++)
12.5
50
10
10
10
10
2
1
1
.125
.5
norB (Δ)
norC (++)
12.5
12.5
50
50
10
10
10
10
10
10
10
10
2
2
1
1
1
1
.125
.125
.5
.5
norC (Δ)
12.5
50
10
10
10
10
.125
.5
norD (++)
norD (Δ)
12.5
12.5
50
50
10
10
10
10
10
10
10
10
.125
.125
.5
.5
.5
.5
2
2
.25
.5
.25
1
1
.125
2
.5
1
1
.5
2
Abbreviations: Hoe, Hoechst 33 342; MDR, multidrug resistance; MIC, minimum inhibitory contribution; Palmitol, palmitoleic acid; Putr, putrescine; Spe, spermine;
Sped, spermidine; Tetra, tetracycline; Unde, undecanoic acid.
pump, overexpression of which increased the MICs of all fatty
acids, except palmitic acid, 4- to 8-fold, and the MIC of tetracycline 64-fold. Relative to the parental strain, the tet38 mutant
had a limited change of MIC of 2-fold or less for all compounds, suggesting limited expression of tet38 under basal conditions in vitro. Overexpression of norB showed a more limited
effect, with a 2- to 4-fold increase in the MIC of fatty acids
except palmitic and palmitoleic acids. No change was seen in
MICs for the norB mutant except for a 2fold decrease with undecanoic acid. There was no change in the fatty acid MICs in
the other mutants.
In strain MW2 overexpression of tet38 from plasmid pLI50tet38 caused an increase of 2-fold and 4-fold in fatty acid MICs,
and 16-fold in tetracycline MIC (Table 2). The lower magnitude of increases in MICs with plasmid LI50-tet38 in MW2
than in RN6390, is likely related to the 4-fold lower level of
tet38 transcripts found in MW2 ( pLI50-tet38) than in RN6390
( pLI50-tet38; data not shown).
We determined the MICs of spermidine, putrescine, and
spermine for RN6390 and MW2 and their pump mutants.
Only the RN6390 norC mutant showed a decrease (4-fold) in
MICs to both polyamines. Overexpression of norC in RN6390
1488
•
JID 2014:209 (1 May)
•
Truong-Bolduc et al
and MW2 resulted in an 8-fold increase in MICs of spermidine,
putrescine, and spermine.
Effects of Fatty Acids and Polyamines on Efflux Pump Gene
Expression
The tet38 expression has been shown to increase in subcutaneous abscesses relative to growth in laboratory conditions [20].
We thus tested the ability of fatty acids, which are also present
in abscesses, to induce expression of tet38 (Table 3). Exposure
of RN6390 to sub-MIC concentrations of palmitoleic, oleic, linoleic, and undecanoic acids resulted in increases in tet38 transcript levels of 2-fold or more, with the greatest effect seen with
palmitoleic acid (16-fold at 0.25 × MIC = 10 µM). Only undecanoic acid at 0.25-fold MIC (=17.5 µM) caused an increase in
norB transcript levels (3.3-fold). In contrast, exposure of
RN6390 to spermidine at 0.25-fold MIC caused no change in
norC transcript levels.
A 2-fold or less change in tet38 transcription was found following exposure of MW2 to fatty acids, and no change was
seen following polyamine exposure (Table 3). Basal transcript
levels of tet38 in RN6390 and MW2 without fatty acid exposure
were similar, however. Differences in genetic background between
Table 3. Overexpression of MDR Efflux Pump Transcripts of
Staphylococcus aureus RN6390 and MW2 Under Induction by
Fatty Acids
MDR Efflux
Pumps (Fold
Change)a
Fatty acids
Fold × MIC
tet38
norB
1
1
1
1
S. aureus RN6390
Saturated fatty acids
Palmitic acid
0.125 × MIC
0.25 × MIC
Lauric acid
0.125 × MIC
1.3
1.2
0.25 × MIC
1.5
1.5
Unsaturated fatty acids
Palmitoleic acid
0.125 × MIC
5.4
0.5
Oleic acid
0.25 × MIC
0.125 × MIC
16
1.5
0.25
1
2
1.5
2.5
2.5
1.5
1.5
0.25 × MIC
Linoleic acid
0.125 × MIC
0.25 × MIC
Fatty acid with antimicrobial activity against S. aureus
Undecanoic acid
0.125 × MIC
0.25 × MIC
1.5
4.6
1.1
3.3
S. aureus MW2
Saturated fatty acids
Palmitic acid
0.125 × MIC
1
1
0.25 × MIC
1
1
0.125 × MIC
0.25 × MIC
1
1
1
1
Palmitoleic acid
0.125 × MIC
0.25 × MIC
2
2
1
1
Oleic acid
0.125 × MIC
1.2
1
Linoleic acid
0.25 × MIC
0.125 × MIC
1.5
1.5
1
1
0.25 × MIC
1.5
1
Lauric acid
Unsaturated fatty acids
Fatty acid with antimicrobial activity against S. aureus
Undecanoic acid
0.125 × MIC
2
0.25 × MIC
2
1
1
Gene gmk was used as internal control. Each assay was done in triplicate.
Abbreviations: MDR, multidrug resistance; MIC, minimum inhibitory contribution.
a
Fold change, ratio of transcripts with vs without induction by fatty acids.
RN6390 and MW2, which differ in sigB, rsbU, and other genes,
could play a role in the differences in fatty acid induction of
tet38 transcripts between these 2 strains.
Tet38-mediated Transport of Hoechst 33 342, Fatty Acids, and
Tetracycline in Everted Membrane Vesicles
The S. aureus NorA efflux pump heterologously expressed in
E. coli has been shown to transport the quinolone norfloxacin
in everted membrane vesicles [27]. To determine if Tet38
Figure 1. A, Hoechst 33 342 transport in Tet38-incorporated everted
membrane vesicles energized by lactate (0.5 mM). The binding capacity of
membrane vesicles was determined by monitoring fluorescence in the
presence of vesicles and in increasing concentrations of Hoechst 33 342.
Addition of the protonophore nigericin (2.7 µM) reversed the transport of
Hoechst. The arrows indicate the addition of lactate or nigericin to the
mixture. The squares represent the control assay using everted membrane
vesicles prepared from E. coli BL21 ( pTrcHis2). Each assay was done in
triplicate. B, Double reciprocal plots of Hoechst 33 342 transport in Tet38incorporated everted membrane vesicles. The increasing concentrations of
Hoechst 33 342 between 50 and 200 nM produced linear increases in fluorescence signal, reflecting linear membrane binding capacity. The kinetic
of Hoechst 33 342 transport by Tet38 was saturable, with an estimated
Vmax of 100 nmol/mg/min and an estimated Km of 0.40 µM.
functions as an efflux transporter of fatty acids and tetracycline,
we expressed and tested Tet38 in the E. coli everted membrane
vesicle system. The Hoechst 33 342 dye, which selectively
fluoresces in the membrane environment and to which tet38
conferred resistance (Table 2), was transported from everted
membrane vesicles prepared from Tet38-expressing cells as
measured by a lactate-dependent reduction in fluorescence.
Initial (20 seconds) rates of transport of Hoechst 33 342 were
linear within the concentration range of 50–200 nM and were
dependent on the presence of tet38. There was no change in
Hoechst 33 342 fluorescence in a vesicle preparation from E. coli
BL21( pTrcHis2) cells (Figure 1A). Hoechst 33 342 transport by
Tet38 was saturable with an estimated Km of 0.40 µM, a value
similar to that of transport by NorA [27]. Vmax was estimated
to be 100 nmol/mg vesicle protein/min.
We measured the ability of fatty acids and tetracycline to
compete with transport of Hoechst 33 342. Tetracycline produced
Role of Efflux Pumps in S. aureus Survival
•
JID 2014:209 (1 May)
•
1489
Figure 2. Competition for Hoechst 33 342 transport in Tet38-incorporating everted membrane vesicles by tetracycline, palmitoleic acid, undecanoic acid,
and palmitic acid. Each assay was done in triplicate. A, Double reciprocal plots of inhibition of Hoechst 333 42 transport in Tet38-incorporating everted
membrane vesicles by tetracycline. The reaction has an apparent Ki of 5.2 µM. Squares, Tetracycline = 70 nM (0.25-fold MIC); triangles, Tetracycline = 35 nM
(0.125-fold MIC); circles, no drug. B, Double reciprocal plots of inhibition of Hoechst 33 342 transport in Tet38-incorporating everted membrane vesicles by palmitoleic acid. The reaction has an apparent Ki of 9 µM. Squares, Palmitoleic = 10 µM (0.25-fold MIC); triangles, Palmitoleic = 5 µM (0.125-fold MIC); circles, no
drug. C, Double reciprocal plots of inhibition of Hoechst 33 342 transport in Tet38-incorporating everted membrane vesicles by undecanoic acid. The reaction
has an apparent Ki of 13.3 µM. Squares, Undecanoic = 17.5 µM (0.25-fold MIC); triangles, Undecanoic = 8.75 µM (0.125-fold MIC); circles, no drug. D,
Absence of inhibition of Hoechst 33 342 transport in Tet38-incorporating everted membrane vesicles by palmitic acid. The double reciprocal plots show no variation. Squares, Palmitic = 200 µM (5-fold MIC); triangles, Palmitic = 40 µM (1-fold MIC); circles, no drug. Abbreviation: MIC, minimum inhibitory concentration.
competitive inhibition of Hoechst 33 342 transport with an apparent Ki = 5.2 µM, providing direct evidence that tetracycline
is a substrate of Tet38 (Figure 2A). Palmitoleic and undecanoic
acids, which had increases in MICs with overexpression of
tet38, exhibited patterns of competitive inhibition with apparent Ki values of 9 µM and 13.3 µM, respectively (Figure 2B and
2C). Both fatty acids also exhibited complete transport inhibition at their respective MICs (40 and 70 µM) (data not shown).
Palmitic acid, resistance to which was not conferred by tet38,
exhibited no inhibition of Tet38-mediated transport of Hoechst
33 342 at 1- and 5-fold MICs (40 and 200 µM; Figure 2D).
Thus, Tet38 functions as a selective transporter of palmitoleic
and undecanoic acids as well as tetracycline.
in the ability of these strains to adhere to skin, as determined
by recovery of bacteria 1 hour after skin exposure (2.3 × 104
and 2.2 × 104 CFU/mouse for MW2 and MW2 Δtet38,
Mouse Skin Colonization Model
Skin survival was tested with strain MW2, a USA400 isolate
that reliably produces abscesses after subcutaneous injection
[19, 20]. After 24 hours of skin surface exposure, recovery of the
MW2 Δtet38 (3.6 × 103 CFU/mouse) was 5-fold less than that
of the parental strain (1.7 × 104 CFU/mouse; Figure 3). Complementation of tet38 in the mutant by plasmid-encoded tet38
resulted in survival similar to that of the parental strain
(2.0 × 104 CFU/mouse). This effect was not due to differences
1490
•
JID 2014:209 (1 May)
•
Truong-Bolduc et al
Figure 3. In vivo model to study the antibacterial activity of fatty acids
on mouse skin. An initial inoculum of 107 Staphylococcus aureus MW2
and the mutants MW2Δtet38, MW2ΔnorB, MW2ΔnorC were applied on
the mouse skin. The surviving bacteria were recovered after 24 hours. A 5fold reduction in the number of recovered S. aureus was found associated
with the mutant MW2Δtet38. MW2: wild-type S. aureus strain; MW2Δtet38,
MW2ΔnorB, MW2ΔnorC: tet38, norB, and norC isogenic mutants Each assay
was repeated 8 times (P < .00002). Abbreviation: CFU, colony-forming unit.
respectively) but appears instead to be due to differences in
their ability to survive on skin surface over time. In contrast,
norB and norC mutants of MW2 showed no difference in survival on skin at 24 hours (1.75 × 104 and 1.76 × 104 CFU/
mouse, respectively; Figure 3). We also tested RN6390 for survival on mouse skin. Although its Δtet38 mutant (QT7) also
showed somewhat lower survival on skin relative the parental
strain, the parental strain itself, which in contrast to MW2 is
unreliable at forming subcutaneous abscesses, survived poorly
on mouse skin, with a 10-fold reduction in CFU between 1 and
24 hours and could not be used reliably to test the role of tet38
in skin colonization.
DISCUSSION
Studies of efflux pumps of S. aureus have been largely focused
on those that confer resistance to antimicrobial agents [33, 34],
but their natural functions are poorly understood. Tet38 and
NorB, which had the strongest effects on resistance to antibacterial fatty acids, were originally identified by their antimicrobial
resistance phenotypes, to tetracycline and to fluoroquinolones,
respectively [25]. Interestingly, their expression is coordinately
regulated by the MgrA transcriptional activator/repressor. The
Nor efflux pumps and to a lesser extent Tet38 have broad substrate profiles, suggesting the possibility that their ability to
confer resistance to antimicrobials may in a number of cases be a
secondary or coincidental function. Their potential nonresistance natural functions are highlighted by prior findings that in
the absence of antimicrobials 3 S. aureus drug-resistance pumps,
Tet38, NorB, and NorD, selectively contribute to survival in subcutaneous abscesses [19, 20], possibly triggered by the ability of
low free iron, low pH, or reduced oxygen tension to induce their
expression [19, 30, 35].
One natural function of efflux pumps is thought to be as
“membrane vacuum cleaners” that remove harmful substances
from the membrane, a function that could be facilitated in
some cases by their broad substrate profiles [36–38]. Substrates
appear to be able to enter pumps from both the membrane
itself as well as from the cytosol [27]. The interaction of long
chain fatty acids with the bacterial membrane led us to postulate that some native efflux pumps may have antibacterial fatty
acids as natural substrates and thus may contribute to the
ability of S. aureus to survive in the abscess environment and
on the skin surface. Of the 5 pumps tested, Tet38 conferred resistance (5-fold) to a range of antibacterial fatty acids, with the
greatest effect for palmitoleic and undecanoic acids. NorB had
a weaker (2- to 4-fold) and more limited effect, and the other
pumps tested had no demonstrable ability to confer resistance.
Our findings further indicate that the fatty acid resistance phenotype of Tet38 is attributable to its ability to transport substrate fatty acids, a mechanism of resistance not previously
reported for antibacterial fatty acids in S. aureus. Efflux pump
systems such as the AcrAB of E. coli or FarAB of gonococci,
were previously shown to contribute to fatty acids resistance
[39, 40]. We also demonstrated directly that Tet38 functions as
a transporter of tetracycline. The substrates of Tet pumps have
generally been thought to be limited to members of the tetracycline class of antimicrobials, and tet genes have been most commonly found on plasmids. Our findings further suggest that
chromosomally encoded Tet pumps, like Tet38, may in fact
have other natural functions.
The expression of both tet38 and norB in strain MW2 is
induced in abscesses [20], which contain fatty acids, including
palmitoleic and undecanoic acids [5]. In the case of MW2,
however, induction of tet38 by fatty acids in vitro is poor, and
other environmental triggers for gene expression, such as low
pH (skin pH = 5.5) and low oxygen, had no effect on tet38 transcript level (data not shown). These triggers found in abscesses
have been shown to be inducers of norB [30, 35]. In contrast, in
strain RN6390 fatty acids induced expression of tet38 and to a
lesser extent norB, indicating that regulatory mechanisms and
induction pathways may differ among strains. Fatty acids are
the first identified natural compound inducers of tet38, which
is not induced by exposure to subinhibitory concentrations of
tetracycline [25]. Palmitoleic and undecanoic acids were the
strongest inducers of tet38, and linoleic acid had a weaker effect
in strain RN6390.
Linoleic acid induced tet38, mmpL, and mdeA pump gene
expressions in strain MRSA252 [21]. This induction was not
seen in RN6390 (data not shown), but no role for MmpL and
MdeA in resistance to fatty acids in S. aureus has been demonstrated [41]. Internalization of S. aureus into epithelial cells is
also associated with increased expression of tet38 [42]. Fatty
acids, such as sapienic and linoleic acids, may also induce expression of other genes such as proteases associated with increased severity of skin infection [43]. Following exposure to
fatty acids, the transcript levels of fadABDE (fatty acid degradation operon) were measured by RT-PCR. No significant increase was found (data not shown).
Spermidine and spermine are polyamines that are present on
human skin and like antibacterial fatty acids have activity against
S. aureus [12, 13]. Strain USA300 is a notable exception with relative resistance due to a polyamine-modifying N-acetyltransferase
encoded on the arginine catabolic mobile element [18]. Mouse
skin normally contains low levels of the catabolic enzyme spermidine/spermine-N-acetyltransferase (SSAT), a rate-limiting enzyme of the polyamine metabolism [44]. Efflux pump expression
appeared to have little effect on susceptibility to polyamines, with
the exception that overexpression of norC caused an 8-fold increase in resistance to polyamines in S. aureus RN6390. Notably,
expression of norC, like that of tet38 and norB, is coordinately
regulated by MgrA [26]. In contrast to the ability of Tet38 substrate fatty acids to induce expression of tet38, expression of norC
was not affected by subinhibitory concentrations of polyamines.
Role of Efflux Pumps in S. aureus Survival
•
JID 2014:209 (1 May)
•
1491
Because the roles of Tet38 and NorB in promoting bacterial
survival in an abscess environment have been established, we
tested their contribution and that of NorC to the ability of
S. aureus to colonize skin surfaces, a necessary step prior to
skin infection. In the adapted mouse model, tet38 but not norB
or norC mutants exhibited a 5-fold reduction in survival of
strain MW2 on skin. The reduction in skin survival in the tet38
mutant was fully complemented by plasmid-encoded tet38.
Thus, the Tet38 transporter itself contributes to the ability of
S. aureus to survive on skin, and this effect is likely due to its
ability to confer resistance to antibacterial fatty acids. The lack
of effect seen in the norB and norC mutants suggests that these
pumps do not play major roles in the ability of S. aureus to colonize mouse skin. Colonization of mouse skin was tested with
strain MW2, a clinically derived community MRSA strain that
readily forms subcutaneous abscesses, because RN6390, a laboratory strain that does not reliably form abscesses, also survived
poorly on mouse skin. Although fatty acids did not induce expression of tet38 in strain MW2, it is possible that other as yet
undefined conditions on the skin surface enhance tet38 expression. Our findings further indicate that fatty acid substrate induction is not required for the fitness benefit of tet38 in survival
on mouse skin.
The emergence of resistance to current antimicrobial agents
highlights the importance of discovery and development of new
approaches. Our findings raise the possibility that agents inhibiting Tet38 and possibly other efflux pumps might serve as potentiators of antibacterial fatty acids and might be used to
enhance the activity of regimens used to decolonize the skin of
patients with recurrent staphylococcal infections. As is the case
with NorB [20], Tet38 combines the ability to confer resistance
to antimicrobials with the ability to facilitate S. aureus survival
in abscesses and on skin.
Notes
Acknowledgments. The authors thank Irene Kochevar, Tom Gisel, and
Frank Doyle (Massachusetts General Hospital, Wellman Center for Photomedicine, Boston, MA) for the use of their spectrofluorometer. They thank
Abhisek Routray (Channing Laboratory, Boston, MA) for his assistance
with the French press. They thank Yin Wang (Massachusetts General Hospital, Infectious Diseases Units, Boston, MA) for her technical support.
Financial support. This work was supported by the National Institutes
of Health, Public Health Service grant [R37-AI23988 to D. C. H].
Potential conflicts of interest. All authors: No reported conflicts.
All authors have submitted the ICMJE Form for Disclosure of Potential
Conflicts of Interest. Conflicts that the editors consider relevant to the
content of the manuscript have been disclosed.
References
1. Drake DR, Brogden KA, Dawson DV, Wertz PW. Thematic review
series: skin lipids. Antimicrobial lipids at the skin surface. J Lipid Res
2008; 49:4–11.
2. Wang CS, Hartsuck J, McConathy WJ. Structure and functional properties of lipoprotein lipase. Biochim Biophys Acta 1992; 1123:1–17.
1492
•
JID 2014:209 (1 May)
•
Truong-Bolduc et al
3. Parsons JB, Yao J, Frank MW, Jackson P, Rock CO. Membrane disruption by antimicrobial fatty acids releases low-molecular-weight proteins
from Staphylococcus aureus. J Bacteriol 2012; 194:5294–304.
4. Shryock TR, Kapral FA. The production of bactericidal fatty acids
from glycerides in staphylococcal abscesses. J Med Microbiol 1992; 36:
288–92.
5. Dye ES, Kapral FA. Characterization of a bactericidal lipid developing
within staphylococcal abscesses. Infect Immun 1981; 32:98–104.
6. Kelsey JA, Bayles KW, Shafii B, McGuire MA. Fatty acids and monoacylglycerols inhibit growth of Staphylococcus aureus. Lipids 2006;
41:951–61.
7. Desbois AP, Smith VJ. Antibacterial free fatty acids: activities, mechanisms of action and biotechnological potential. Appl Microbiol Biotechnol 2010; 85:1629–42.
8. Sanderson NM, Guo BQ, Jacob AE, Handley PS, Cunniffe JG, Jones
MN. The interaction of cationic liposomes with the skin-associated
bacterium Staphylococcus epidermidis: effects of ionic strength and
temperature. Biochim Biophys Acta Bio-Membr 1996; 1283:207–14.
9. Nakatsuji T, Kao MC, Fang JY, et al. Antimicrobial property of lauric
acid against Propionibacterium acnes: its therapeutic potential for inflammatory acne vulgaris. J Invest Dermatol 2009; 129:2480–8.
10. Knapp HR, Melly MA. Bactericidal effects of polyunsaturated fatty
acids. J Infect Dis 1986; 154:84–94.
11. Dolezalova M, Janis R, Svobodova H, Kasparkova V, Humpolicek P,
Krejc J. Antimicrobial properties of 1-monoacylglycerols prepared from
undecanoic (C11:0) and undecenoic (C11:1) acid. Eur J Lipid Sci
Technol 2010; 112:1106–14.
12. Baze PE, Milano G, Verrando P, Renee N, Ortonne JP. Distribution of
polyamines in human epidermis. Br J Dermatol 1985; 112:393–6.
13. Baze PE, Milano G, Verrando P, Renee N, Ortonne JP. Polyamine levels
in normal human skin: a comparative study of pure epidermis, pure
dermis, and suction blister fluid. Arch Dermatol Res 1983; 275:218–21.
14. Zhang M, Wang H, Tracey KJ. Regulation of macrophage activation
and inflammation by spermine: a new chapter in an old story. Crit Care
Med 2000; 28:N60–6.
15. Greenway DL, Dyke KG. Isolation and properties of a linoleic acidresistant mutant of Staphylococcus aureus. J Gen Microbiol 1980; 118:
267–70.
16. Clarke SR, Mohamed R, Bian L, et al. The Staphylococcus aureus
surface protein IsdA mediates resistance to innate defenses of human
skin. Cell Host Microbe 2007; 1:199–212.
17. Kohler T, Weidenmaier C, Peschel A. Wall teichoic acid protects Staphylococcus aureus against antimicrobial fatty acids from human skin. J
Bacteriol 2009; 191:4482–4.
18. Joshi GS, Spontak JS, Klapper DG, Richardson AR. Arginine catabolic
mobile element encoded speG abrogates the unique hypersensitivity of
Staphylococcus aureus to exogenous polyamines. Mol Microbiol 2011;
82:9–20.
19. Ding Y, Fu Y, Lee JC, Hooper DC. Staphylococcus aureus NorD, a
putative efflux pump coregulated with the opp1 oligopeptide permease, contributes selectively to fitness in vivo. J Bacteriol 2012; 194:
6586–93.
20. Ding Y, Onodera Y, Lee JC, Hooper DC. NorB, an efflux pump in
Staphylococcus aureus MW2, contributes to bacterial fitness in abscesses. J Bacteriol 2008; 190:7123–9.
21. Kenny JG, Ward D, Josefsson E, et al. The Staphylococcus aureus response to unsaturated long chain free fatty acids: survival mechanisms
and virulence implications. PLoS ONE 2009; 4:e4344.
22. Manna A, Cheung AL. Characterization of sarR, a modulator of sar expression in Staphylococcus aureus. Infect Immun 2001; 69:885–96.
23. Hsieh PC, Siegel SA, Rogers B, Davis D, Lewis K. Bacteria lacking a
multidrug pump: a sensitive tool for drug discovery. Proc Natl Acad Sci
U S A 1998; 95:6602–6.
24. Ng EY, Trucksis M, Hooper DC. Quinolone resistance mediated by
norA: physiologic characterization and relationship to flqB, a quinolone
resistance locus on the Staphylococcus aureus chromosome. Antimicrob
Agents Chemother 1994; 38:1345–55.
25. Truong-Bolduc QC, Dunman PM, Strahilevitz J, Projan SJ, Hooper
DC. MgrA is a multiple regulator of two new efflux pumps in Staphylococcus aureus. J Bacteriol 2005; 187:2395–405.
26. Truong-Bolduc QC, Strahilevitz J, Hooper DC. NorC, a new efflux
pump regulated by MgrA of Staphylococcus aureus. Antimicrob Agents
Chemother 2006; 50:1104–7.
27. Yu JL, Grinius L, Hooper DC. NorA functions as a multidrug efflux
protein in both cytoplasmic membrane vesicles and reconstituted proteoliposomes. J Bacteriol 2002; 184:1370–7.
28. Sau S, Sun J, Lee CY. Molecular characterization and transcriptional
analysis of type 8 capsule genes in Staphylococcus aureus. J Bacteriol
1997; 179:1614–21.
29. Niemeyer DM, Pucci MJ, Thanassi JA, Sharma VK, Archer GL. Role
of mecA transcriptional regulation in the phenotypic expression of
methicillin resistance in Staphylococcus aureus. J Bacteriol 1996; 178:
5464–71.
30. Truong-Bolduc QC, Liao C-H, Villet R, et al. Reduced aeration
affects the expression of the NorB efflux pump of Staphylococcus
aureus by posttranslational modification of MgrA. J Bacteriol 2012;
194:1823–34.
31. Truong-Bolduc QC, Hooper DC. Phosphorylation of MgrA and its
effect on expression of the NorA and NorB efflux pumps of Staphylococcus aureus. J Bacteriol 2010; 192:2525–34.
32. Kugelberg E, Norstrom T, Petersen TK, Duvold T, Andersson DI,
Hughes D. Establishment of a superficial skin infection model in mice
by using Staphylococcus aureus and Streptococcus pyogenes. Antimicrob
Agents Chemother 2005; 49:3435–41.
33. Poole K. Efflux-mediated resistance to fluoroquinolones in gramnegative bacteria. Antimicrob Agents Chemother 2000; 44:2233–41.
34. Poole K. Efflux-mediated resistance to fluoroquinolones in grampositive bacteria and the mycobacteria. Antimicrob Agents Chemother
2000; 44:2595–9.
35. Truong-Bolduc QC, Bolduc GR, Okumura R, et al. Implication of the
NorB efflux pump in the adaptation of Staphylococcus aureus to growth
at acid pH and in resistance to moxifloxacin. Antimicrob Agents Chemother 2011; 55:3214–9.
36. Konings WN. The cell membrane and the struggle for life of lactic acid
bacteria. Antonie Van Leeuwenhoek - Int J Microbiol 2002; 82:3–27.
37. Konings WN, Albers SV, Koning S, Driessen AJM. The cell membrane
plays a crucial role in survival of bacteria and archaea in extreme environments. Antonie Van Leeuwenhoek - Int J Microbiol 2002; 81:61–72.
38. Konings WN, Lolkema JS, Bolhuis H, Van Veen HW, Poolman B,
Driessen AJM. The role of transport processes in survival of lactic acid
bacteria: energy transduction and multidrug resistance. Antonie Van
Leeuwenhoek - Int J Microbiol 1997; 71:117–28.
39. Lee EH, Shafer WM. The farAB-encoded efflux pump mediates resistance of gonococci to long-chained antibacterial fatty acids. Mol Microbiol 1999; 33:839–45.
40. Rosenberg EY, Bertenthal D, Nilles ML, Bertrand KP, Nikaido H. Bile
salts and fatty acids induce the expression of Escherichia coli AcrAB
multidrug efflux pump through their interaction with Rob regulatory
protein. Mol Microbiol 2003; 48:1609–19.
41. Domenech P, Reed MB, Barry CE III. Contribution of the Mycobacterium tuberculosis MmpL protein family to virulence and drug resistance.
Infect Immun 2005; 73:3492–501.
42. Garzoni C, Francois P, Huyghe A, et al. A global view of Staphylococcus
aureus whole genome expression upon internalization in human epithelial cells. BMC Genomics 2007; 8:171.
43. Arsic B, Zhu Y, Heinrichs DE, McGavin MJ. Induction of the staphylococcal proteolytic cascade by antimicrobial fatty acids in community
acquired methicillin resistant Staphylococcus aureus. PLoS ONE 2012;
7:e45952.
44. Pegg AE, Casero RA Jr. Current status of the polyamine research field.
Methods Mol Biol 2011; 720:3–35.
Role of Efflux Pumps in S. aureus Survival
•
JID 2014:209 (1 May)
•
1493