Download - Wiley Online Library

v?&&J
IMMUNOLOGY AND
MEDICAL
MICROBIOLOGY
ELSEVIER
FEMS Immunology and Medical Microbiology
I3 ( 1996) 8 l-86
Coagulase-negative staphylococci isolated from two cases of
toxic shock syndrome lack superantigenic activity, but induce
cytokine production
G. Lina ‘. *, A. Fleer b, J. Etienne ‘, T.B. Greenland
” lGpl,nrtmrrtt
de Rrcherchr
en BactCriologie
Midicale
(UPR ES). Faclrlti
dr MCdrcine
‘, F. Vandenesch
Akis
Carrel.
Rue G~ti//awne
a
Porndin.
69372 LJY~ Cede.r 08, Frcm1.e
’ Eijhman- Wirzkler Laboratoy
’ Lrrboratoire
d’bnmuwlo~ie
of Medical
et Biologic
Microbiology
and Wilhelmina
3501 CA Utrecht.
The Nrther1and.s
Pulmantrire.
HGpitcd Curdiologiyw
Received 25 May 1995; revised I9 September
Chiidrerz ‘5 Hospital,
PO Bar IXOOY.
Lo~tis Prtrdel. BP L,wm Cedes 03. Frtrncu
1995: accepted
19 September
1995
Abstract
Two strains of Sruph$ococ~rcs
epidernzidis
isolated from patients with toxic shock symptoms have been reported to carry
genes related to S. uurer~s enterotoxins B and C by dot-blot hybridisation, although the corresponding superantigenic toxins
were not detected immunologically.
We here show that these strains produce no superantigens capable of stimulating
proliferation of human mononuclear leukocytes or rabbit splenocytes, and that no DNA homologous to the seb or set genes
can be detected by PCR. However, stimulation of human monocytes by whole killed bacteria induced dose-dependent
production of the cytokines TNFcr. IL-I /3 and IL-6. which may be responsible for the clinical symptoms in these patients.
Keword.s:
Tumor necrosis factor-m: Interleukin- 1p: Interleukin-6:
Staph$ocowu.s
1. Introduction
Coagulase-negative
staphylococci are increasingly
being recognised as important pathogens, especially
for patients in surgical, haemato-oncological
or intensive care units [I]. Stuphvlococcu.s epidermidis is
frequently responsible for nosocomial infections and
can cause shock symptoms which are clinically se-
Corresponding author. Present address: Laboratoire Central de
Microbiologic.
BItiment
10, Hapital Edouard Herriot. Place
d’Arsonva1, 69437 Lyon Cedex 03. France. Tel.: +33 72 1 I 07
62: Fax: +33 72 11 07 64.
09X%8244/96/$1
5.00 Q 1996 Federation
SSDI 09’8.8244(95)00087-9
of European
Microbiological
epidermidk
vere [X] in up to 30% of patients with bacteraemia;
half of these patients die [3].
Shock may be induced by several distinct mechanisms. Cases associated with Gram-negative bacteria
result from the overstimulation
of macrophages by
lipopolysaccharide
(LPS) present in the outer membrane. The activated macrophages produce excessive
amounts of cytokines such as tumor necrosis factor
alpha (TNFcx) and interleukins
1p and 6 (IL- 1 p;
IL-6) [4] which contribute to the shock symptoms.
Certain Gram-positive
bacteria can induce shock
through a similar mechanism where peptidoglycan
and teichoic acid cell wall components
stimulate
cytokine production by monocytes. The shock thus
Societies.
All rights reserved
G. Lina et al. / FEMS Immunology
82
and Medical
produced by S. aureus and S. epidermidis in experimental animals closely resembles that induced by
Escherichia coli [5,6]. Other Gram-positive
bacteria
induce shock through the production of extracellular
toxins with superantigenic
properties. These induce
polyclonal MHC-dependent
T cell proliferation with,
again, release of TNFo, IL-l p, IL-6 and y-interferon, generally accounting for the observed symptoms of toxic shock syndrome (TSS) [7,8].
Two strains of S. epidermidis isolated from patients whose symptoms fitted the CDC definition for
toxic shock syndrome (TSS) induced by superantigen [9] gave a positive signal on dot-blot hybridisation with oligoprobes recognising
S. aureus genes
coding for enterotoxins -B and -C (SEB and SEC),
but antisera to SEB or SEC failed to recognise
components in the culture supematants [lo]. An antigenitally
novel, but genetically related toxin was
envisaged. In this paper we re-examine the evidence
for toxin production by these two S. epidennidis
strains, and test the alternative hypothesis of cytokine induction in macrophages through cell wall
components.
2. Materials and methods
2.1. Bacterial
strains
S. epidennidis strains N860094 and N900057 derive from blood cultures of two patients with TSS
symptoms (fever, hypotension,
rash and multiple
organ failure leading to death in one case); strain 354
was used as an unrelated control. Defined strains of
S. aureus were used to evaluate enterotoxin production: RN450 (sea - , seb - , set - 1; FRIS6 (sea - ,
seb + , set-1);
FR1137 (sea-,
seb-,
secf).
Strains were stored at - 80°C until used.
2.2. PCR amplification
of seb and set
Genomic DNA was extracted from staphylococcal
cultures [ll] and used as template for amplification
using primers previously shown to be specific for
seb and set [ 12,131. Seb 1 (5’~GAGAGTCAACCAGATCCTAAACCAG-3’1,
seb2
(5’-ATACCAAAAGCTATTCTCATTTTCT-3’1,
set 1 (5’-
Microbio1og.v
13 119961
RI -86
GACATAAAAGCTAGGAATTT-3’)
and sec2 (5’AAATCGGATTAACATTATCC-3’1
were synthesised on a Applied Biosystems 39 1 DNA synthesiser
(Perkin-Elmer,
Montagny-le-Bretonneux,
France).
After amplification for 30 cycles (1 min denaturation
at 94°C 1 min annealing at 55°C and 1 min extension at 72°C) the products were analysed by electrophoresis through 1% agarose gels (Sigma, Saint
Quentin Fallavier, France).
2.3. Mitogenic
natants
actioity
qf staphylococcal
super-
Exoproteins were prepared from overnight cultures of staphylococci
in brain-heart
infusion by
precipitation with 5 ~01s. of 98% ethanol for 2 days
at 4°C. After drying at 2O”C, the precipitate was
resuspended in l/50 vol. pyrogen-free distilled water, filtered through Ultrafree UFCCTHK (100 kDa
cutoff) cartridges and concentrated on MC UFC3LGC (10 kDa cutoff) filters (both from Millipore,
Molstein, France).
Human peripheral blood mononuclear cells were
prepared from blood of healthy donors (Centre de
Transfusion,
Lyon. France)
on Ficoll Hypaque
(Lymphoprep, Nycone, Norway) gradients [ 141 and
resuspended in RPM1 1640 medium (Gibco, Grand
Island, NY) supplemented with 10% heat-inactivated
foetal calf serum at lo5 cells ml-‘. Individual wells
of 96-well microtitre plates were seeded with 100 ~1
of cell suspension and 100 ~1 of staphylococcal
supernatant to give final concentrations of 0, 1/ 100,
l/10 and l/l relative to the initial staphylococcal
culture. After incubation for 48 h at 37°C in humidified 5% CO? in air, I &i of [3H]thymidine (1 Ci
mM_‘; CEA, Saclay, France) in 10 ~1 was added to
each well and the culture was continued for 24 h.
Cells were harvested on glass-fibre filters and their
radioactivity was evaluated in a Packard scintillation
counter. Results are expressed as cpm k standard
error of the mean in triplicate cultures.
Alternatively,
splenocytes were prepared by gently teasing apart spleens aseptically removed from
New Zealand rabbits (ESD, ChPtillon sur Chalaronne,
France). After removal of aggregates by passage
through nylon wool, the suspensions were adjusted
to 10’ cells ml-’ and incubated with staphylococcal
supematant as described above.
G. Lina et al. / FEMS Immunology
2.4. Cytokine production after staphylococcal
lation of human monocytes
and Medical
3. Results
PCR amplification from bacterial DNA of the two
S. epidetmidis
strains associated with TSS using
primers specific for the seb and set genes from S.
epidermidis
Brain heart solution
S. epidermidis N900057
S. epidermidis
N870094
S. artreu.~ FRIS6
S. rrureus RN420
undiluted
l/100
l/10
undiluted
l/l00
l/10
undiluted
l/100
l/10
undiluted
l/100
l/IO
undiluted
Measures are the mean of three determinations.
set
seb
Fig. I. PCR amplification of staphylococcal
toxin genes srb and
SYCfrom S. rpidermidis
isolated from TSS patients and S. ~II~~IL~
reference strains. Lanes A. S. aureus RN450 (srh - : SK - ):
Lanes 3, S. aure~.< FRIS6 (seh + : src - 1: Lanes C, S. (twe~s
FRI 137 ( seb - ; WC+ ): Lanes D. S. epidennidis N900057; Lanes
E, S. epidermidis N870094. seb: amplification using primers seb
1 and seb 2: SK: amplification using primers set I and set 2.
aureus produced no detectable signal, whereas the
control S. aureus strains gave a clearly positive
result (Fig. I). This is in contradiction with observations that DNA from the S. epidermidis strains can
hybridise with oligonucleotide probes for these genes
[lo]. We attempted amplifications
where one of the
PCR primers for each gene was replaced by an
oligonucleotide
previously used as probe, and amplifications using lower annealing temperatures but no
positive signals were obtained (data not shown).
Mutations in the gene regions corresponding
to the
primers could nevertheless have prevented amplification, so we tested the biological activity of culture
supematants from the two strains of S. epidermidis.
incorporation
by rabbit splenocytes or human monocytes
isolated from TSS patients or S. awew reference strains
Stimulant added
83
13 f 19Y6181-86
stimu-
S. epidermidis colonies were freshly inoculated
into Muller-Hinton
broth and incubated at 37°C for
18 h. Harvested
cells were washed
X 3 in
phosphate-buffered
saline, pH 7.4, counted spectrometrically and inactivated by heating at 60°C for 30
min. Suspensions for stimulation of human monocytes were adjusted to the equivalent of 5 X 10’”
CFU ml-‘. Incubation
on blood agar plates confirmed the absence of living bacteria in the heated
suspensions.
Human peripheral blood monocytes
were prepared and incubated with the killed bacteria as described in [6]. Briefly, 100 ~1 of bacteria1 suspension
at the indicated concentrations
was added to IO6
adherent monocytes in 96-well flat-bottomed culture
plates. After 18 h incubation, the supernatants were
removed and centrifuged, then cytokine concentrations were evaluated by enzyme immunoassay
as
described in [14]. Measurements are expressed as ng
ml-’ of cytokine.
Table I
[‘H]Thymidine
Microbinlq~y
(PBMC)
incubated
with supernatants
from cultures
Rabbit splenocytes
(cpm f SEM )
PBMC
(cpm & SEM)
667 i 109
889 i 358
750 + 363
724 f 92
903 of- 62
837 + J8
775 + 94
5241 + 418
16217 + 510
24915 + 396
967 + 102
672 f 50
334+
6
I81 +
21
168+
9-8
208 f
8
224 +
6
191 +
16
119+
109
22s+ 17
1’67 f I87
8570 + 596
15395 * 1831
1x2+
lh
15x+
II
214+
16
cpm. counts per minute: SEM. standard
error of the mean. Strains: a\ in Fig. I
of S.
G. Linn et al. / FEMS Immunology and Medical Micmbiologv I3 f 1996) 81-86
84
Table 2
Induction of TNFa, IL-lb
and IL-6 expression
incubation with killed S. epidermidis
Medium 199
S. epidemidis
354
S. epidemidis N900057
S. epidemidis N870094
1x104
1 x 105
IX lob
IX 107
IX 101
IX 105
1 x IOh
I x 10’
1XlOl
1 x 105
IX IO6
IX IO’
by PBMC
TNF
IL-
I
IL-6
< < 0.2
2.3
4.8
10.7
15.6
2.0
3.6
7.1
15.7
2.3
4.0
7.7
15.0
0.3
16.3
26.1
34.3
36.7
19.8
25.4
32.0
36.8
23.0
7.0
31.2
36.2
0.2
8.2
8.2
13.2
15.7
7.6
8.5
13.3
18.3
5.6
13.6
12.7
13.6
on
Data are expressed in ng ml-’ as determined by enzyme immunoassay. Strains as in Fig. 1; S. epidermidis 354 is a reference
strain.
Superantigen expression leads to proliferation of
human peripheral
blood monocytes
or of rabbit
splenocytes. When tested in these systems the S.
epidemidis
supematants,
at any concentration,
caused no mitogenic response in the cells over controls. S. atireus FRIS6, as expected, produced a
vigourous proliferative
response, while S. aureus
RN450 which does not produce superantigen,
was
inactive (Table 1). We conclude that strains N900057
and N860094 of S. epidermidis do not produce
superantigens or mitogenic exotoxins.
Possible
induction
of cytokine
expression
in
monocytes by components of the bacterial cell wall
was therefore investigated
by incubating
human
monocytes
with suspensions
of killed washed S.
epidermidis. Table 2 shows that both TSS associated
strains and a control strain (354) of S. epidermidis
induce a dose-dependent
release of TNFcu, IL- 1p
and IL-6 into the supematant by human monocytes.
Very similar levels of release of each cytokine was
observed on stimulation by the same concentration
of all three strains tested.
4. Discussion
The mechanisms by which S. epidennidis causes
shock are at present unclear. It has been suggested
that it might produce a superantigenic
toxin, and
indeed it does secrete exoproteins and toxins such as
S-haemolysin, cytotoxins and Dnase which resemble
their counterparts from S. aureus [ 15,161. Enterotoxin-producing
isolates of S. epidermidis have been
reported from animal samples [ 171, and an initial
study of 7 human cases of CNS-related TSS suggested that only TSST- l-producing strains were involved [lS]. Later studies have, however, shown that
these strains neither produce TSST-1 toxin nor possess the homologous genes [ 19,201.
The S. epidermidis strains N900057 and N870094
were initially thought to have caused TSS in the
patients from which they were isolated through an
exoprotein
because their DNAs appeared to hybridise with oligoprobes specific for the S. aureu,s
seh and set genes. However, immunological
assays
detected no SEB or SEC in the culture supematants,
suggesting antigenic variation [ 101. We have re-examined the possible presence of seb or set DNA in
these strains by PCR, and find no signal even under
relaxed conditions. We therefore conclude that the
previous results obtained using oligoprobes were false
positives due to inadequate specificity of the method,
although dot-blot hybridisation has been widely used
for the detection of bacterial genes. including those
for TSST- 1 and enterotoxins [ 10,11,2 11.
It remained possible that our S. epidermidis isolates produced an unrelated superantigenic exotoxin.
so we tested supernatants
of mature cultures for
mitogenic activity on susceptible human and rabbit
cells. No mitogenic activity was observed at any
concentration,
providing strong evidence that these
strains do not secrete superantigenic
toxins, or indeed any other mitogenic substance.
An alternative mechanism for induction of TSS
involves interaction between bacterial cell wall components and monocytes. leading to the release of
cytokine mediators. Gram-negative
bacteria such as
E. coli induce shock through LPS in their outer
membrane. and S. epidermidis can cause apparently
identical experimental shock in rabbits [5]. In both
cases the percentage of animals suffering shock was
similar, and symptoms were accompanied by TNFQ
and IL- 1p release. Recently peptidoglycans, teichoic
acid and lipoteichoic acid from S. epidermidis cell
walls have been shown to induce cytokine production in human monocytes [6,14.22]. We show here
G. Lina et al. / FEMS Irnmur~olog~ and Medical
that whole
killed S. epidermidis
is a powerful inducer of TNFcu, IL- 1j3 and IL-6 expression in human adherent monocytes.
It might be questioned whether cytokine release in
the absence of superantigen
stimulation of T cells
can provide a sufficient stimulus for TSS. It has been
shown that immunisation against TNFQ protects animals from SEB-induced
TSS, and that TNFa and
IL-I /3 are major mediators of shock [7,8,23]. In
addition,
superantigen-induced
TSS requires the
presence of T cells, since nude mice are not susceptible, but T cell proliferation
is not required, since
cyclosporin treatment does not protect [24,25]. Release of cytokines.
especially TNFcr and IL-I p.
either through cell wall interactions with monocytes
or through superantigen stimulation of T cells, thus
appears to be a crucial element in the induction of
TSS.
It may be noted that S. e,uidernzidis strain 354
which was derived from a catheter-related
bacteremia without known TSS, was as efficient an
inducer of cytokine release as the two TSS-related
strains. The incidence of TSS in human patients and
in experimental animals is generally less than half of
the potential cases. Although the capacity to induce
excessive cytokine release appears necessary for the
induction of TSS by a bacterium, other factors must
intervene in the progression to clinical shock. S.
epidermidis cell wall components
appear to be a
sufficiently
powerful stimulus to cytokine release
from monocytes to explain clinical TSS associated
with this bacterium.
Acknowledgements
We are grateful to P. Schlievert for scientific
advice. and B. Jaulhac and S. Foumier for their
technical cooperation and assistance.
References
[I] Herwaldt. L.A. ( 1988) Coagulase-negative
staphylococci: important nosocomial pathogens. Curr. Opin. Infect. Dis. 1.
755-763.
[2] Fidalgo. S., Vazquez, F.. Mendoza. M.C., Perez. F. and
MCndez, F.J. ( 1990) Bacteremia due to Sfaph~lococc~ts rpi-
Microhiologx
13 f IYY61 81-86
85
microbiologic. epidemiologic, clinical and prognostic features. Rev. Infect. Dis. 12, 520-528.
[3] Martin, M.A.. Pfaller. M.A. and Wenzel. R.P. (1989) Coagulase-negative staphylococcal bacteremia. Mortality and hospital stay. Ann. Intern. Med. 110. 9-16.
[J] Freudenberg,
M.A. and Galanos,
C. (1990) Bacterial
lipopolysaccharides:
structure, metabolism and action. lnt.
Rev. Immunol. 6. 207-22 I.
[5] Wakabayashi. G., Gelfand. J.A., Jung. W.K., Connolly, R.J..
Burke. J.F. and Dinarello. C.A. (1991) Srupb~lococcus epidermidis induces complement activation, tumor necrosis factor and interleukin-1, a shock-like state and tissue injury in
rabbits without endotoxemia.
Comparison
to E.scherkhia
w/i. J. Clin. Invest. 87, 1925-1935.
161Timmerman. C.P.. Mattsson, E.. Martinez-Martinez. L.. De
Graaf. L.. Van Strijp. J.A.G., Verbrugh. H.A., Verhoef, J.
and Fleer. A. (1993) Induction of release of tumor necrosis
factor from human monocytes by staphylococci and staphylococcal peptidoglycans.
Infect, Immun. 61. 4167-4172.
[71 Parsonnet. J., Hickman, R.K., Eardley. D.D. and Pier, G.B.
(1985) Induction of human interleukin-I
by toxic shock
syndrome toxin-l. J. Infect. Dis. 151. 514-522.
[81 Fast. D.J.. Schlievert. P.M. and Nelson. R.D. (1989) Toxic
shock syndrome-associated
staphylococcal
and atreptococcal
pyrogenic toxins are potent inducers of tumor necrosis factor
production. Infect. Immun. 57. 291-294.
191 CDC (1980) Follow-up on toxic-shock syndrome. MMWR
29. 441-1-15.
1101 Jaulhac. B.. Bes. M., Bornstein. N., Piemont, Y.. Brun, Y.
and Fleurette. J. (19921 Synthetic DNA probes for detection
of genes for enterotoxins A. B. C. D, E and TSST-I in
staphylococcal
strains. J. Appl. Microbial. 72, 386-392.
II II Ausubel. F.M.. Brent. R.. Kingston, R.E.. Moore. B.D..
Seidman, J.G., Smith. J.A. and Struhl. K. (1987) Analysis of
DNA sequences by blotting and hybridization.
In: Current
Protocols in Molecular Biology (Janssen. K.. Ed.). Greene
Publishing Associates and Wiley-Interscience,
New York.
[I21 Wilson. I.G.. Cooper. J.E. and Gilmour. A. ( 1991) Detection
of enterotoxingenic
Srapltvlococcrrs LIII~~USm dried skimmed
milk: use of the polymerase chain reaction for amplification
and detection of staphylococcal
enterotoxin genes entE and
enrCI and the thermonuclease
gene nn~‘. Appl. Environ,
Microbial. 57. 1793-798.
[I31 Johnson. W.M., Tyler. SD.. Ewan. E.P.. Ashton. F.E., Pollard, D.R. and Rozee. K.R. (1991) Detection of genes for
enterotoxins.
exfoliative toxins and toxic shock syndrome
toxin 1 in Srtrph&~oc~cus nnrens by the polymerase chain
reaction. J. Clin. Microbioi. 29. 426-30.
[IdI Mattsson. E.. Verhage, L.. Rollof. J.. Fleer. A.. Verhoef. J.
and van Dijk. H. (1993) Peptidoglycan
and teichoic acid
from Stc~ph~loco~ws rpidennidis
stimulate human monocytes to release tumour necrosis factor-a. interleukin-I p and
interleukin-6. FEMS Immunol. Med. Microbial. 7. 28 I-288.
of coagulase-negative
[I51 Gemmell. C.G. (1987) Exo-proteins
staphylococci as possible virulence factors. Zentralbl. Bakteriol. A 266, 93- 102.
iI61 McKevitt. A.I., Bjornson. G.L., Mauracher. C.A. and
dermidis:
86
[17]
[18]
[ 191
[20]
[21]
G. Lina et al. / FEM.7 Immunolog_v and Medical Microbiology
Scheifele, D.W. (1990) Amino acid sequence of a deltalike
toxin from Staph$ococcus
epidermidis. Infect. Immun. 58.
1473-1475.
Marin, M.E., Del Carmen de la Rosa, M. and Comejo. I.
(1992) Enterotoxigenicity
of staphylococcus
strains isolated
from Spanish dry-cured hams. Appl. Environ, Microbial. 58.
1067- 1069.
Crass, B.A. and Bergdoll. M.S. (1986) Involvement of coagulase-negative
staphylococci
in toxic shock syndrome. J.
Clin. Microbial. 23, 43-45.
Parsonnet, J., Harrison, A.E., Spencer, S.E., Reading, A..
Parsonnet, KC. and Kass, E.H. (1987) Nonproduction
of
toxic shock syndrome toxin 1 by coagulase-negative
staphylococci. J. Clin. Microbial. 25. 1370-1372.
Kreiswirth, B.N., Schlievert, P.M. and Novick. R.P. (19871
Evaluation of coagulase-negative
staphylococci for ability to
produce toxic shock syndrome toxin 1. J. Clin. Microbial.
25, 2028-2029.
Neill, R.J., Fanning, G.R.. Delahoz, F., Wolff, R. and Gem-
[22]
[23]
[24]
[25]
13 (19%) 81-86
ski, P. (1990) Oligonucleotide
probes for detection and differentiation
of Staph$ococcus
aurerts strains containing
genes for enterotoxins A, B and C and toxic shock syndrome
toxin]. J. Clin. Microbial. 28, 1514-518.
Heumann, D., Barras, C.. Severin, A., Glauser. M.P. and
Tomasz. A. (1994) Gram-positive
cell walls stimulate synthesis of Tumor Necrosis Factor alpha and Interleukin-6 by
Human Monocytes. Infect. Immun. 62, 2715-2721.
Miethke, T.. Gaus, H., Wahl, C., Heeg. K. and Wagner. H.
(1992) T-cell-dependent
shock induced by a bacterial superantigen. Chem. Immunol. 55, 172-184.
Marrack. P., Blackman. M., Kushnir, E. and Kappler. J.
(1990) The toxicity of staphylccoccal
enterotoxin B in mice
is mediated by T cells. J. Exp. Med. 171, 455-464.
Lee, P.K., Deringer, J.R., Kreiswirth, B.N., Novick, R.P. and
Schlievert. P.M. (1991) Fluid remplacement
protection of
rabbits challenged
subcutaneously
with toxic shock syndrome toxins. Infect. Immun. 59, 879-884.