Subinhibitory cerulenin inhibits staphylococcal exoprotein production by blocking transcription rather than by blocking secretion
Rajan P. Adhikari and
Richard P. Novick
Molecular Pathogenesis Program, Skirball Institute of Biomolecular Medicine and Department of Microbiology, New York University Medical Center, 540 First Avenue, New York, NY 10016, USA
Correspondence
Richard P. Novick
novick{at}saturn.med.nyu.edu
 |
ABSTRACT
|
---|
Cerulenin is an antibiotic that inhibits fatty acid synthesis by covalent modification of the active thiol of the chain-elongation subtypes of
-ketoacyl-acyl carrier protein synthase. It also inhibits other processes that utilize essential thiols. Cerulenin has been widely reported to block protein secretion at sub-MIC levels, an effect that has been postulated to represent interference with membrane function through interference with normal fatty acid synthesis. This study confirms the profound reduction in extracellular proteins caused by low concentrations of the antibiotic, and shows by Northern blot hybridization that this reduction is due to interference with transcription. By exchanging promoters between entB, a gene that is inhibited by cerulenin, and entA, a gene that is not, it was also shown that the antibiotic does not block secretion. Subinhibitory concentrations of cerulenin were also found to block transcriptional activation of at least two regulatory determinants, agr and sae, that function by signal transduction. Interference with the activation of these and other regulatory determinants probably accounts for much of the inhibitory effect on exoprotein production of sub-MIC concentrations of cerulenin.
Abbreviations: AIP, autoinducing peptide; TSST-1, toxic shock syndrome toxin 1
 |
INTRODUCTION
|
---|
Cerulenin is an antibiotic with a broad spectrum of inhibitory activities, all or most of which can be attributed to its unusual mode of action, namely the inhibition of fatty acid synthesis (Goldberg et al., 1973
) and related processes, including sterol synthesis (Greenspan & Mackow, 1977
), protein acylation (Jochen et al., 1995
; Straub et al., 2002
), synthesis of polyketide antibiotics (Hiltunen & Soderhall, 1992
), synthesis of homoserine lactone autoinducers (Val & Cronan, 1998
), and synthesis of myristic aldehyde (Byers & Meighen, 1989
), the substrate of bacterial luciferases. Its common mechanism is covalent attachment to the active thiol group of one or another of the key enzymes in the affected pathways (Price et al., 2001
). Its effects on lipid synthesis have led to the idea that it perturbs the cytoplasmic membrane, with important consequences for certain membrane functions, especially the secretion of proteins (Jacques, 1983
). Thus subinhibitory concentrations of cerulenin have been reported to inhibit the secretion of staphylococcal
-haemolysin (Saleh & Freer, 1984
), enterotoxin B (EntB) (Altenbern, 1977
), and other superantigens, Bacillus subtilis
-amylase, protease and levansucrase (Mantsala, 1982
), yeast
-galactosidase (Martinez et al., 1982
) and Escherichia coli
-lactamase (Mantsala & Lehtinen, 1982
). Our interest in this phenomenon was occasioned by our recent demonstration that two staphylococcal superantigens, EntB and toxic shock syndrome toxin 1 (TSST-1), are autorepressors (Vojtov et al., 2002
), whereas a third, EntA, is not. Perhaps subinhibitory concentrations of cerulenin could help to determine whether other exoproteins are autorepressors: the precursor of an autorepressor would not accumulate intracellularly if secretion were blocked, whereas the precursor of a non-autorepressing exoprotein would accumulate.
We began the present study by revisiting the reported interference by subinhibitory concentrations of cerulenin with exoprotein secretion and found that the antibiotic blocks exoprotein synthesis and therefore would not be useful for the identification of autorepressors. Analysis of the effects of subinhibitory concentrations of cerulenin on gene fusions revealed further that the antibiotic does not block secretion after all, and that it blocks synthesis at the level of transcription. We find that many exoprotein genes are blocked by subinhibitory concentrations of cerulenin, a rare few are stimulated and others are unaffected, and that the effects of subinhibitory concentrations of cerulenin on exoprotein profiles vary dramatically among Staphylococcus aureus strains. Additionally, subinhibitory concentrations of cerulenin inhibit the transcription of certain global regulatory genes such as agr and sae, so its effects on exoprotein genes may be mediated via global regulators. Finally, since S. aureus pathogenesis depends on a large set of extracellular proteins, whose synthesis is controlled by global regulators, it was predicted that subinhibitory cerulenin would attenuate staphylococcal pathogenesis, and this was confirmed in a murine sepsis model.
 |
METHODS
|
---|
Bacterial strains and plasmids.
These are listed in Table 1
.
Media and growth conditions.
The standard culture medium was CYGP without glucose (Novick, 1991
). Cell growth was monitored with either a KlettSummerson colorimeter with a green (540 nm) filter (Klett) or with a THERMOmax microplate reader (Molecular Devices) at 650 nm (OD650). For determination of exoprotein profiles, culture samples (110 ml) were centrifuged in an Eppendorf centrifuge. The supernatant was recentrifuged to remove any residual organisms, then precipitated with a 10 % volume of 50 % trichloracetic acid, and the pellet analysed by SDS-PAGE according to the method of Laemmli (1970)
. All samples were equalized to the density of the culture. Western blots were carried out by a standard protocol by using anti-TSST-1 (Toxin Technology), anti-EntB and anti-EntA (Sigma-Aldrich) raised in rabbit as primary antibody, and anti-rabbit Ig with horseradish peroxidase (Amersham) as secondary antibody. Signals were detected by ECL (Amersham).
DNA procedures.
Amplification reactions (PCRs) were carried out by using different primers (see Table 2
). For switching promoters between entA and entB, primers containing PstI and BamHI restriction sites at the 5' end were designed. For cloning coding regions of entA and entB, primers containing BamHI and KpnI restriction sites at the 5' end were used. Chromosomal DNA isolated from strain S6C in the case of the entB gene, and RN8530 in the case of entA, was used as template. Plasmid and chromosomal DNA were isolated by using a QIAprep Spin Miniprep Kit from Qiagen. PCR products were purified by using a QIAquick PCR Purification Kit also from Qiagen. PCR products were digested with KpnI, PstI and BamHI and cloned into the multiple cloning site of pCN50 (Charpentier et al., 2004
). Clones in pCN50 were transformed into E. coli DH5
and then moved into RN4220 by electroporation (Novick, 1991
). All plasmids were transferred from S. aureus RN4220 to other S. aureus strains by standard transduction techniques by using phage 80
(Novick, 1991
).
Luciferase assay.
Strains with luciferase reporter were grown in CYGP broth. To 200 µl bacterial culture in a white microtitre plate was added 10 µl 1 % tetradecenal (Bedoukian Research) in ethanol. Bioluminescence was measured with a Molecular Devices Lmax384 luminometer, using a delay time of 3 s and an integration time of 5 s. Results are expressed as arbitrary light units (RLU) and all luciferase results are normalized to the cell content of the sample.
RNA preparation.
Cell pellets were treated with RNA Protect reagent (Qiagen) and mechanically disrupted by agitation with glass beads using the Bio 101 FastPrep Apparatus. RNA was purified using the Qiagen RNeasy kit, and its integrity checked by agarose gel electrophoresis (Weinrick et al., 2004
).
Northern blot hybridization.
RNA samples corresponding to equal numbers of cells were separated by gel electrophoresis through 1 % denaturing agarose (MOPS/formaldehyde), vacuum-blotted to Hybond-N+ membranes (Amersham), and UV cross-linked. Blots were hybridized overnight to [
32P]dATP-labelled, PCR-generated probes. Washed blots were exposed to phosphorimager screens which were read by a Molecular Dynamics Phosphorimager. Primers (Integrated DNA Technologies) are listed in Table 2
.
Murine infection model.
Bacterial samples, 1·5x108 c.f.u. (S. aureus strain LS-1), in 0·1 ml buffered normal saline, were administered by tail vein to 25 g Swiss Webster mice. Cerulenin (in ethanol) was added to the input inoculum to give 0, 5, 10 or 40 mg kg1. Cerulenin alone, 40 mg kg1, served as control for cerulenin toxicity. Mice were housed in microisolator cages and fed a standard diet with water ad libitum. They were observed for visible ill effects for 24 h.
 |
RESULTS
|
---|
Determination of subinhibitory concentrations of cerulenin
Since cerulenin inhibits bacterial growth at widely differing concentrations depending on the species and the growth conditions, we first analysed its effects on the growth of several different S. aureus strains in our standard CY medium. As shown in Fig. 1
, at 5 or 10 µg ml1, cerulenin does not affect the growth rate of our standard strain, RN6734, though it slightly reduces the final yield, whereas at higher concentrations, it sharply inhibits growth. We determined the MIC of RN6734 for cerulenin to be 125 µg ml1. The effects of subinhibitory cerulenin on other strains were similar, except that final yields were usually not reduced. Results for S6C, for example, are also shown in Fig. 1
. Accordingly, we used a concentration of 5 µg cerulenin ml1 added at a Klett reading of 50 (K50) for most of the experiments described below.
Effects of subinhibitory cerulenin on overall cytoplasmic and extracellular protein patterns and on superantigens
To evaluate interstrain variability in the effects of subinhibitory cerulenin on the production of cytoplasmic and extracellular proteins, and to confirm reported effects on the above-mentioned superantigens, we analysed post-exponential culture supernatants from four strains, RN6734, which does not produce any known superantigen (see Fig. 2a, b
, lanes 1 and 2), plus native producers of TSST-1, EntB and EntA, respectively, by SDS-PAGE, staining the gels with Coomassie brilliant blue. Whereas the cytoplasmic protein profiles of these strains were largely unaffected by cerulenin at 5 µg ml1, as shown in Fig. 2(a)
, the antibiotic generally caused a dramatic reduction in the number and intensity of the exoprotein bands, as shown in Fig. 2(b)
. Additionally, a few bands seemed unaffected by subinhibitory concentrations of cerulenin and a very few others seemed to be stimulated. The exoprotein profiles were similar for strains RN6734, RN4282 (TSST-1) and S6C (EntB), which are closely related, but quite different for RN8530 (EntA), which is not closely related to the other three. Note that EntB and TSST-1 are absent from the supernatants of cultures grown with subinhibitory concentrations of cerulenin. This was confirmed by Western immunoblotting (Fig. 2c
), which shows also that EntA is unaffected by the antibiotic.

View larger version (105K):
[in this window]
[in a new window]
|
Fig. 2. (a, b) Whole-cell protein (a) and exoprotein (b) profiles of CYGP broth cultures without glucose and with and without 5 µg cerulenin ml1 added at K50. Samples were taken after a further 6 h incubation (t6). M, PAGE ruler; lanes 18, samples without (lanes 1, 3, 5 and 7) and with (lanes 2, 4, 6 and 8) 5 µg cerulenin ml1 of strains RN6734 (lanes 1 and 2), RN4282 (lanes 3 and 4), S6C (lanes 5 and 6) and RN8530 (lanes 7 and 8). (c) Western blot analysis of extracellular protein in the presence and absence of 5 µg cerulenin (cer.) ml1 for TSST-1 in strain RN4282, EntB in strain S6C and EntA in strain RN8530. (d) Exoprotein profile of different S. aureus strains whose genome has been sequenced. Cultures were grown in CYGP broth without glucose and without (1, 3, 5, 7, 9, 11 and 13) and with (lanes 2, 4, 6, 8, 10, 12 and 14) 5 µg cerulenin ml1 added at K50. Samples were taken at t6. Lanes 1 and 2, strain NCTC 8325; lanes 3 and 4, strain COL; lanes 5 and 6, strain MRSA252; lanes 7 and 8, strain N315; lanes 9 and 10, strain Mu50; lanes 11 and 12, strain Sanger 476; lanes 13 and 14, strain MW2.
|
|
We next tested the effects of subinhibitory concentrations of cerulenin on a wider variety of staphylococcal strains, especially those whose genomes have been sequenced. These exoprotein profiles, shown in Fig. 2(d)
, again illustrate wide variation both in the presence and in the absence of subinhibitory concentrations of cerulenin. Note that our cultures of two of the strains, COL (lanes 3 and 4) and mu50 (lanes 9 and 10) are unexpectedy agr-defective (unpublished data). We attribute this to the frequent occurrence of agr defects in clinical isolates under laboratory conditions (Somerville et al., 2002
). Since many of the exoproteins whose synthesis is inhibited by subinhibitory cerulenin are agr upregulated, both the paucity of exoprotein bands and the relative lack of cerulenin inhibition in these strains are attributable to the agr defects. For MRSA252, although the exoprotein pattern is typical of S. aureus, there is also very little effect of cerulenin. We have no explanation for this except to suggest that the putative transducer of cerulenin inhibition may be deficient in this strain. A few of the unidentified bands that were blocked by subinhibitory concentrations of cerulenin were excised and identified by mass spectroscopy. These were shown to be a major extracellular antigen (Map), the major staphylococcal autolysin (Atl) and a serine protease-like protein (Spl); additionally, the band suspected to be EntB was confirmed. As the identities of most of the bands seen in these exoprotein profiles are unknown, an important initiative will be a definitive analysis of the extracellular proteome of S. aureus, with special reference to the effects of subinhibitory antibiotics.
Subinhibitory concentrations of cerulenin block exoprotein synthesis rather than secretion
Earlier reports of blockage of secretion by subinhibitory concentrations of cerulenin did not formally rule out blockage of synthesis, though one report suggested this as a possibility. In order to test the above hypothesis regarding autorepressors, we needed first to reinvestigate this possibility, which was done by switching the promoters and coding regions of two superantigen genes, entA and entB. As noted, EntB is not secreted in the presence of subinhibitory concentrations of cerulenin while EntA is secreted. Tests of the effects of subinhibitory concentrations of cerulenin on these two constructs, in comparison with the native genes, are shown in Fig. 3
(a): subinhibitory concentrations of cerulenin blocked the appearance of either EntA or EntB, when the respective genes were under entBp control, but not when they were under entAp control. In other words, the target of subinhibitory concentrations of cerulenin inhibition seemed to be entBp and not EntB itself, which, once made, is secreted in the presence of the drug. We noted that in these gels, with the entA promoter, there was less EntB in the supernatant in the presence than in the absence of cerulenin. This difference was not seen in an agr-null background (not shown); its basis is presently under study.

View larger version (88K):
[in this window]
[in a new window]
|
Fig. 3. Switching promoters and coding regions between entA and entB. Constructs were prepared in the RN6734 background and exoprotein samples were prepared from CYGP culture supernatants at t6 (a) without and (b) with 5 µg cerulenin ml1. M, protein size markers; lane 1, pRN9977; lane 2, pRN9978; lane 3, pRN9979; lane 4, pRN9980; lane 5, pRN9981.
|
|
Subinhibitory cerulenin blocks transcription
As noted above, subinhibitory cerulenin blocks synthesis of EntB, and probably also of TSST-1, but not EntA. To analyse the defect in synthesis, we performed a Northern blot hybridization analysis on whole-cell RNA prepared from cells grown with and without cerulenin. As shown in Fig. 4
(a), cerulenin at 5 µg ml1 greatly inhibited entB and tst transcription, but stimulated entA transcription, consistent with its inhibition of synthesis of the former two proteins but not of the latter. It is suggested, in conclusion, that subinhibitory cerulenin affects EntB and TSST-1 production primarily at the level of transcription.

View larger version (41K):
[in this window]
[in a new window]
|
Fig. 4. (a) Northern blot hybridization of strains RN4282 (lanes 1 and 2), S6C (lanes 3 and 4) and RN8530 (lanes 5 and 6), without (lanes 1, 3 and 5) and with (lanes 2, 4 and 6) 5 µg cerulenin ml1. Panel I, 16S rRNA, panel II, lanes 1 and 2, tst probe; lanes 3 and 4, entB probe; lanes 5 and 6, entA probe. (b) Effect of 5 µg cerulenin ml1 added at K50 on different global regulators and exoprotein gene transcription in strain RN6734 as shown by Northern blot hybridization. Cultures were grown in CYGP broth without glucose and samples were taken at 0, 3 and 6 h. Lanes: 1, without cerulenin at t0; 2, with 5 µg cerulenin ml1 at t0; 3, without cerulenin at t3; 4, with 5 µg cerulenin ml1 at t3; 5, without cerulenin at t6; 6, with 5 µg cerulenin ml1 at t6 probed by 16S rRNA, RNAIII, rot, sarA, sae, hla, sspA, splC, geh, fabF, map and spa.
|
|
Effect of subinhibitory cerulenin on transcription of other exoprotein genes
Given the above results, it seemed likely that the effects of subinhibitory cerulenin on other exoprotein genes would also be at the level of transcription. Northern blot hybridization analysis of several other genes encoding exoproteins previously found to be stimulated, inhibited or unaffected by cerulenin is shown in Fig. 4(b)
. As can be seen, the genes hla (
-haemolysin), splC, geh (glycerol ester hydrolase) and map were strongly inhibited, whereas the genes sspA (serine protease) and spa (protein A) were affected minimally. We also analysed one gene (fabF) encoding a cytoplasmic protein, fatty acid synthetase. This gene was chosen because of the known inhibition of its product by cerulenin. fabF is transcribed only very early during growth and was moderately stimulated by subinhibitory concentrations of cerulenin an effect that could be the result of a modest inhibition of fatty acid synthesis at the low cerulenin concentration used.
As there is no obvious direct connection between cerulenin and the transcription process, it seemed likely that subinhibitory concentrations of cerulenin were affecting one or more of the well-known global regulators that control transcription; in this connection, it seemed significant that many of the genes whose transcription patterns were examined are controlled by agr. Given the possibility that the antibiotic affects membrane function, regulatory genes that involve transmembrane signalling could well be affected. Accordingly, we Northern blotted these same RNA gels with probes specific for agr (RNAIII) and sae, both of which function by transmembrane signalling, and for rot and sarA, which do not. As can be seen in Fig. 4(b)
, subinhibitory concentrations of cerulenin blocked agr induction, most strongly at t6, blocked induction of the two larger sae transcripts at t3 and t6, and stimulated the sarA and rot transcripts at t6 (and slightly stimulated the sarB and C transcripts, also at t6). With the notable exception of sspA, these results are largely consistent with the known effects of mutations affecting the four regulatory genes on transcription of these exoprotein genes, supporting the conclusion that subinhibitory concentrations of cerulenin act largely or exclusively through regulatory genes. The effects of subinhibitory concentrations of cerulenin on sspA transcription are not understandable in terms of the known regulation of this gene since it is upregulated by agr and downregulated by sarA, it should have decreased sharply in the presence of subinhibitory concentrations of cerulenin. Further study of this apparent anomaly is planned.
agr inhibition by cerulenin
Since agr activation is required for sae activation, i.e. downregulation of sae would follow from downregulation of agr, we focussed on agr. We have shown above that subinhibitory concentrations of cerulenin inhibit transcription from the agrp3 promoter. The experiment reported in Fig. 5
(a) showed that subinhibitory concentrations of cerulenin also inhibit transcription from the agrp2 promoter. As both of these promoters are activated by the agr autoinduction circuit, which involves the synthesis and secretion of an autoinducing peptide (AIP), subinhibitory concentrations of cerulenin could affect either the synthesis of this peptide or the response to it through the agr signalling module. To determine whether subinhibitory concentrations of cerulenin inhibit either of these steps in agr activation, we cloned the agrBD module, responsible for AIP synthesis (Ji et al., 1997
), behind the entA promoter, which, as shown above, is insensitive to cerulenin. RN7206 containing this module was grown in the presence of cerulenin (10 µg ml1) or in the absence of the drug and its supernatant used to test for AIP-dependent activation of the AgrAC signalling module in the presence or absence of cerulenin at the same concentration, using the agrp3-lux construct as a reporter. The results of this test were that cerulenin at 10 µg ml1 did not detectably inhibit either the synthesis/secretion or the action of the AIP (not shown). Thus cerulenin must act by inhibiting a regulatory system that is upstream of agr and required for its activation. Since cerulenin is very unlikely to affect transcription by directly interacting with DNA, it is suggested that the drug must block the function of a pre-existing regulatory unit, presumably by inactivating a cerulenin target, such as fatty acid synthase. A simple test of this possibility was a test for reversal of cerulenin inhibition by fatty acids. As shown in Fig. 5(b)
, heptadecanoic acid partially reverses the inhibition of exoprotein production and of agr activation by cerulenin at 10 µg ml1, consistent with the blockage of an unidentified step in transcription requiring fatty acid synthesis.
Effects of subinhibitory concentrations of cerulenin on exoprotein production by regulatory gene mutants
Although it is clear from the preceding section that agr is an important target of cerulenin action with respect to the regulation of staphylococcal virulence, it is also clear, as shown for example in Fig. 6
, lanes 12, that agr is not the whole story here. For example, many bands appear in the agr-null profile, representing proteins whose synthesis is not agr-upregulated, and many of these are sensitive to subinhibitory concentrations of cerulenin. It thus seemed likely that subinhibitory concentrations of cerulenin would affect regulatory genes other than agr, and would affect them independently of agr. Although, as shown above (Fig. 3b
), subinhibitory concentrations of cerulenin block sae activation, this may be a direct consequence of agr inhibition. It also seemed likely that a mutation in any regulatory gene that was directly affected by subinhibitory concentrations of cerulenin would eliminate the effect of the antibiotic on the subset of exoproteins regulated by that gene (such an effect, of course, could be seen only with genes that were downregulated by the regulatory gene in question). Accordingly, we determined exoprotein profiles for a series of additional regulatory mutants in the presence and absence of subinhibitory concentrations of cerulenin. These results are shown in Fig. 6
, where it can be seen first that each of the regulatory mutations has a major effect on the exoprotein profile, that each has a unique effect, and that with one exception, subinhibitory concentrations of cerulenin have a substantial effect on each of these patterns. The exception is the arlR mutation, whose exoprotein profile is affected minimally by subinhibitory concentrations of cerulenin. Since arlRS downregulates norA, a multidrug exporter that could theoretically export cerulenin, this result could simply represent reduced sensitivity to cerulenin owing to enhanced export of the antibiotic. This possibility was ruled out by a determination of cerulenin MICs for arl-negative and arl-positive strains, which differed by twofold at most. Alternatively, it is possible that inhibition of arlRS could account for at least some of the cerulenin effects, since arlRS upregulates agr. A test of this possibility is planned. In order to be informative, this test would have to be of the effects of cerulenin on arlRS function, not of its effects on arlRS transcription.

View larger version (63K):
[in this window]
[in a new window]
|
Fig. 6. Exoprotein profile of different global regulator null mutants in CYGP broth culture without glucose and without (lanes 1, 3, 5, 7, 9 and 11) and with (lanes 2, 4, 6, 8, 10 and 12) 5 µg cerulenin ml1 added at K50. Samples were taken at t6. Lanes 1 and 2, strain RN7206; lanes 3 and 4, strain RN9388; lanes 5 and 6, strain RN9524; lanes 7 and 8, strain RN9822; lanes 9 and 10, strain RN9896; lanes 11 and 12, strain RN9808.
|
|
Prevention of murine lethality by subinhibitory concentrations of cerulenin
The profound effect of subinhibitory concentrations of cerulenin on the production of staphylococcal exoproteins, including toxins and other virulence factors, suggested that cerulenin might be able to interfere with bacterial infections at subinhibitory levels, even though it is not an attractive antibiotic at conventional inhibitory levels. Accordingly, we tested it for the attenuation of virulence with S. aureus strain LS-1, a mouse-adapted strain used widely for tests of virulence (Bremell et al., 1991
). Fig. 7
(a) shows the effects of subinhibitory concentrations of cerulenin on the exoprotein pattern of this strain and of its agr-null derivative. As can be seen, subinhibitory concentrations of cerulenin had a profound effect on the exoprotein patterns of both strains. We also introduced an agr tester plasmid containing an agrp3-lux fusion into strain LS-1 to confirm inhibition of agr by subinhibitory concentrations of cerulenin in this strain. Although a subinhibitory concentration of cerulenin had no effect on the growth rate of LS-1 (not shown), it totally blocked agr activation (Fig. 7b
).
For the in vivo test, we used an intravenous inoculum of 1·5x108 c.f.u., which was lethal for three of three mice within 24 h. Injection of cerulenin, at 5, 10 or 40 mg kg1, along with the bacteria, resulted in survival of all of three mice. Note that the lowest dose corresponds to the standard subinhibitory dose in vitro, assuming that cerulenin is distributed uniformly within the mouse. Since we have no information on the fate of cerulenin in the mouse, we do not know whether the lower doses of cerulenin, co-injected with the inoculum, were growth inhibitory for the bacteria. The highest dose, however, would clearly correspond to an inhibitory dose in vitro. Note that a mouse given this dose but no bacteria showed no grossly visible ill effects after 18 h. Further study on this point is in progress.
 |
DISCUSSION
|
---|
In this study, we have revisited an earlier view that the antibiotic cerulenin, at subinhibitory concentrations, inhibits the secretion of bacterial toxins and other exoproteins, with the view of using this inhibition to test staphylococcal exoproteins for autorepression. It rapidly became apparent, however, that subinhibitory cerulenin inhibits the synthesis of staphylococcal exoproteins and would be of no use in identifying autorepressors. It acts at the level of exoprotein gene transcription, though it has little, if any, effect on the production of cytoplasmic proteins. By exchanging promoters between entB, whose transcription was blocked by cerulenin, and entA, whose transcription was not blocked, we showed that EntB as well as EntA is secreted in the presence of subinhibitory concentrations of cerulenin. The level of extracellular EntB was not as high as in the absence of the drug; this difference was not observed in the agr-null background (data not shown). We conclude, nevertheless, that interference with secretion is not the primary mode of action of subinhibitory concentrations of cerulenin on bacterial exoproteins. To extend this finding, we evaluated the effects of subinhibitory concentrations of cerulenin on transcription of several other exoprotein genes. Transcription was strongly inhibited for hla, geh, map and sspL, while sspA and spa were affected only minimally. On the basis of these results, we suggest that the profound effect of subinhibitory concentrations of cerulenin on staphylococcal exoprotein patterns is largely or entirely the result of blockage of transcription.
Since cerulenin acts by inhibiting fatty acid synthesis and related processes, there is no obvious means for it to block transcription directly, suggesting that it might act through regulatory genes whose activation is dependent on pre-existing functions rather than on primary transcription. We addressed this by testing for the effects of subinhibitory concentrations of cerulenin on certain regulatory genes. We observed that subinhibitory concentrations of cerulenin profoundly inhibit the transcriptional activation of two of these, agr and sae, both of which are activated by transmembrane signalling, but had a moderate stimulatory effect on two others, sarA and rot, which do not involve transmembrane signalling. Since cerulenin inhibition of agr activation was partially reversed by heptadecanoic acid, we suggest that the antibiotic may inhibit agr signalling by modifying the composition of the membrane. Since agr signalling involves two pre-transcriptional steps, maturation of the AIP (Lina et al., 1998
) and activation of the receptor (Lina et al., 1998
; Novick et al., 1995
), both of which involve transmembrane proteins, either of these processes could be the target of cerulenin inhibition. Preliminary experiments, however, suggest that neither of these processes is sensitive to cerulenin (unpublished data). Therefore, some upstream pre-transcriptional process may be the target.
It is clear from the various exoprotein profiles presented that agr is responsible for only a portion of the subinhibitory concentrations of cerulenin effects on exoprotein synthesis. It is likely that one or more additional regulatory genes are involved; however, there is presently only a slight hint that activation of any regulatory determinant could involve a cerulenin-sensitive process. This is that the results with two strains, MRSA252 and the arlR mutant, show only minimal effects of subinhibitory concentrations of cerulenin on exoprotein profiles. Thus activation of the arlRS signalling pathway, or of an unknown regulatory element that is defective in MRSA, may be directly affected by subinhibitory concentrations of cerulenin. Since arlRS actually downregulates most of the exoprotein genes in its regulon, the arlR mutant generates a characteristic exoprotein profile that is easily seen to be resistant to the effects of subinhibitory concentrations of cerulenin. The mechanism of activation of arlRS, however, is not presently known. Incidentally, is it not remarkable that cerulenin blocks the formation of staphylococcal peptide autoinducers and of the homoserine lactone autoinducers used by Gram-negative bacteria, and does so by apparently different mechanisms? Finally, subinhibitory concentrations of cerulenin appear to have a dramatic effect on lethal staphylococcal sepsis in the mouse. Because these results raise the possibility that cerulenin could serve as an anti-infective, despite its interference with certain eukaryotic functions, it is planned to investigate this infection model in more detail.
 |
ACKNOWLEDGEMENTS
|
---|
We are grateful to Ruzhong Jin for performing the animal experiments and to Jesse S. Wright, III, for construction of the agrp3-lux strain. This work was supported by NIH grants R01-AI14372 and R01-AI42783 to R. P. N.
 |
REFERENCES
|
---|
Altenbern, R. A. (1977). Extreme sensitivity of staphylococcal enterotoxin B and C production to inhibition by cerulenin. Antimicrob Agents Chemother 11, 906908.[Medline]
Bremell, T., Lange, S., Yacoub, A., Ryden, C. & Tarkowski, A. (1991). Experimental Staphylococcus aureus arthritis in mice. Infect Immun 59, 26152623.[Medline]
Byers, D. M. & Meighen, E. A. (1989). Inhibition of Vibrio harveyi bioluminescence by cerulenin: in vivo evidence for covalent modification of the reductase enzyme involved in aldehyde synthesis. J Bacteriol 171, 38663871.[Medline]
Charpentier, E., Anton, A. I., Barry, P., Alfonso, B., Fang, Y. & Novick, R. P. (2004). Novel cassette-based shuttle vector system for gram-positive bacteria. Appl Environ Microbiol 70, 60766085.[Abstract/Free Full Text]
Fournier, B., Klier, A. & Rapoport, G. (2001). The two-component system ArlS-ArlR is a regulator of virulence gene expression in Staphylococcus aureus. Mol Microbiol 41, 247261.[CrossRef][Medline]
Giraudo, A. T., Cheung, A. L. & Nagel, R. (1997). The sae locus of Staphylococcus aureus controls exoprotein synthesis at the transcriptional level. Arch Microbiol 168, 5358.[CrossRef][Medline]
Goldberg, I., Walker, J. R. & Bloch, K. (1973). Inhibition of lipid synthesis in Escherichia coli cells by the antibiotic cerulenin. Antimicrob Agents Chemother 3, 549554.[Medline]
Greenspan, M. D. & Mackow, R. C. (1977). The effect of cerulenin on sterol biosynthesis in Saccharomyces cerevisiae. Lipids 12, 729740.[Medline]
Hiltunen, M. & Soderhall, K. (1992). Inhibition of polyketide synthesis in Alternaria alternata by the fatty acid synthesis inhibitor cerulenin. Appl Environ Microbiol 58, 10431045.[Abstract]
Jacques, N. A. (1983). Membrane perturbation by cerulenin modulates glucosyltransferase secretion and acetate uptake by Streptococcus salivarius. J Gen Microbiol 129, 32933302.[Medline]
Ji, G., Beavis, R. & Novick, R. P. (1997). Bacterial interference caused by autoinducing peptide variants. Science 276, 20272030.[Abstract/Free Full Text]
Jochen, A. L., Hays, J. & Mick, G. (1995). Inhibitory effects of cerulenin on protein palmitoylation and insulin internalization in rat adipocytes. Biochim Biophys Acta 1259, 6572.[Medline]
Kreiswirth, B. N., Lofdahl, S., Betley, M. J., O'Reilly, M., Schlievert, P. M., Bergdoll, M. S. & Novick, R. P. (1983). The toxic shock syndrome exotoxin structural gene is not detectably transmitted by a prophage. Nature 305, 709712.[CrossRef][Medline]
Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680685.[Medline]
Lina, G., Jarraud, S., Ji, G., Greenland, T., Pedraza, A., Etienne, J., Novick, R. P. & Vandenesch, F. (1998). Transmembrane topology and histidine protein kinase activity of AgrC, the agr signal receptor in Staphylococcus aureus. Mol Microbiol 28, 655662.[CrossRef][Medline]
Mantsala, P. (1982). Inhibition of protein secretion by cerulenin in Bacillus subtilis. J Gen Microbiol 128, 29672972.[Medline]
Mantsala, P. & Lehtinen, H. (1982). Secretion of beta-lactamase by Escherichia coli in vivo and in vitro: effect of cerulenin. Antonie Van Leeuwenhoek 48, 353364.[CrossRef][Medline]
Martinez, J. P., Elorza, M. V., Gozalbo, D. & Sentandreu, R. (1982). Regulation of alpha-galactosidase synthesis in Saccharomyces cerevisiae and effect of cerulenin on the secretion of this enzyme. Biochim Biophys Acta 716, 158168.[Medline]
Novick, R. P. (1991). Genetic systems in staphylococci. Methods Enzymol 204, 587636.[Medline]
Novick, R. P. & Jiang, D. (2003). The staphylococcal saeRS system coordinates environmental signals with agr quorum sensing. Microbiology 149, 27092717.[CrossRef][Medline]
Novick, R. P., Ross, H. F., Projan, S. J., Kornblum, J., Kreiswirth, B. & Moghazeh, S. (1993). Synthesis of staphylococcal virulence factors is controlled by a regulatory RNA molecule. EMBO J 12, 39673975.[Abstract]
Novick, R. P., Projan, S. J., Kornblum, J., Ross, H. F., Ji, G., Kreiswirth, B., Vandenesch, F. & Moghazeh, S. (1995). The agr P2 operon: an autocatalytic sensory transduction system in Staphylococcus aureus. Mol Gen Genet 248, 446458.[Medline]
Price, A. C., Choi, K. H., Heath, R. J., Li, Z., White, S. W. & Rock, C. O. (2001). Inhibition of beta-ketoacyl-acyl carrier protein synthases by thiolactomycin and cerulenin. Structure and mechanism. J Biol Chem 276, 65516559.[Abstract/Free Full Text]
Saleh, F. A. & Freer, J. H. (1984). Inhibition of secretion of staphylococcal alpha toxin by cerulenin. J Med Microbiol 18, 205216.[Abstract]
Somerville, G. A., Chaussee, M. S., Morgan, C. I., Fitzgerald, J. R., Dorward, D. W., Reitzer, L. J. & Musser, J. M. (2002). Staphylococcus aureus aconitase inactivation unexpectedly inhibits post-exponential-phase growth and enhances stationary-phase survival. Infect Immun 70, 63736382.[Abstract/Free Full Text]
Straub, S. G., Yajima, H., Komatsu, M., Aizawa, T. & Sharp, G. W. (2002). The effects of cerulenin, an inhibitor of protein acylation, on the two phases of glucose-stimulated insulin secretion. Diabetes 51, S91S95.[Abstract/Free Full Text]
Tegmark, K., Karlsson, A. & Arvidson, S. (2000). Identification and characterization of SarH1, a new global regulator of virulence gene expression in Staphylococcus aureus. Mol Microbiol 37, 398409.[CrossRef][Medline]
Tremaine, M. T., Brockman, D. K. & Betley, M. J. (1993). Staphylococcal enterotoxin A gene (sea) expression is not affected by the accessory gene regulator (agr). Infect Immun 61, 356359.[Abstract]
Val, D. L. & Cronan, J. E., Jr (1998). In vivo evidence that S-adenosylmethionine and fatty acid synthesis intermediates are the substrates for the LuxI family of autoinducer synthases. J Bacteriol 180, 26442651.[Abstract/Free Full Text]
Vojtov, N., Ross, H. F. & Novick, R. P. (2002). Global repression of exotoxin synthesis by staphylococcal superantigens. Proc Natl Acad Sci U S A 99, 1010210107.[Abstract/Free Full Text]
Weinrick, B., Dunman, P. M., McAleese, F., Murphy, E., Projan, S. J., Fang, Y. & Novick, R. P. (2004). Effect of mild acid on gene expression in Staphylococcus aureus. J Bacteriol 186, 84078423.[Abstract/Free Full Text]
Wright, J. S., 3rd, Jin, R. & Novick, R. P. (2005). Transient interference with staphylococcal quorum sensing blocks abscess formation. Proc Natl Acad Sci U S A 102, 16911696.[Abstract/Free Full Text]
Received 7 April 2005;
revised 4 June 2005;
accepted 13 June 2005.
Copyright © 2005 Society for General Microbiology.