New Architectures for Tet-On and Tet-Off Regulation in Staphylococcus aureus

Abstract

Inducible expression is a valuable approach for the elucidation of gene functions. Here, we present new configurations of the tetracycline-dependent gene regulation (tet) system for Staphylococcus aureus. To provide improved and expanded modes of control, strains and plasmids were constructed for the constitutive expression of tetR or a variant allele, rev-tetR^r2. The encoded regulators respond differently to the effector anhydrotetracycline (ATc), which causes target gene expression to be induced with TetR or repressed with rev-TetR. To quantify and compare regulation mediated by episomal or chromosomal (rev-)tetR constructs, expression from a chromosomal P_xyl/tet-gfpmut2 fusion was measured. Chromosomally encoded TetR showed tight repression and allowed high levels of dose-dependent gene expression in response to ATc. Regulatory abilities were further verified using a strain in which a native S. aureus gene (zwf) was put under tet control in its native chromosomal location. Tight repression was reflected by transcript amounts, which were barely detectable under repressed conditions and high in ATc-treated cells. In reporter gene assays, this type of control, termed Tet-on, was more efficient than Tet-off regulation, in which addition of ATc causes downregulation of a target gene. The latter was achieved and quantified by direct rev-TetR control of P_xyl/tet-gfpmut2. Additionally, TetR was used in trans to control the expression of antisense RNA for posttranscriptional gene silencing. Induction of antisense RNA expression of the fabI gene caused pronounced growth retardation lasting several hours. These results demonstrate the efficiency of the new tet systems and their flexible use for different purposes.

Staphylococcus aureus is a frequent colonizer of human skin and mucosal surfaces but is also an opportunistic pathogen. Superficial infections, such as impetigo, carbuncles, or scaled-skin syndrome, are attributed to S. aureus; furthermore, these bacteria are capable of infecting internal organs of the human body to, e.g., elicit osteomyelitis, endocarditis, or sepsis. The quest for the elucidation of gene-function relationships in S. aureus is facilitated by the availability of various molecular biology techniques (47); in particular, obtaining deletion mutants is aided by a number of protocols (2, 4, 10). Naturally, in the case of essential genes, the respective deletion mutants cannot be cultured. Those genes, however, represent promising targets for anti-infective compounds needed to combat strains that withstand treatment with current antibiotics (41).

To circumvent deletion, recalcitrant genes can be put under inducible expression control within the chromosome (16, 31, 49). The most prominent gene regulation systems for S. aureus employ the following promoters and regulators: P_xyl and XylR (53), Pspac and LacI (31), and P_xyl/tet and TetR (32). All three systems are characterized by induction of target gene expression upon detachment of the effector-bound repressors from DNA but differ in terms of efficiency and regulatory windows (54). It was shown that a one-plasmid-based TetR-dependent (tet) regulation system displayed considerable leakiness in the repressed state. However, this architecture, in which tetR is expressed under the control of the autoregulated promoter PR* while the adjacent divergently oriented promoter, P_xyl/tet, mediates conditional expression of a target gene, is the most prominent for tet regulation in S. aureus to date. Both PR* and the Bacillus subtilis P_xylA-derived P_xyl/tet promoter (20) are vested with tetO sequences as cognate sites for TetR, whose DNA binding capacity is abrogated by inducers like tetracycline (Tc) or anhydro-Tc (ATc) (14). Expression of a target gene in canonical prokaryotic tet control is thus achieved upon TetR induction by (A)Tc.

If the gene of interest proves to be essential for growth or during the course of infection, a strain then requires the permanent presence of an inducer to ensure survival or to sustain virulence, respectively. It would, however, be beneficial to have such genes active by default and to be able to shut them down rapidly at a given time point. In TetR-regulated antisense RNA expression, short fragments transcribed under P_xyl/tet control from the nontemplate strand of a target gene lead to its downregulation in a posttranscriptional fashion (32). With the emergence of mutant TetR variants, this outcome can also be achieved by direct tet regulation. Among the transcriptional repressors used for transgene regulation to date, TetR is unique in displaying a reversed phenotype when mutated at selected positions (51). Such reverse TetR regulators (rev-TetR) require ATc as a corepressor to bind tetO and thus to silence a downstream target gene. rev-TetR has been applied in Gram-positive and Gram-negative bacteria (22, 31, 48), frequently employing the rev-TetR variant r2 (rev-TetR^r2), characterized by the amino acid exchanges E15A, L17G, and L25V (51).

Although tet regulation systems in higher eukaryotes employ other cis elements, modified TetR-based regulators, and distinct effectors (7), it seems timely to adapt established nomenclature of eukaryotic tet regulation to bacterial systems, as previously suggested (19). Thus, Tet-on control prevails in a configuration in which the addition of an effector (in this case, ATc) leads to gene expression in control, whereas Tet-off defines the silencing of a target gene under these conditions. New setups are presented in this study to enable both types of control and to provide improved regulation capacities.

MATERIALS AND METHODS

Chemicals, enzymes, molecular weight markers, and oligonucleotides.

Chemicals were generally purchased from Merck (Darmstadt, Germany), AppliChem (Darmstadt, Germany), or Sigma (Munich, Germany) at the highest purity available. ATc was purchased from Acros (Geel, Belgium). DNA enzymes and molecular weight markers were obtained from Fermentas (St. Leon-Rot, Germany), New England Biolabs (Frankfurt/Main, Germany), Roche Diagnostics (Mannheim, Germany), Peqlab (Erlangen, Germany), or GE Healthcare (Munich, Germany). Lysostaphin was obtained from Dr. Petry Genmedics (Reutlingen, Germany). Oligonucleotides were purchased from biomers.net (Ulm, Germany) and are given in Table 1.

TABLE 1.

Oligonucleotides used in this study

Primer	Sequence (5′→3′)
DP10gh	CTATTTGCAACAGTGCCGT
fabI_fw	TGAAAGGTACCCATATGTCATCATGGGAATCG
fabI_rev	TTTTAGGTCGACGAGCCACAATTGTTAATGAG
G6PDH-NorthT7_rev	CTAATACGACTCACTATAGGGAGACATGTGGTTTTGCACCATATC
G6PDH-North_for	GACGCGTTTATGGAACATG
gfp_fdh_fw	TTATCTTGATCATAAGGGTAACTATT
gfp_fdh_rev	TGACACCTGATCAACTGGTAATG
Karcfw_pcx19	ATTATAGATCTGAATTCAAAGATAAAAGGAGG
Komp_arc_rw	ATATTGAGCTCATCACCTTAAATTTTACTG
lip_fw	TTAATAGGTACCTTACTATT
lip_rev	TCACTTAAGCGTCTGCTAAAT
Pt17_tet_fw	TCAAGAGGATCCACTCCAAATATAGCT
Pt17_tet_rev	ACTGTGAGGATCCGGATGAGTAGGTA
pxt_rev	GGGATCCTCTAGAGTCGACCTG
pxt_fw	TCTGGTAAGCTTGAAGTTACCAC
SA0171KOP1F	AAGTCGAATTCCCACAATCACAAATCATCAC
SA0171KOP2R	AATATGGATCCCCCTTGAATTATTGTTAAATTC
SA0171KOP3F	TATTAGTCGACGCTAGCGATTAACGCTTTC
SA0171KOP4R	AATTAGATATCTGATAACGACTTGCATGCCTC
tet arc native fw	ATTATGCTTAGCAGACTATTTGCAACAGTGC
tet arc native rev	AATTTGGACCATCTGTCATGTCTATTCCTCTAGAGTCGACCTGC
tet-G6PDH_B_for	TATATGTCGACCATATTGTATGACCTACTGAATGG
tet-G6PDH_B_rev	TATATAAGCTTTTCGTAATCAATCCTTGTCATTCG
tet-G6PDH_A_for	ATTATGAGCTCTGTTATTTGGCAAGCATTGTTTATTAC
tet-G6PDH_A_rev	TTATATCCGGACGATTCATGTATCATGTTTAATGTG
tet_pCX_fw	GATCTGTACACTCGAGAGGATCCACTCCAAATATAGCT
tet_pCX_rev	CGAGGCATCACTTGCCGAACGGTAAGGAACCCAGACTGT

Bacterial strains, growth conditions, and manipulations.

Bacteria were grown in liquid or on solid BM (6), BOG (BM without glucose), or tryptic soy broth (TSB) (Sigma, Munich, Germany). Antibiotics were used, where appropriate, at the following final concentrations: ampicillin, 100 μg/ml; kanamycin, 15 μg/ml; chloramphenicol, 10 μg/ml; and spectinomycin, 150 μg/ml. As the effector for (rev-)TetR, ATc (prepared as a 10 mM stock solution in 70% ethanol) was added to bacterial cultures at a final concentration of 0.4 μM unless otherwise stated. Escherichia coli was made competent and transformed using standard techniques (26). Competent B. subtilis cells were obtained as described previously (38), and the integration of plasmid DNA into the amyE locus was verified by the loss of amylase activity. S. aureus cells were subjected to electroporation according to the method of Augustin and Götz (3). Transformation of Staphylococcus carnosus was achieved using protoplast cells (23). Plasmids cloned in E. coli were shuttled through the restriction-deficient S. aureus RN4220 or tetR-expressing derivatives thereof (see below) prior to transformation of S. aureus SA113 or its descendants. Chromosomal integration of plasmid DNA into S. aureus was achieved as described previously (10). lox-flanked aphAIII kanamycin resistance cassettes were excised from modified strains' genomes using plasmid pRAB1 carrying cre, encoding a site-specific recombinase (40). Exchange of DNA between different S. aureus strains was achieved by bacteriophage φ11 transduction (27). The strains and plasmids used in this study are listed in Table 2.

TABLE 2.

Bacterial strains and plasmids used in this study

Strain or plasmid	Relevant characteristic(s)	Reference
B. subtilis
WH557	trpC2 lacA::Pt17-tetR lox66-aphAIII-lox71	9
WH557-gfp	trpC2 lacA::Pt17-tetR lox66-aphAIII-lox71 amyE::Pxyl/tet-gfpmut2	9
WH558	trpC2 lacA::Pt17-tetR lox66-aphAIII-lox71 amyE::Pxyl/tet-lacZ	9
E. coli
DH5α	λ− ϕ80dlacZΔM15 Δ(lacZYA-argF)U169 recA1 endA1 hsdR17(rK− mK−) supE44 thi-1 gyrA relA1	26
XL1-Blue	hsdR17(rK− mK+) recA1 endA1 gyrA96 thi-1 supE44 relA1 lac [F′ proAB lacIq ZΔM15 Tn10 (Tetr)]	Stratagene
S. aureus
SA113 (ATCC 35556)	NCTC8325 derivative; agr negative; 11-bp deletion in rbsU	29
RN4220	NCTC8325-4 derivative; acceptor of foreign DNA	29
RAB133	arcABDCR::lox72	40
RAB171	SA113 fdh::Pxyl/tet-gfpmut2	This work
RAB180	RN4220 lip::Pt17-tetR lox66-aphAIII-lox71	This work
RAB190	RN4220 lip::Pt17-tetR lox72	This work
RAB200	SA113 lip::Pt17-tetR lox66-aphAIII-lox71	This work
RAB210	SA113 lip::Pt17-tetR lox72	This work
RAB211	SA113 lip::Pt17-tetR lox72 fdh::Pxyl/tet-gfpmut2	This work
RAB215	SA113 lip::Pt17-tetR lox66-aphAIII-lox71 Pxyl/tet-zwf	This work
RAB216	SA113 lip::Pt17-tetR lox72 Pxyl/tet-zwf	This work
RAB220	SA113 lip::Pt17-revtetR lox66-aphAIII-lox71	This work
RAB230	SA113 lip::Pt17-revtetR lox72	This work
RAB231	SA113 lip::Pt17-revtetR lox72 fdh::Pxyl/tet-gfpmut2	This work
RAB235	SA113 lip::Pt17-revtetR lox66-aphAIII-lox71 Pxyl/tet-zwf	This work
RAB236	SA113 lip::Pt17-revtetR-r2 lox72 Pxyl/tet-zwf	This work
S. carnosus
TM300	Wild type	50
Plasmids
pBT2	cat bla; E. coli/Staphylococcus shuttle vector; thermosensitive ori for staphylococci	10
pCX19	cat xylR lip; Staphylococcus vector	28
pIC156	Source of specR	52
pRAB1	cat bla PpagA-cre; pBT2 derivative	40
pRAB2(-r2)	cat bla ′lip′-Pt17-(rev-)tetR-lox66-aphAIII-lox72-′lip′; pBT2 derivative	This work
pRAB4	cat bla ′SAOUHSC_00139-141-Pxyl/tet-gfpmut2-specR-SAOUHSC_00143′; pBT2 derivative	This work
pRAB6	cat bla ble Pxyl/tet-AS ′fabI′; pRB473 derivative	This work
pRAB7(-r2)	cat Pt17-(rev-)tetR; pCX19 derivative	This work
pRAB8	cat PR*-tetR Pxyl/tet-arcABDCR; pCX19 derivative	This work
pRAB9	cat bla SAOUHSC_01599-lox66-aphAIII-lox72-Pxyl/tet-zwf; pBT2 derivative	This work
pRB473	cat bla ble; E. coli/Staphylococcus shuttle vector	11
pWH105-gfpmut2	bla ′amyE′-cat-Pxyl/tet-gfpmut2-′amyE′; pWH105 derivative	This work
pWH1925-r2	Source of revtetR	51
pWH1935-1a	Source of tetR	9
pWH1935-2	Source of Pxyl/tet	9

Isolation, manipulation, and detection of nucleic acids and proteins.

Plasmid DNA was prepared from E. coli or staphylococci by using commercially available kit systems from Peqlab (Erlangen, Germany) or Qiagen (Hilden, Germany). Staphylococcus cells were treated with lysostaphin at a final concentration of 12.5 μg/ml for 30 min at 37°C. Chromosomal DNA of S. aureus or S. carnosus was prepared by phenol-chloroform extraction or by using the InstaGene system (Bio-Rad, Munich, Germany). DNA sequencing was carried out at GATC (Constance, Germany). For detection and quantification of tet-controlled zwf transcript amounts (see below), RNA preparation and Northern blots were conducted as described previously (18, 21, 45) using primers G6PDH-North_for and G6PDH-NorthT7_rev. The hybridization signals were detected using Lumi-Film (Roche, Mannheim, Germany). Soluble protein fractions of relevant strains were subjected to (rev-)TetR-directed Western blotting using rabbit polyclonal antibodies diluted 1:20,000 (35) in combination with the ECL Western Blotting Detection System (GE Healthcare, Munich, Germany). Chemiluminescent signals were detected using a Konica Medical Film Processor QX-150U machine (Konica Minolta, Munich, Germany) and Hyperfilm ECL (GE Healthcare, Munich, Germany).

Fluorescence measurements.

Cells from 2 ml of mid-log-phase cultures grown in BOG with or without ATc were harvested, washed, and resuspended in 200 μl phosphate-buffered saline (PBS). Cell densities and fluorescence were determined using 150 μl of each sample in a flat-bottom 96-well microtiter plate (Greiner Bio-One, Nürtingen, Germany) using a GENios microplate reader (Tecan, Crailsheim, Germany). Raw fluorescence data were obtained by using filters for excitation at a λ of 485 nm and emission at a λ of 535 nm. The values obtained were correlated with the respective optical density at 595 nm (OD₅₉₅) and then subtracted from background obtained from cells without gfpmut2 grown with or without ATc.

Construction of a plasmid for conditional complementation of S. aureus ΔarcABDCR.

Unless otherwise stated, plasmids were propagated and cloned in E. coli DH5α or XL1-Blue throughout this study. The final constructs were verified by analytical restriction and sequencing. A plasmid for tet-regulated complementation of an arc operon deletion strain (40) was cloned as follows. The arcABDCR genes were amplified from chromosomal DNA of S. aureus with primers Karcfw_pcx19 and Komp_arc_rw. The 5.7-kbp Ecl136I/BglII-digested product was cloned into the BamHI/SmaI-digested plasmid pCX19 (28). A 0.84-kbp P_xyl/tet-PR*-tetR regulatory region generated from pWH1935-1a with primers tet_arc_native_fw and tet_arc_native_rev was inserted into the obtained plasmid via AhdI/BlpI to yield pRAB8.

Cloning of vectors for episomal expression of tetR or rev-tetR.

Throughout this article, rev-tetR^r2 and rev-TetR^r2 are referred to as rev-tetR and rev-TetR, respectively, for convenience. A 0.86-kbp PCR product consisting of Pt17-rev-tetR was obtained from pRAB2-r2 using the primers tet_pCX_fw and tet_pCX_rev. The undigested fragment was cloned into the SmaI/ScaI-digested staphylococcal vector pCX19 (28). The resulting plasmid, in which rev-tetR had the same orientation as the plasmid's cat gene, was termed pRAB7-r2. To clone pRAB7 for tetR expression, a fragment containing the wild-type (wt) regulator gene was amplified from plasmid pWH1935-1a (9) using primers DP10gh and pxt_rev. A 510-bp XbaI/Eco47III part of tetR was ligated with the likewise cut pRAB7-r2 to replace the codons responsible for the reverse phenotype with those of wt TetR.

Construction of strains for chromosomal expression of tetR or rev-tetR.

To provide homologous regions for chromosomal integration of (rev-)tetR, a major part of S. aureus lip was amplified with primers lip_fw and lip_rev. The obtained sequence was introduced into the allelic replacement vector pBT2 (10) via Acc65I/AflII to yield pBT2-lip. A sequence containing the synthetic promoter Pt17, tetR, and an aphAIII cassette flanked by the lox66 and lox71 sequences (1) was obtained from B. subtilis WH558 (9) by PCR using primers Pt17_tet_fw and Pt17_tet_rev. The BamHI-cut fragment was inserted into pBT2-lip, in which the enzyme cuts once within the lip sequence. Cloning of Pt17-containing sequences was generally performed in S. carnosus, because propagation in E. coli usually resulted in massive deletions within the synthetic promoter region. One candidate in which tetR was colinear to the plasmid backbone cat resistance marker was designated pRAB2. To obtain pRAB2-r2, tetR of pRAB2 was partially replaced with an Eco47III/XbaI fragment of rev-tetR of pWH1925-r2. The Pt17-(rev-)tetR lox66-aphAIII-lox72 sequences were integrated into SA113 lip, yielding strain RAB200 (tetR) or RAB220 (rev-tetR), respectively. In vivo Cre recombinase treatment (40) led to the excision of the aphAIII sequence. The obtained kanamycin-sensitive strains were designated RAB210 (tetR) and RAB230 (rev-tetR).

Construction of strains with chromosomally borne gfpmut2 for tet control.

In order to obtain a chromosomal integration mutant for P_xyl/tet-gfpmut2, the S. aureus fdh gene (SAOUHSC_00142) encoding formate dehydrogenase was chosen. To construct an fdh replacement vector, a 1.1-kbp fragment upstream of fdh was amplified using primers SA0171KOP1F and SA0171KOP2R, which introduced EcoRI and BamHI restriction sites, respectively. A 1.25-kbp spectinomycin resistance cassette was obtained from plasmid pIC156 (52) via BamHI/SalI cleavage. To provide a downstream homologous sequence, a PCR product of 0.99 kbp obtained with the primers SA0171KOP3F and SA0171KOP4R was cut with SalI/EcoRV and cloned into pBT2. The resulting vector was cut with SalI/EcoRI to insert the upstream and marker fragments to obtain pBT2-fdh. pWH105-gfpmut2, a derivative of pWH105 (35) in which lacZ has been replaced by the triple gfp mutant gfpmut2 (12) via HindIII/BlpI, was used as a template to amplify P_xyl/tet-gfpmut2 using the primers gfp_fdh_fw and gfp_fdh_rev. BclI digestion was used to ligate the fragment with the BamHI-cut pBT2-fdh. One candidate in which gfpmut2 had been inserted divergently to cat was designated pRAB4. Insertion of the P_xyl/tet-gfpmut2 portion into SA113, RAB210, and RAB230 gave rise to RAB171 (P_xyl/tet-gfpmut2), RAB211 (tetR P_xyl/tet-gfpmut2), and RAB231 (rev-tetR P_xyl/tet-gfpmut2), respectively.

Construction of strains for chromosomal tet regulation of the S. aureus zwf gene.

Primers tet-G6PDH_B_for and tet-G6PDH_B_rev were used to amplify and modify the zwf gene (SAOUHSC_01599; glucose-6-phosphate 1-dehydrogenase), including its Shine-Dalgarno sequence, from the chromosomal DNA of S. aureus. The SalI/HindIII-digested PCR product (1.2 kbp) was ligated into likewise cut pBT2 (10), creating pBT2B. Primers tet-G6PDH_A_for and tet-G6PDH_A_rev were used to obtain a 1.6-kbp upstream flanking region of zwf. P_xyl/tet with a divergent lox66-aphAIII-lox71 cassette was obtained from the plasmid pWH1935-2 by Kpn2I/SalI digestion and ligated with both the SacI/Kpn2I-restricted PCR fragment and SacI/SalI-digested pBT2B, resulting in pRAB9. Prior to chromosomal integration into SA113-derived strains, pRAB9 was shuttled through the tetR expression strain RAB190, which was constructed like RAB210, albeit in an RN4220 background. Insertion of the P_xyl/tet-zwf portion of pRAB9 into RAB210 gave rise to RAB215. The P_xyl/tet-zwf sequence was transduced into RAB230 with phage φ11, resulting in RAB235. Cre-dependent removal of the aphAIII fragment (40) yielded the kanamycin-sensitive strains RAB216 (tetR P_xyl/tet-zwf) and RAB236 (rev-tetR P_xyl/tet-zwf).

Construction of a plasmid for antisense tet regulation of fabI.

P_xyl/tet was amplified from pWH1935-2 (9) with the primers pxt_fw and pxt_rev. The XbaI/HindIII-digested 131-bp fragment was cloned into the shuttle vector pRB473 (11). The resulting intermediate construct was cut with Acc65I and SalI, and a similarly restricted 372-bp PCR product containing part of the fabI gene [SAOUHSC_00947; enoyl-(acyl carrier protein) reductase] generated with primers fabI_fw and fabI_rev from chromosomal S. aureus DNA was ligated to yield plasmid pRAB6. After propagation in the kanamycin-resistant RAB190 antecessor RAB180, pRAB6 was used to transform RAB210.

RESULTS

Construction and analysis of new tetR and rev-tetR expression plasmids and strains.

In order to conditionally complement an S. aureus arc operon deletion strain (termed RAB133 [40]), plasmid pRAB8 was constructed. It carries tetR under the control of the autoregulated promoter PR* and a divergent P_xyl/tet promoter controlling transcription initiation of the arcABDCR genes. The encoded arginine deiminase pathway enzymes lead to degradation of arginine with concomitant release of ammonia (46). Initial experiments conducted with S. aureus RAB133(pRAB8) under anaerobic conditions showed an increase in pH in the absence of ATc, indicative of ammonia production (data not shown). This gave reason to assume that repression of the plasmid-borne arc operon in the absence of inducer was not extremely pronounced.

This, among other considerations, prompted us to generate new tet regulation configurations that (i) show increased efficiency and (ii) exploit different regulatory architectures. The basic concept included constitutive expression of tetR and its spatial uncoupling of the target gene. To provide a new plasmid for tetR expression, pRAB7, which harbors tetR downstream of the synthetic σ^A-dependent promoter Pt17, was constructed (9). Due to its origin of replication, this plasmid is probably present at about 15 copies per Staphylococcus cell (30). In combination with a target gene cloned downstream of a tet-regulatable promoter, this setup was expected to obey Tet-on logic. In order to exploit rev-TetR as a Tet-off regulator in S. aureus, a derivative of pRAB7 containing rev-tetR^r2, was cloned to express TetR E15A L17G L25V with reverse behavior.

As an alternative configuration, we sought to have the repressors also expressed from the S. aureus chromosome. The dispensable lip gene seemed to be a suitable locus of (rev-)tetR integration, so replacement vectors were used to construct strains RAB210 and RAB230 for Pt17-dependent expression of tetR or rev-tetR, respectively, from the chromosome. Initial Western blot analyses showed that RAB210 TetR levels were comparable to those of Bacillus subtilis WH557, which also harbors a chromosomal Pt17-tetR fusion (9). TetR was more abundant than rev-TetR (not shown). To provide a quantifiable readout system for the new regulation systems, a reporter gene under tet control was integrated into a different locus of the S. aureus chromosome. In previous studies, it had been shown that an S. aureus formate dehydrogenase (fdh)-deficient strain had an unaffected phenotype concerning growth and pH under aerobic conditions (our unpublished results). Hence, a fragment containing a P_xyl/tet-gfpmut2 sequence was integrated into the fdh genes of SA113, RAB210, and RAB230 to obtain RAB171 (no tetR), RAB211 (tetR and gfpmut2 control), and RAB231 (rev-tetR and gfpmut2 control). For reasons of comparison, B. subtilis WH557-gfp was constructed, which, next to Pt17-tetR, also harbors P_xyl/tet-gfpmut2 in a distinct chromosomal site. Finally, S. aureus RAB171 was transformed with pRAB7 and pRAB7-r2 for episomal (rev-)tetR expression to complete the set of reporter strains. Upon quantification of ATc-dependent fluorescence intensity, clear (rev-)TetR-dependent regulation of the reporter gene was observed. In the case of episomal-regulator expression, the efficiencies of Tet-on [RAB171(pRAB7)] and Tet-off [RAB171(pRAB7-r2)] control were comparable. For TetR, ∼127 relative fluorescence units in the repressed state and ∼710 when cells were grown with ATc were determined, resulting in an induction factor (IF) of ∼5.6. In the case of rev-TetR, values averaged around ∼903 and ∼105 in the derepressed state and for corepression with ATc, respectively (IF, ∼8.6). Measurements of strains with chromosomally located (rev-)tetR revealed a particularly efficient regulation capacity for the Tet-on system (RAB211). Here, relative fluorescence of approximately 2,016 determined with ATc was accompanied by an extremely low level of about 23 in the repressed state, resulting in an IF of ∼88. RevTetR-mediated Tet-off control quantified with RAB231 resulted in values of ∼1,098 (without ATc) and ∼345 (with ATc), resulting in an IF of ∼3.2 (Fig. 1). B. subtilis WH557-gfp gave values of ∼79 in the repressed state and ∼1,115 when induced by ATc (IF, ∼14) (data not shown). In order to check for dose dependence of the response, the fluorescences of RAB211 and RAB231 cells grown with different ATc concentrations (5 nM, 0.1 μM, and 0.4 μM) were quantified. ATc dose dependence appeared to be pronounced with RAB211, where the addition of 0.1 μM ATc elicited about half the response of cells grown with 0.4 μM inducer. The response was less concentration dependent with rev-TetR in RAB231, for which corepression with 5 nM ATc could hardly be enhanced by concentrations of up to 0.4 μM ATc (Fig. 2). To correlate the regulation capacities with cytosolic-regulator amounts, Western blot analyses of RAB171(pRAB7), RAB171(pRAB7-r2), RAB211, and RAB231 were conducted, which confirmed that TetR was abundantly expressed, irrespective of the episomal or chromosomal localization of the gene. Signal intensities were comparable in both configurations, with somewhat smaller amounts of rev-TetR (Fig. 3).

FIG. 1.

Regulation capacities of strains RAB171 (with either pRAB7 or pRAB7-r2) for episomal expression and RAB211 and RAB231 for chromosomal expression of (rev-)tetR as determined by a chromosomal P_xyl/tet-gfpmut2 fusion. The white bars represent relative fluorescence values obtained without ATc, and the black bars show the results of cultures grown with 0.4 μM ATc. The error bars represent the standard deviations of triplicate determinations.

FIG. 2.

ATc dose dependence of target gene response by TetR and rev-TetR. The values obtained for RAB211 (wt TetR) and RAB231 (rev-TetR) are shown. The error bars represent the standard deviations of triplicate determinations.

FIG. 3.

Western blot of 35 μg of soluble protein obtained from strains RAB171(pRAB7-r2) with an episomal location or RAB211 and RAB231 with a chromosomal location of (rev-)tetR. A signal obtained with approximately 50 ng of purified TetR is shown as a positive control for comparison. Protein extracts of RAB171 were used as a negative control.

Chromosomal Tet-on and Tet-off regulation of a native S. aureus gene.

Ribose and 2-deoxyribose sugars are major constituents of nucleic acids. When unavailable, ribose components are supplied by decarboxylation of hexoses in the pentose phosphate pathway. We aimed to regulate the glucose-6-phosphate 1-dehydrogenase-encoding gene zwf in its native chromosomal locus. To this end, the (rev-)TetR addressable P_xyl/tet promoter was positioned in the chromosomes of RAB210 and RAB230, upstream of zwf's Shine-Dalgarno sequence. zwf directed Northern blots were conducted with RAB216 (tetR and zwf control), RAB236 (rev-tetR and zwf control), RAB210, and RAB230. In the last two, only weak zwf expression was observed. In the case of Tet-on control (RAB216), strong signals were observed in the induced state, whereas repression appeared to be very efficient, as a specific product was virtually undetectable. In RAB236 without ATc, zwf mRNA was about as abundant as that in induced RAB216 cells, whereas corepression had only moderate effects (Fig. 4). The appearance of two specific zwf signals speaks in favor of two distinct terminators. This assumption, as well as further analyses of zwf gene dosage effects, will be addressed in further studies.

FIG. 4.

Northern blots directed against zwf. The strains analyzed contained the unaffected gene (RAB210 or RAB230) or P_xyl/tet-zwf fusions controlled by chromosome-borne TetR (RAB216) or rev-TetR (RAB236). The lanes display transcript amounts obtained from cultures grown with ATc (+) or without the compound (−). Analyses were conducted with total RNA from log-phase (A) or early-stationary-phase (B) cultures. (C) Representative loading controls with 23S and 16S rRNA.

Episomal antisense control by trans-encoded TetR.

Several previous studies employed tet regulation in S. aureus to express antisense (AS) RNA for posttranscriptional gene silencing. To test the efficiency of RAB210 as a host for the control of episomal AS-RNA expression, fabI, which encodes enoyl-(acyl carrier protein) reductase, involved in fatty acid biosynthesis, was chosen as a target. A part of fabI was attached in inverse orientation to P_xyl/tet in plasmid pRAB6. The impact of AS-fabI expression on the fitness of RAB210 was analyzed in liquid media. Although growth of the ATc-treated RAB210(pRAB6) culture eventually accelerated, pronounced retardation was observed during the first ∼9 h postinoculation. Control cultures of RAB210 harboring the empty vector pRB473 treated equally did not show any growth difference, ruling out cytotoxicity of ATc (Fig. 5).

FIG. 5.

Effects of episomal ′fabI′ AS expression controlled by chromosomally encoded TetR. The growth curves of RAB210(pRAB6) and the growth phenotype of RAB210(pRB473) (no AS expression) (control) are shown. The filled symbols indicate the presence of 0.4 μM ATc.

DISCUSSION

tet regulation has been employed in more than a dozen different Gram-positive and Gram-negative bacterial species to date (8) and is particularly popular for studying genes in S. aureus. This study provides second-generation tet systems for enhanced repression, multiple modes of application, and rev-TetR control (Fig. 6). S. aureus tet regulation plasmids, such as pYJ335 (32), pYH4 (33), or pALC2073 (5), exploit the promoter P_xyl/tet with a single tetO sequence between the −35 and −10 sites (20). Although a comparable one-plasmid setup has shown considerable leakiness (54) and P_xyl/tet equipped with an additional tetO site conferred tighter repression in B. subtilis (20, 35), this second version of P_xyl/tet has notably been neglected for use in S. aureus to date. Our observations with pRAB8, however, suggested that this two-tetO version also produced discernible transcript amounts when repressed. Hence, the target gene promoter may not be the most critical parameter for the efficiency of tet control.

FIG. 6.

Overview of different configurations applied for S. aureus tet regulation in this study. The bent arrows denote promoters involved in tet regulation, and the rectangles represent tetO sites. lox72 sequences are shown as triangles. The arrows representing (rev-)tetR are black, and genes subject to tet control are gray. The dotted lines indicate the genetic sources of repressors that are directed to control of the respective target promoters. (A) A one-plasmid setup was used to control expression of arcABDCR. Here, tetR regulates its own transcription by not only promoter PR*, but also P_xyl/tet in cis. (B) pRAB7 and pRAB7-r2 were applied to episomal (rev-)TetR expression to control a chromosomal P_xyl/tet-gfpmut2 fusion. The latter is represented within the adjacent genes SAOUHSC_00141 and SAOUHSC_00143. Chromsomally embedded (rev-)tetR configurations were applied to control a reporter gene (C), a gene in its native chromosomal locus (D), or an episomally localized antisense fragment (E). In all cases, a double-tetO version of P_xyl/tet was addressed by the repressors, which were expressed by Pt17 (B to E). Within itself, each representation is drawn to scale.

In E. coli, B. subtilis, or mycobacteria (15, 35, 44), constitutive tetR expression provided tight repression, presumably by increasing the cellular TetR-to-tetO ratio. In a very recent study, Corrigan and Foster improved the regulation capacities of a one-plasmid configuration for S. aureus (13). Adaptation of the −10 sequence of the tetR-driving promoter PR* to fit the consensus was unintentionally accompanied by inactivation of the overlapping tetO site in plasmid pRMC1. We reckon that the latter effect mainly accounted for improved repression by providing increased amounts of TetR. In fact, the resulting promoter largely resembles Pt17, which provided valuable services in B. subtilis (9, 35) and also expressed tetR well in S. aureus (Fig. 3). At the moment, the reason why chromosomal expression of TetR provided more pronounced regulation than the pRAB7 system despite comparable cellular protein amounts remains elusive. Since previous studies had shown that the lipase activity of an in vitro-grown SA113 Δlip strain was unaffected (our unpublished results), lip was chosen as a locus for tetR integration. In contrast to previous studies (16, 24), which inserted tetR into the intragenic phage L54a attachment site of geh, encoding an active lipase (39), the lipase activity of RAB210 should be unaffected, which may have implications for future virulence studies (42).

Regarding its induction factor, RAB211 does not at first glance meet the superb regulation capacities of tet regulation reporter strains, such as E. coli DH5αZ1/P_LtetO-1 (IF, ∼5,050) or B. subtilis WH558 (IF, ∼314) (9, 44). The impacts of different reporter genes on the determination of this parameter should not be underestimated, as exemplified by correlating the IF of B. subtilis WH557-gfp in retrospect with that of WH558. The two strains share common tet architectures, but in contrast to WH558 (P_xyl/tet-lacZ), an IF of only ∼14 was determined with the P_xyl/tet-gfpmut2 sequence. The calculated IF of ∼88, measured with a P_xyl/tet-gfpmut2 fusion, thus underscores RAB211's salient regulation efficiency.

Although not found to be essential for S. aureus in two different studies (17, 34), initial deletion approaches for the zwf gene failed in our laboratory (unpublished results). Hence, zwf was placed under tet control in its native chromosomal locus for analysis of transcript amounts, as well as for further future examinations. Northern blots of P_xyl/tet-zwf strains confirmed the particularly tight repression capacities of the chromosomal tetR expression architecture. As only extremely small transcript amounts were detected in the repressed state (Fig. 4), RAB210-mediated target gene repression by chromosomally located tetR may be capable of mimicking knockout phenotypes. On the other hand, in accordance with data of Zhang and colleagues (54), P_xyl/tet appeared to produce very large amounts of target transcript when induced. Besides adjusting the physiologic conditions for target gene expression by applying lower ATc concentrations (Fig. 2), it might also be practicable to design and apply alleviated versions of P_xyl/tet for future applications.

Using a fabI fragment that was largely congruent with one described previously (33), we could demonstrate that the AS-RNA-dependent Tet-off mode (32-34) was also functional when TetR was supplied in trans by RAB210 (Fig. 5). Possible explanations for the mitigation of ATc-dependent growth include effector decomposition, tetR mutations rendering the repressor induction defective, or promoter down-mutations in P_xyl/tet. By employing rev-TetR, this study marks the establishment of a second mode of Tet-off regulation in S. aureus. Evidently, rev-TetR allows the control of the flow of genetic information at a point further upstream than AS-RNA. The repression capacities exerted by rev-TetR were considerably less pronounced than chromosomal Tet-on control (Fig. 1). It is tempting to speculate that larger amounts of rev-TetR may enhance corepression capacities. Promising approaches include codon adaptation (37), the use of different promoters driving rev-tetR (25, 35), and the adaptation of single-chain TetR (36) to a reverse variant. For RAB231 (rev-TetR-controlled gfpmut2), only minor changes in reporter activity were determined between 5 nM and 0.4 μM ATc (Fig. 2). Previous studies of E. coli and Mycobacterium smegmatis (25, 48) had demonstrated graded responses of rev-TetR control to be most distinctive between 0 and ∼6 nM ATc. It can be assumed that dosable target gene expression using rev-TetR in S. aureus may be elicited within a similar range of ATc concentrations.

Figure 6 summarizes various versatile architectures for tet regulation in S. aureus. For the purpose of tet-dependent one-plasmid-based complementation of deletion strains (Fig. 6A), the recently described pRMC1 and -2 vectors seem to be well suited due to their exquisite regulation capacities (13). In contrast, the architecture of RAB216 may be exemplary for the regulation and analysis of S. aureus genes in their native chromosomal locus (Fig. 6D). Regulated complementation of target genes is also feasible by their ectopic integration, e.g., into an intergenic chromosomal phage attachment site (43). Genetic stability, absence of resistance markers, and tight repression capacities make strains configured like RAB216 attractive for analysis in infection models. Finally, RAB210-based episomal AS control (Fig. 6E) can also be exploited for minimally invasive target gene control. Recently, S. carnosus (22) and Staphylococcus epidermidis (H. Rohde, personal communication) have been shown to be amenable to tet regulation. Hence, it can be assumed that the new configurations can also be employed in these and other staphylococci.