Complex transcriptional organization regulates an Escherichia coli locus implicated in lipopolysaccharide biogenesis

Research in Microbiology 162 (2011) 470e482 www.elsevier.com/locate/resmic

Complex transcriptional organization regulates an Escherichia coli locus implicated in lipopolysaccharide biogenesis Alessandra M. Martorana a,1, Paola Sperandeo b, Alessandra Polissi b, Gianni Deho` a,* a

Dipartimento di Scienze Biomolecolari e Biotecnologie, Universita` degli Studi di Milano, 20133 Milan, Italy b Dipartimento di Biotecnologie e Bioscienze, Universita` di Milano-Bicocca, Milan, Italy Received 12 November 2010; accepted 22 February 2011 Available online 12 March 2011

Abstract The Escherichia coli yrbG-lptB locus ( yrbG kdsD kdsC lptC lptA lptB) encodes genes for outer membrane biogenesis, namely, kdsC and kdsD for biosynthesis of the lipopolysaccharide inner core sugar Kdo, and lptA, lptB, and lptC for lipopolysaccharide transport to the outer membrane. Three promoters (yrbGp, kdsCp and the sE-dependent lptAp) have been previously identified by genetic analysis. In this work, we show that transcription of this locus generates an array of overlapping mRNAs and we characterize the two intralocus promoter regions. In the kdsCp region, we identified three promoters (kdsCp1, kdsCp2, and kdsCp3) scattered within about 600 nt in the 30 coding region of kdsD. The lptAp region is composed of two closely spaced promoters, lptAp1 and lptAp2. The former had been previously identified as a sE-dependent promoter. Interestingly, lptAp1 is not activated by several stressful conditions that normally induce the sEdependent envelope stress response, whereas it seems to respond to conditions affecting lipopolysaccharide biogenesis, thus implying a specialized sE-dependent LPS stress signaling pathway. Ó 2011 Institut Pasteur. Published by Elsevier Masson SAS. All rights reserved. Keywords: Escherichia coli; Operon regulation; Lipopolysaccharide biogenesis; Kdo; Lipopolysaccharide stress; sE

1. Introduction Outer membrane (OM) biogenesis in Gram-negative bacteria is a complex task that coordinates biosynthesis and transport of different molecular species (phospholipids, lipopolysaccharides (LPS), outer membrane proteins (OMPs) and lipoproteins) with cell homeostasis, growth, division and differentiation. A master regulator of several genes implicated in this process is the alternative RNA polymerase sigma factor sE. sE plays a key role in the cell envelope stress response (Ades, 2008; Alba and Gross, 2004). This factor is normally maintained in the cell in an inactive form by membrane-

* Corresponding author. Tel.: þ39 02 5031 5019; fax: þ39 02 5031 5044. E-mail addresses: [email protected] (A.M. Martorana), [email protected] (P. Sperandeo), [email protected] (A. Polissi), [email protected] (G. Deho`). 1 Present address: Dipartimento di Biotecnologie e Bioscienze, Universita` di Milano-Bicocca, Milan, Italy.

spanning anti-s factor RseA. Upon accumulation of unfolded proteins in extracytoplasmic space as a consequence of a stress, a protease cascade is activated, leading ultimately to RseA degradation and consequent sE activation. It has been proposed that sE activity is also sensitive to the status of LPS, although the molecular mechanism has not been elucidated (Tam and Missiakas, 2005). Finally, sE may be significantly upregulated in response to intracellular signals related to growth phase and nutrient availability as cells enter stationary phase (Costanzo and Ades, 2006). This activation pathway depends on the alarmone ppGpp (a general signal of starvation stress also required for activation of sS and sN alternative sigma factors upon entry into stationary phase) and appears to be independent of the envelope stress signaling pathway (Costanzo et al., 2008; Costanzo and Ades, 2006). In the last few years, the Escherichia coli yrbG-lptB locus, which contains, in the given order, yrbG-kdsD-kdsC-lptC-lptAlptB, has been identified and characterized (Sperandeo et al., 2009). In addition to yrbG, a non-essential gene encoding

0923-2508/$ - see front matter Ó 2011 Institut Pasteur. Published by Elsevier Masson SAS. All rights reserved. doi:10.1016/j.resmic.2011.03.007

A.M. Martorana et al. / Research in Microbiology 162 (2011) 470e482

a putative cation exchanger, the locus codes for functions implicated in LPS biogenesis, namely two enzymes (KdsC and KdsD) of the biosynthetic pathway of Kdo (an essential sugar moiety of the LPS inner core) and three (LptA, LptB, and LptC) out of seven proteins known to compose the LPS transporter (Lpt), a molecular device that ferries LPS from the inner to the outer membrane across the periplasmic space (Ruiz et al., 2008; Sperandeo et al., 2007, 2008; Wu et al., 2006). A promoter upstream of yrbG (yrbGp), showing similarity with the housekeeping sD-dependent promoter consensus sequence, has been documented by primer extension, whereas two additional promoters (kdsCp and lptAp) were identified by genetic analysis (Sperandeo et al., 2006); moreover, lptAp was shown to be dependent on sE (Sperandeo et al., 2007). In this work, we dissected the transcriptional organization of the locus and found that three major overlapping operons are transcribed from the three promoter regions to the end of the locus and yonder within the downstream rpoN operon. Both promoter regions internal to the locus are composed of multiple promoters. Interestingly, the sE-dependent promoter upstream of lptA does not respond to any type of stress that activates the sE-dependent extracytoplasmic stress pathway except a subset of stressful conditions affecting LPS, thus implying an LPSspecific sE-dependent stress response pathway.

471

30 min. For overexpression of OmpC, the cultures were grown in LD at 37  C up to OD600 ¼ 0.2 and split into two subcultures with and without 0.2% arabinose. The incubation of the non-induced and arabinose-induced subcultures was then continued at 37  C for 30 min. For lptAB and lptC depletion, bacterial cultures at OD600 ¼ 0.2 were harvested by centrifugation, washed twice with LD and diluted 100-fold in LD with or without arabinose. Samples were taken at subsequent time points as indicated. 2.3. b-galactosidase assay Twenty ml of exponential culture at OD600 ¼ 0.4 were collected by centrifugation, resuspended in 500 ml of TEDP (0.1 M TriseHCl, 1 mM EDTA, 1 mM DTT and 1 mM PMSF) and disrupted by sonication (220 s pulses at 40% amplitude). The samples were spinned at 12,000 rpm at 4  C to remove cell debris. The protein concentration in the crude extracts was determined by the Coomassie (Bradford) protein assay kit (THERMO Scientific). b-galactosidase activity of the extracts was measured as described (Miller, 1992). Specific activity was expressed as nmoles of o-nitrophenol-b-galactopyranoside converted to o-nitrophenol min1 mg of protein1. 2.4. Northern blotting and primer extension analyses

2. Materials and methods 2.1. Bacterial strains and plasmids Bacterial strains and plasmids used in this work are described in Table 1 and oligonucleotides in Table 2. pRS415derivative plasmids were constructed by cloning into the EcoRIeBamHI-digested vector PCR fragments amplified from genomic BW25113 DNA with pairs of primers as listed in Table 1 and digested with EcoRIeBamHI. The inserts were controlled by sequencing. LD broth (Ghisotti et al., 1992), LB (Bertani, 1951) and M9 minimal medium (Kunz and Chapman, 1981) have been described. hoLB (hypotonic LB) is 0.2% NaCl LB. 0.2% glucose or 0.2% glycerol, 0.2% L-arabinose (as inducer of the araBp promoter), 0.25% casamino acids, 0.5 mM IPTG (isopropyl-ß-D-thiogalactopyranoside), 100 mg/ml ampicillin (Amp), 50 mg/ml kanamycin (Kan), and 25 mg/ml chloramphenicol (Cam) were added as required. Solid media contained 1% agar. 2.2. Growth conditions and extracytoplasmic stress If not otherwise indicated, bacterial cultures were grown in LD with the required supplements at 37  C. For chemical and osmotic stresses, exponential cultures were split into 20 ml subcultures in prewarmed flasks containing the stress agents and incubation was continued at 37  C for 30 min. For hypotonic stress, the cultures were grown in LB with ampicillin up to an OD600 of 0.2, harvested by centrifugation washed with 0.2% NaCl, resuspended in 1/10 of hoLB, diluted 1:10 in prewarmed LB or hoLB, and reincubated at 37  C for

Total RNA extraction, northern blot analysis and synthesis of radiolabeled riboprobes were previously described (Briani et al., 2007; Deho` et al., 1992). The specific riboprobes were obtained by in vitro transcription with T7 RNA polymerase of DNA fragments obtained by PCR amplification of genomic BW25113 DNA with the primers listed in Table 2. Primer extension analysis was performed as described (Sperandeo et al., 2006). A DNA sequencing ladder generated with the same primer used for retrotranscription using the Taq DNA Polymerase, Sequencing Grade (PROMEGA) was run in parallel with PE samples in the denaturing polyacrylamide gel electrophoresis (PAGE). As sequencing template, either a plasmid harboring the target DNA region or a fragment of BW25113 genomic DNA obtained by PCR with specific primers listed in Table 2 was used. Autoradiographic images were obtained by phosphorimaging using ImageQuant software (Molecular Dynamics). 2.5. Total LPS extraction and analysis LPS was extracted from 2 OD600 unit culture samples by a mini-phenol-water extraction technique, separated by tricine-SDS-PAGE and silver-stained as described (Sperandeo et al., 2007). 3. Results 3.1. Transcriptional profile of the yrbG-lptB locus Attempts to assess the transcriptional profile of the yrbGlptB locus by northern blotting using riboprobes specific for

472

A.M. Martorana et al. / Research in Microbiology 162 (2011) 470e482

Table 1 Bacterial strains and plasmids. Strain or plasmid

Relevant characters/origin

Source or reference (Sperandeo et al., 2006) (Sperandeo et al., 2006) (Sperandeo et al., 2007) (Datsenko and Wanner, 2000) (Hanahan, 1983) (Baba et al., 2006) (Baba et al., 2006) (Carzaniga et al., 2009) This work

KG244 KG245 KG246

BW25113 araBp-lptC BW25113 araBp-lptA BW25113 DrseA D(araD-araB)567 ΔlacZ4787(::rrnB-3) lrph-1 Δ(rhaD-rhaB)568 hsdR514 fhuA2 D(argF-lacZ )U169 phoA glnV44 F80 D(lacZ )M15 gyrA96 recA1 relA1 endA1 thi-1 hsdR17 BW25113 DrpoN::kan BW25113 DrpoS::kan rne-3071 zce-726::Tn10 BW25113 rne-3071 zce-726::Tn10; by P1 transduction from KG241, selection for TetR, screening for thermosensitivity at 44  C BW25113 zce-726::Tn10; as KG243, screening for growth at 44  C BW25113 DrpoN; from JW3169 by kan cassette curing as described (Datsenko and Wanner, 2000) BW25113 DrpoS; from JW5437 by kan cassette curing as described (Datsenko and Wanner, 2000)

Plasmids pAM1 pAM2 pAM3 pAM4 pAM5 pAM6 pAM7 pAM8 pAM9 pAM10 pAM11 pAM12 pAM13 pAM14 pAM15 pAM16 pAM17 pAM18 pAM19 pAM21 pAM22 pAM23 pAM24 pAM25 pAM26 pAM27 pAM28 pAM29 pAM30 pAM31 pAM32 pBA114 pBAD33 pCP20

pRS415 derivative, carries E. coli 3340228 pRS415 derivative, carries E. coli 3340131 pRS415 derivative, carries E. coli 3340051 pRS415 derivative, carries E. coli 3339890 pRS415 derivative, carries E. coli 3339706 pRS415 derivative, carries E. coli 3339099 pRS415 derivative, carries E. coli 3339099 pRS415 derivative, carries E. coli 3341138 pRS415 derivative, carries E. coli 3341295 pRS415 derivative, carries E. coli 3341328 pRS415 derivative, carries E. coli 3341138 pRS415 derivative, carries E. coli 3341260 pRS415 derivative, carries E. coli 3341138 pRS415 derivative, carries E. coli 3339706 pRS415 derivative, carries E. coli 3339575 pRS415 derivative, carries E. coli 3339390 pRS415 derivative, carries E. coli 3339280 pRS415 derivative, carries E. coli 3339890 pRS415 derivative, carries E. coli 3339706 pRS415 derivative, carries E. coli 3339490 pRS415 derivative, carries E. coli 3339731 pRS415 derivative, carries E. coli 3339577 pRS415 derivative, carries E. coli 3339099 pRS415 derivative, carries E. coli 3339706 pRS415 derivative, carries E. coli 3339660 pRS415 derivative, carries E. coli 3339099 pRS415 derivative, carries E. coli 3339831 pRS415 derivative, carries E. coli 3339782 pRS415 derivative, carries E. coli 3339782 pRS415 derivative, carries E. coli 3340073 pRS415 derivative, carries E. coli 3339936 pBAD33 ompC, CamR pACYC184 ori, araC-araBp, CamR lcI857 lP-FLP Rep(Ts) CamR AmpR

Strain BB-3 BB-4 BB-11 BW25113 DH5a JW3169 JW5437 KG241 KG243

pLPT-AB pQE31-S1 pREP4 pRS415

e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e e

3340279 3340279 3340279 3340279 3340279 3339727 3339935 3341434 3341434 3341434 3341325 3341325 3341277 3339935 3339935 3339935 3339935 3340229 3340052 3339802 3339802 3339802 3339802 3339802 3339802 3339681 3339935 3339935 3339850 3340229 3340073

fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment fragment

(primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers (primers

AP101-FG1973) FG1974-FG1973) AP100-FG1973) FG1975-FG1973) FG1889-FG1973) FG2154-FG2155) FG2154-FG2156) AP69-AP70) FG2062-AP70) FG2064-AP70) AP69-FG2061) FG1907-FG2061) AP69-FG2063) FG1889-FG2156) FG2326-FG2156) FG2327-FG2156) FG2328-FG2156) FG1975-FG2338) FG1889-FG2336) FG1888-FG2337) FG2335-FG2337) FG2326-FG2337) FG2154-FG2337) FG1889 - FG2337) FG2451-FG2337) FG2154-FG2452) FG2454-FG2156) FG2453-FG2156) FG2453-FG2455) FG2457-FG2338) FG2456-FG2336)

ColEI ori, lptAp-lptA lptB (3341138 e 3342781), AmpR; from pRS415 by replacing the EcoRIeStuI fragment (lac operon) with the EcoRIeStuI digested AP069-FG1908 PCR product of BW25113 DNA ColEI ori; overexpresses ribosomal protein S1 upon IPTG induction pACYC184 derivative carrying the lacI gene pBR322 derivative; harbors the entire lac operon without promoter; AmpR

each of the three promoter proximal regions (YRBGR, LPTCR and LPTAR, respectively; Fig. 1) produced very faint and smeared signals, probably because of low abundance and instability of the transcripts of this locus in reference strain BW25113 (not shown). In conditions thought to stabilize mRNA, namely, upon over expression of the ribosomal protein

This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work (Alba and Gross, 2004) (Guzman et al., 1995) (Cherepanov and Wackernagel, 1995) This work (Sukhodolets and Garges, 2003) Qiagen (Simons et al., 1987)

S1 (Briani et al., 2008) and in an RNase E thermosensitive mutant at non-permissive temperature (Ghora and Apirion, 1978), the signals (as well as the smears) became more intense (Fig. 2) and enabled a tentative transcriptional map of this locus (Fig. 1, bottom). In both conditions, RNA probe YRBGR detected an mRNA species of about 4.5 kb (Fig. 2A),

A.M. Martorana et al. / Research in Microbiology 162 (2011) 470e482

473

Table 2 Oligonucleotides. Name

Sequencea (50 to 30 )

Use/description

AP036 AP037 AP069 AP070 AP100 AP101 FG1888 FG1889 FG1892 FG1893 FG1899 FG1900 FG1907 FG1908 FG1973 FG1974 FG1975 FG1988 FG2000 FG2001 FG2005 FG2006 FG2007 FG2061 FG2062 FG2063 FG2064 FG2154 FG2155 FG2156 FG2326 FG2327 FG2328 FG2335 FG2336 FG2337 FG2338 FG2367 FG2421 FG2451 FG2452 FG2453 FG2454 FG2455 FG2456 FG2457

ctaatacgactcactatagggcATTGGTCAGCTTGGCTTTATC ATGAGTAAAGCCAGACGTTGGG cgagaggaattcGGTCCGTAAAAGCAGATAAAGCC cgacgcggatccATTAAGGCTGAGTTTGTTTG cgagaggaattcACGCCGTGTCTTCGATATGG cgagaggaatTCGGTGTGTTACATATGCATG cgagaggaattCGGCAACGTTTGCCAGCACCG cgagaggaattCGCAGATGTGCATCTGTGTGT ctaatacgactcactatagggATCGCTGCGACTAAGTTGTC CGCGTCTATTCTTTGCCGAA ctaatacgactcactatagggGGTAGAACGTTGCCGGTTTA CCGTAACCGGAGACACTGAT ccggaattccggAGATCAATCTGGTGACGCAG cttccggaagATCGCCAGTTGTTGGCTAAGC cgcggatccgcgGAATCTTTACACTACGCCTG ccggaattccggATACGTGTGCGCCCTGGCAT ccggaattccggCTTCTGCTGCGCGTAAACGA ACGGCCAGTGAATCCGTAAT b TGCTGGTCCGATTCAATGTG AAGTGAGCTGGCAAGCACAA CTATAACGTCGGCGCTGACA ACCGTCACGAACATTGAACG TAAGTTCAGTGCCTCAACGG cgacgcgaattccAAATGTTGTTCCGTATAACG cgacgcgaattcGACCTCGTCACGTTATACGG cgacgcgaattCTGCGTCACCAGATTGATCTG cgacgcgaatTCCAGCGGTCTGAAAATGCGCG ccggaattcATCCACTGGCGTACAGTCGT cgcggatccAACACACAGATGCACATCTG cgcggatccGGATCTCATCGCCCGTATGC cgagaggaattcCAGGATGTGGTGATTGCTATC cgagaggaattcATCAGAATTTCACGCTTGCC cgagaggaattcACGCATTATGTCGCACGTAG ccggaattccggAGTAGCAAAGAAGCCTGTC cgcggatccgCGAAGACACGGCGTAAATCACC cgcggatccgcgATCGCCCATAACCAGCGTG cgcggatccgcgCGAGTAAATGGTCGCCATCG AGATTAAGGCTGAGTTTGTTTG AAGCCGCGTGCTTTTAACAG cgagaggaattCCGTTAATCTGCATCACCGGT cgacgcggatccACCGGTGATGCAGATTAACG cgagaggaattcCGCCACGCTGGTTATGGGCG cgagaggaattcCGCGGCTTTACTGCTGAAGA cgacgcggatccTCTTCAGCAGTAAAGCCGCG cgagaggaattcCGCATGTTAAGAAAACGGCC cgagaggaattcCGATATGGGCGTGGATGTTC

LPTCP riboprobe with AP037 LPTCP riboprobe with AP036 Plasmid construction; sequencing fragment Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction; sequencing fragment YRBGP riboprobe with FG1893 YRBGP riboprobe with FG1892 LPTAP riboprobe with FG1900 LPTAP riboprobe with FG1899 Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction PE on pAM derived transcripts 50 -end Sequencing fragment PE for lptAp 50 -end PE for kdsCp 50 -end Sequencing fragment PE for kdsCp 50 -end Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction PE for lptAp 50 -end PE for kdsCp 50 -end Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction Plasmid construction

a b

Upper case, E. coli genomic sequences; lower case, T7 promoter sequence; underlined lower case, restriction sites or restriction site-compatible overhangs. Sequence is from pRS415 (NCBI Accession Number U03449).

which is compatible with a transcript covering the entire locus. This idea is supported by the fact that a transcript of comparable molecular weight was also detected by the LPTCR and LPTAR riboprobes (Fig. 2B and C). Signals of lower molecular weight in the three panels may be imputed to transcripts terminated within the locus, processed and/or initiated at promoters downstream of yrbGp. No signals coincident with the 4.5 kb transcript and the shorter ones shown in Fig. 2 were found using a probe internal to the downstream gene rpoN (data not shown), suggesting that the above transcripts terminate within the yrbG locus. In contrast, longer than 4.5 kb

signals could be observed with all probes, thus indicating that transcripts traversing the yrbG locus may extend into the downstream rpoN locus. Overall, these data suggest that this locus is organized into three overlapping operons, yrbG, kdsC, and lptA, each extending at least to lptB. 3.2. Characterization of the kdsCp region To identify the kdsCp promoter predicted by previous genetic analysis (Sperandeo et al., 2006), we assayed lacZ expression in a set of fusions with different portions of the

474

A.M. Martorana et al. / Research in Microbiology 162 (2011) 470e482

Fig. 1. Map of the E. coli yrbG-lptB locus. Upper part: the genes of the locus are drawn to scale as open rectangles with arrowhead. Coordinates in the kb scale are obtained from the E. coli genomic sequence (GenBank NC_000913) by subtracting 3337. Promoters are indicated by bent arrows. Middle part: riboprobes (closed bars) and oligonucleotides (arrowheads) used for transcriptional mapping. Lower part: transcriptional map (see text). Approximate transcript length, indicated in kb on the left, was determined by comparison to molecular weight of internal standards (rRNAs); the 50 ends of the transcripts were mapped by primer extension.

kdsC upstream region (Fig. 3, pAM1 to pAM5). Promoter activity has been reported in a Shigella flexnerii genomic clone containing the 30 -end of yrbG and part of kdsD (Bartoleschi et al., 2002); this putative promoter activity was thus tested with lacZ fusions with the E. coli homologous region (pAM7) and different portions thereof (Fig. 3). The results are presented in Fig. 3. Overall, results suggested a rather complex promoter region upstream of kdsC that extends within the 30 -half of kdsD. Slight but significant b-galactosidase activity was detected in

pAM1 and increasingly higher enzymatic levels were expressed by pAM2 to pAM5, which harbor progressively longer extensions at the 50 -end. This suggested that pAM1 contains a kdsC-proximal promoter (henceforth indicated as kdsCp1) and that either an activating region or an additional promoter(s) is located upstream of kdsCp1. The latter hypothesis seemed to be supported by pAM18 and pAM32, which did not contain kdsCp1 and expressed lacZ at substantial levels, suggesting the presence of a second promoter, kdsCp2.

Fig. 2. Northern blot analysis of transcripts from yrbG-lptB locus. BW25113/pREP4/pQE31S1 was grown in LD at 37 with aeration up to OD600 0.4 and S1 expression induced with 1 mM IPTG; KG243 (rnets) was grown in LD at 30  C with aeration up to OD600 0.4 and quickly transferred to 44  C. Aliquots were sampled immediately before (0) and at different times (indicated in min above the lanes) after IPTG induction or the temperature upshift; RNA was extracted, resolved by denaturing 1.5% agarose gel electrophoresis and analyzed by northern blot hybridization with the radiolabeled riboprobes YRBGR (A), LPTCR (B) and LPTAR (C). The size of the rRNA and RNAs is indicated (in kb) on the left.

A.M. Martorana et al. / Research in Microbiology 162 (2011) 470e482

475

Fig. 3. Characterization of the kdsCp promoter region. The region is drawn to scale and coordinates on the chromosome bar are given in nt from the kdsC start codon (NC_000913 3,340,295). Promoter transcription start points are indicated as bent arrows. In the lower part, the genomic regions cloned in pRS415 derivative plasmids are drawn. Coordinates of primer extension signals are reported on the chromosome bar relative to the kdsC start codon. The positions of the 50 -ends determined by PE using oligo FG1988 (at the 50 -end of lacZ ) and FG2007 or FG2421 (on the genomic sequence) are indicated by downward solid and open arrowheads, respectively. Signals detected in the plasmidless strain are reported on the chromosome bar. Large and small symbols indicate comparatively stronger and weaker extension signals, respectively; v: position of a predicted signal not experimentally demonstrated on the specific plasmid marked or in the plasmidless strain (marks on the chromosome bar). The panel combines the data of several PE experiments some of which are shown in Fig. 4. On the right, b-galactosidase specific activity expressed by the plasmids is reported. The values are the average and the standard deviation of at least three independent experiments.

The kdsC distal region of the fragments cloned in pAM5 (which contains both kdsCp1 and kdsCp2) and pAM19 (which contains kdsCp2) overlapped 219 bp of the kdsC proximal region of pAM7. Both pAM5 and pAM19 expressed higher b-galactosidase level than pAM4 and pAM32, respectively, thus suggesting the presence of a promoter upstream of the pAM4 and pAM32 fragments. This promoter (kdsCp3) was delimited by a series of deletions of the pAM7 fragment on either side; from the data reported in Fig. 3, it appears that kdsCp3 was not present in pAM22, whereas it seemed to be harbored by pAM25. Noticeably, all fragments extending upstream of the left end of pAM14 and pAM25 exhibited lower b-galactosidase activity. Further evidence of the three kdsCp promoters was obtained by primer extension analysis. Preliminary approaches to detect the 50 -end of the transcripts with a primer in the 50 end region of kdsC on RNA extracted from the wild type strain were unsuccessful because of very faint and unreliable signals obtained (not shown). To enhance the signals and (at least partially) avoid interference from processing of transcripts originating at the upstream promoters, we performed primer extension on RNA extracted from strains harboring the above plasmids using primers complementary to the 50 -end of the reporter gene (FG1988, Table 2). To map lacZ-distal signals,

additional primers designed against the genomic sequence (Fig. 1 and Table 2) were also used. The results of several experiments are summarized in Fig. 3 and the most significant examples are reported in Fig. 4AeE. Coordinates are from the kdsC start codon. A 47 signal, present in pAM1 through pAM5 (Fig. 4A), indicated that a kdsCp1 was contained in the region harbored by pAM1. The features of an “extended 10” promoter (Mitchell et al., 2003), which may compensate for a poor 35 element, were found in the promoter region (Fig. 4F). The additional signal detected at 56 in pAM4 and pAM5 could be a processing site of mRNA from the upstream promoter(s) since it was missing in pAM1, 2, and 3 (Fig. 4A), which lacked kdsCp2. Primer extension using plasmidless RNA strains failed to reveal the kdsCp1 transcription start point even in the rnets mutant at 42  C (data not shown). This negative result might depend on the weakness of kdsCp1 and/ or on some silencing mechanism acting on the chromosomal copy of this promoter. Primer extension on pAM18 with the primer in lacZ gave a signal at 97 and a second upstream signal; this was mapped at 263 using a more closely located primer and RNA from pAM18 (Fig. 4B) and the plasmidless strain (Fig. C). The 263 position was preceded by reasonable 10 and 35 sD

476

A.M. Martorana et al. / Research in Microbiology 162 (2011) 470e482

Fig. 4. Identification of kdsCp transcription start points. RNA preparations from strains harboring pAM plasmids indicated above the lanes (panels A, B, D, and E); plasmidless BW25113 (panels C and F) or KG243 (rnets; panel F) strains were used for PE analysis with the radiolabeled primer FG1988 (complementary to lacZ; panels A, B and D), FG2007 (genomic; panel C) and FG2421 (genomic; panels E and F), respectively. Coordinates are from the kdsC start codon. DNA sequencing reactions of pAM5 (panels A and E), pAM7 (panel F), pAM24 (panel D) or of a PCR-amplified genomic fragment (primers FG1888-FG2005; panels B and C) obtained with the same oligonucleotides were run in the same gel. G: Sequence of the promoter regions. The 50 -end detected by PE is in bold letters. Putative sD 35/10 hexamers are boxed; the TRTGN motif of extended 10 promoters is underlined.

motifs and could thus represent the kdsCp2 transcription start point (Fig. 4F). Primer extension with pAM4 and pAM5 RNAs gave an upstream signal (X in Fig. 4A) compatible with the 263 end, whereas the 97 signal was absent. The 97 and the 56 signals could be the 50 -end of decay intermediates and the 97 signal could only be detected by primer extension when the 56 processing site was missing (pAM18). A primer extension signal at 554, compatible with kdsCp3, was also detected (oligo in lacZ with pAM23 and pAM24, Fig. 4D, and oligo within kdsD with pAM5 and the tested plasmids harboring this region, Fig. 4E). This signal was also detected in the plasmidless wild type and rnets strains (Fig. 4F). As was the case for kdsCp1, an extended 10 box could be found upstream of 554 (Fig. 4F). This 50 -signal was only 10 nt from the left end of the pAM22 fragments, which did not

express b-galactosidase activity and thus could not encompass the entire promoter, and only 34 nt from the left end of the pAM5, pAM19, pAM14, and pAM25 fragments. The higher b-galactosidase activity exhibited by pAM14 and pAM25, which did not harbor other downstream promoters, might depend on the serendipitous creation of a better-performing 35 box at the junction of the vector with fragments starting at 564 (Fig. 4F, hybrid promoter). An additional primer extension signal, not present in plasmid-bearing strains, was also detected in plasmidless strains at position 579 (Fig. 4F). This may likely represent a 50 -end of decay intermediates of transcripts starting from the upstream promoter yrbGp. Expression of kdsCp1, kdsCp2 and kdsCp3 was neither stimulated by the DrseA mutation (in strain BB-11) nor abolished by rpoN or rpoS mutations in KG245 and KG246,

A.M. Martorana et al. / Research in Microbiology 162 (2011) 470e482

respectively; data not shown), thus suggesting that such promoter activities were not driven by sE, sN, or sS factors. 3.3. Characterization of lptAp A sE-dependent promoter upstream of lptAB was previously identified by genetic means and putative sE 10 and 35 motifs were suggested (Sperandeo et al., 2006; Fig. 5A). To characterize this promoter, we performed primer extension analyses on RNA extracted from the wild type and DrseA isogenic mutant, which overexpresses sE (Missiakas et al., 1997). To enhance the primer extension signal, the above strains harboring the lptAB region with its promoter on a multicopy plasmid (pLPT-AB) were also used. A signal at 54 from lptA (7 nt downstream of the putative sE 10 box) was detected in all samples (Fig. 5B). In the wild type strain without pLPT-AB, it was barely visible with longer exposures (not shown). Moreover, in the DrseA strain the signal was stronger than in the corresponding wild type strain. This signal is consistent with a transcript expressed from the sE-dependent promoter, henceforth designated lptAp1. In addition, a 81 signal, slightly stronger in pLPT-AB carriers, was detected in all strains. This was compatible with the presence of a second promoter, lptAp2. Weak putative 10

477

and 35 sD boxes are suggested in Fig. 5A. To delimit lptAp1 and functionally prove the presence of lptAp2, we assayed lacZ expression in constructs shown in Fig. 5A. Only pAM8 (263 to þ 33) and pAM9 (105 to þ33) exhibited a significant RseA-responsive b-galactosidase activity (Fig. 5C); in contrast, deletion of the putative lptAp1 -35 box (pAM10) abolished lacZ expression, in agreement with the localization of the putative sE 10 and 35 boxes. On the contrary, a very weak b-galactosidase activity was detected in pAM11, in which the lptAp1 -10 box is deleted; this activity further decreased in pAM12 and pAM13 (Fig. 5C, inset). In the DrseA strains harboring such plasmids, b-galactosidase activity was low, although somewhat erratic. Overall, these data suggest that lptAp2 is a very weak (possibly sD-dependent) promoter with a putative activating upstream region. 3.4. Selective response of lptAp1 to different extracytoplasmic stresses The sE-dependent promoters are activated by extracytoplasmic stress (Rhodius et al., 2006). We thus tested whether lacZ under lptAp1 could be induced by stressful conditions known to induce the sE regulon. Surprisingly, overexpression of

Fig. 5. Organization of lptA promoter region. (A) Map of lptAp region. Coordinates in the scale bar (upper) are expressed in nucleotides from the lptA start codon (E. coli 3,341,402). The 35/-10 for sE and sD consensus are drawn to scale and the sequence of the promoter regions is reported above the map. The transcription start points of the promoters are indicated as bent arrows in the map and underlined in the sequence. The genomic fragments cloned in pRS415 as transcriptional fusions with the reporter lacZ gene (plasmids pAM8-pAM13) are shown below. (B) Mapping of lptA promoter transcription start points. PE was performed with the radiolabeled oligonucleotide FG2001 on RNA extracted from BW25113 and its isogenic DrseA mutant BB-11 and their derivatives transformed with pLPT-AB. DNA sequencing reactions of pLPT-AB obtained with the same oligonucleotide were run in the same gel. Coordinates of the PE signals from the kdsC start codon are indicated on the right. (C) b-galactosidase activity expressed from lptAp-lacZ fusions in BW25113 (open bars) and BB-11 (solid bars) strains. Specific activities are reported as average (with standard deviation) of three independent assays. Results obtained with plasmids that map lptAp2 are enlarged in the inset.

478

A.M. Martorana et al. / Research in Microbiology 162 (2011) 470e482

the OM protein OmpC, exposure to 10% ethanol, 0.2% Nadeoxycholate, hyperosmolarity (20% sucrose) and hypoosmolarity (0.2% NaCl) shocks did not increase lacZ expression under either lptAp1 (pAM9) or both lptAp1 and lptAp2 (pAM8; data not shown). To verify that our experimental conditions effectively induced the sE regulon, we assayed by northern blotting the expression of MicA, a sE-regulated small RNA (Udekwu and Wagner, 2007). Exposure of the cell cultures to the above extracytoplasmic stresses increased, albeit at variable levels, the expression of MicA (Fig. 6A), suggesting that regulation of the sE-dependent lptAp1 promoter may differ from other extracytoplasmic stress-inducible promoters. A twofold induction of lptAp1 and an increase in the MicA RNA level, however, occurred upon treatment with ammonium metavanadate (Fig. 7A and Fig. 6A), a phosphatase inhibitor known to induce modifications in LPS fatty acid chains (Zhou et al., 1999), suggesting that lptAp1 might be activated only by a subclass of envelope stresses. Interestingly, metavanadate induced anomalous forms of LPS detectable as higher-molecular weight bands in PAGE (Fig. 7C), that are reminiscent of the colanic acid-decorated LPS produced upon depletion of Lpt proteins (Ma et al., 2008; Sperandeo et al., 2008). We thus tested whether impairment of LPS transport machinery following Lpt protein depletion could induce a sE-dependent stress response and lptAp1. Cultures of E. coli BB-4 and BB-3 conditional expression mutants (whith lptA and lptC, respectively, under control of araBp) harboring the above constructs were grown with arabinose (permissive conditions) and assayed for MicA and b-galactosidase expression upon a shift in an arabinose-free medium (depletion). The MicA level was increased upon depletion (Fig. 6B), indicating that the lack of Lpt proteins induced the extracytoplasmic stress

response. Likewise (Fig. 8A), lptC depletion increased bgalactosidase expression more than fivefold, whereas lptA lptB depletion appeared to have a less marked effect. It should be noted, however, that under the permissive condition, the level of b-galactosidase was significantly higher in BB-4 than in the wild type control and in BB-3. In BB-3 the three lpt genes may be cotranscribed from araBp, whereas in BB-4 lptC is under its natural promoter while lptAB expression is driven by araBp. An unbalanced expression of LptAB versus LptC might cause a chronic envelope stress condition and activate lptAp1, thus accounting for a higher lacZ expression level from the plasmids in BB-4 under permissive conditions and a further moderate increase upon exacerbation of the stress by lptA lptB depletion. Primer extension with a lacZ-primer showed that lptAp1 signal intensity from both pAM8 and pAM9 increased upon lptAB or lptC depletion (Fig. 8B). Likewise, vanadate treatment increased both the chromosomal and pAM8 lptAp1 signal intensities (Fig. 7B), whereas no significant variation was seen in other types of extracytoplasmic stress tested (not shown). Overall our data suggest that the sE-dependent lptAp1 promoter responds to a restricted subclass of extracytoplasmic stresses that seem to directly affect LPS biogenesis. 4. Discussion Polycistronic operons are thought to coordinate expression of functionally related genes operating in a single pathway, and/or to couple different, reciprocally interdependent processes through coordinate expression of key functions. An example of the latter type is the rpsU dnaG rpoD operon, which codes for genes implicated in translation, replication and transcription, respectively (Lupski et al., 1983). In

Fig. 6. Northern blot analysis of micA expression upon extracytoplasmic stress. Stress conditions were performed as described in Materials and Methods. RNA was extracted, resolved by denaturing 6% PAGE and analyzed by northern blot hybridization with the radiolabeled probe against micA transcript (oligonucleotide FG2355 (50 -GAAAAAGGCCACTCGTGAGTG-30 ), complementary to the 30 -end of micA transcript). (A) Extracytoplasmic stress was induced by exposure to E, 10% ethanol; S, 20% sucrose; H, hypotonic medium; V, 25 mM ammonium vanadate; u, untreated cells. (B) Extracytoplasmic stress was induced by depletion of LptA and B (left panel) and LptC (right panel), as described in Materials and Methods.

A.M. Martorana et al. / Research in Microbiology 162 (2011) 470e482

479

Fig. 7. Induction of lptAp1 by ammonium metavanadate. (A) b-galactosidase activity expressed from lptAp-lacZ fusions upon 25 mM NH4VO3 treatment (solid bars) and in untreated cultures (open bars). Specific activity is reported as the average of two independent assays with standard deviation. (B) PE analysis of lptAp1 upon NH4VO3 stress. PE was performed on RNA from BW25113 (left) and its derivative harboring the lptAp1-lacZ fusion (pAM8, right) using the radiolabeled FG2001 and FG1988 oligonucleotides, respectively. DNA sequence reactions of pAM8 and a PCR fragment (AP069-FG2000 primers) obtained with the same primers used for PEs were run in the same gel. (C) LPS profiles of BW25113 (wt) and BB-11 (DrseA) upon treatment with metavanadate. As a control, LPS extracted from depleted (- ara) and non-depleted (þara) BB-4 mutant (araBp-lptA) is shown. LPS was extracted from exponentially growing cell cultures and separated on 18% Tricine-SDS-PAGE and silver-stained.

addition, genetic linkage may favor co-evolution of gene clusters. Differential expression of genes encoded in a polycistronic operon may be regulated at both transcriptional (intraoperonic promoters and transcriptional terminators) and posttranscriptional (mRNA processing, stability and translatability) levels. The yrbG operon encodes, in addition to the putative divalent cation transporter YrbG (Saaf et al., 2001), housekeeping cell functions for OM biogenesis, namely KdsC and KdsD, two enzymes implicated in Kdo biosynthesis, and three genes encoding components of the LPS transport machine (LptA, LptB and LptC) and might thus coordinate LPS synthesis with LPS transport. 4.1. Evidence for a yrbG-lptB operon The six ORFs of the yrbG-lptB locus are either overlapping or spaced by few nucleotides. This physical organization suggested that the yrbGp promoter, previously documented by primer extension, could drive transcription of the entire locus (Sperandeo et al., 2006). Two additional promoter regions (kdsCp and the sE-dependent lptAp) had been identified by genetic analysis (Sperandeo et al., 2006), suggesting differential expression of these genes at least under specific conditions. To shed light on the regulation of this locus we defined a transcriptional map and characterized the internal promoter regions. Northern analysis produced significant signals only in conditions of mRNA stabilization by S1 overexpression (Briani

et al., 2008) or in a thermosensitive RNase E mutant at nonpermissive temperature (Ghora and Apirion, 1978). This suggests that transcription of this locus may not be very efficient and, in any case, mRNAs are very rapidly degraded. Signals of 4.5 kb and longer could be detected with a probe in yrbG. Since, upstream of yrbGp, a divergent operon is found, it is unlikely that these transcripts originate upstream of yrbGp and thus they cover the entire locus. Longer transcripts may extend into rpoN and yonder, consistent with the fact that no intrinsic transcription terminators were detected by inspection of the 47-nucleotides-long lptB-rpoN intergenic region. Thus, the rpoN operon, although provided by its own promoter (Jones et al., 1994; Powell et al., 1995), may also be transcribed from upstream promoters of the yrbG locus. The linkage between the yrbG-lptB and the rpoN loci is evolutionarily well conserved in several g-Proteobacteria orders (Enterobacteriales, Alteromonadales, and Vibrionales), whereas a similar organization, in which only yrbG is missing, is conserved in Pseudomonadales and Xanthomonadales (V. Piccolo and G. Deho`, unpublished). Overall, these observations suggest regulatory coupling between LPS biogenesis and sN gene expression. 4.2. The kdsCp promoter region Although the yrbG-lptB operon encodes three essential genes, inactivation of yrbGp is not lethal, since at least two downstream promoter regions, kdsCp and lptAp, appear to be sufficient to express the essential lptC, lptA, and lptB genes

480

A.M. Martorana et al. / Research in Microbiology 162 (2011) 470e482

Fig. 8. Activation of the sE-dependent promoter lptAp1 upon lptAB and lptC depletion. Promoter activity of lptAp1-lacZ and lptAp2-lacZ fusions in BW25113 (wt), BB4 (araBp-lptA), and BB3 (araBp-lptC ) strains was measured. Exponentially growing cultures (OD600 ¼ 0.2) in LD-arabinose were harvested, washed and diluted one-hundredfold in LD (non-permissive condition, open bars) or LD-arabinose (permissive conditions, solid bars) and grown up to OD600 ¼ 0.4. Samples were collected for b-galactosidase assay (A, average of two experiments) and PE analysis (B and C) with a radiolabeled primer (FG1988) complementary to the lacZ 50 -end.

(Sperandeo et al., 2006). Using transcriptional fusion with lacZ and primer extension analysis, we detected three promoters upstream of kdsC, namely kdsCp1, kdsCp2 and kdsCp3 located within kdsD 47, 263 and 554 nt, respectively, upstream of the kdsC start codon. The role of such promoters is yet to be defined. Their relatively low levels of basal expression suggest that these promoters might be activated under specific environmental conditions. Bartoleschi et al. (2002) isolated a genomic fragment of S. flexneri encompassing the yrbG 30 -end and the kdsD 50 -end (a region where E. coli and S. flexneri share >99% sequence identity), which appeared to contain a promoter activity turned up in HeLa cells. Further studies will be necessary to define whether kdsCp3, which is located at the 30 -end of such a fragment, or another upstream promoter not detected under our growth conditions, is responsible for such activation. 4.3. The lptAp promoter region and the LPS stress response Upstream of lptA we identified two closely spaced promoters. lptAp1, previously predicted by genetic analysis, is a sE-dependent promoter, whereas lptAp2 is a weak s70 promoter. Although lacZ expression driven by the cloned lptAp2 on the reporter plasmids is weak, the primer extension signal of this promoter is stronger than that of lptAp1 and less sensitive to the multiple copies of the plasmid. It is thus possible that this promoter on the chromosome is more

efficient than on the plasmid. This suggests the presence of upstream regulatory regions missing in the plasmid constructs or lptAp2 activation by transcription from upstream promoters. Alternatively, the lptAp2 transcription start point might be coincident with a processing site of transcripts originating at upstream chromosomal promoters. We delimited the extension of lptAp1 and confirmed its sE dependence, as its activity increases up to tenfold in a DrseA mutant. sE is the main regulator of the extracytoplasmic stress response, although other extracytoplasmic stress signaling pathways such as Cpx, Psp, Bae and Rcs are known to operate in E. coli (Rowley et al., 2006). A sE-dependent promoter about 70 nt downstream of lptAp1 and with a predicted transcription start point 20 nt downstream from the lptA start codon was postulated based on a bioinformatics approach (Rhodius et al., 2006; Rhodius and Mutalik, 2010). However, this hypothetical promoter, which would drive lptB (designated yhbG in the cited works) but not lptA expression, was shown to be nearly inactive both in vitro and in vivo (Mutalik et al., 2009; Rhodius and Mutalik, 2010). Moreover, pAM10, which is presumed to contain yhbGp, showed neither sE-independent nor sE-dependent lacZ expression, as compared with pAM9 (Fig. 5C). Thus, lptAp1 is the bona fide sE-dependent promoter that drives expression of the bicistronic lptA operon. Interestingly, lptAp1 appears not to respond to any type of stress known to activate the sE-dependent pathway. In particular, lptAp1 does not respond to OmpC overexpression,

A.M. Martorana et al. / Research in Microbiology 162 (2011) 470e482

known to induce an envelope stress response (Ades et al., 1999), whereas under the same conditions, another sEdependent gene such as micA is activated. On the contrary, we observed that lptAp1 is activated under very specific conditions that affect LPS composition and biogenesis, such as depletion of LptA, LptB and LptC, which are implicated in LPS transport, or by metavanadate treatment. Depletion of any of the seven Lpt proteins composing the LPS transport machine responsible for LPS transport to the cell surface leads to arrest of cell growth, blockage of newly synthesized LPS transport to the OM, formation of abnormal membrane structures in the periplasm and synthesis of an anomalous form of LPS decorated by colanic acid that accumulates in the outer leaflet of the IM (Ma et al., 2008; Sperandeo et al., 2008). Ammonium metavanadate is known to induce covalent modifications of E. coli K12 lipid A, resulting in the accumulation of six major lipid A variants derivatized with palmitate, L-4-aminoarabinose, and/or phosphoethanolamine (Zhou et al., 1999). Such modifications have been implicated in induction of sE regulon (Tam and Missiakas, 2005). In addition we have observed, in metavanadate-treated cells, production of high molecular weight forms of LPS that are reminiscent of those observed in Lpt proteins-depleted cells. Both Lpt protein depletion and ammonium metavanadate treatment appeared to induce the sE regulon, as suggested by the increased abundance of the small RNA MicA. However, the sEdependent lptAp1 promoter is not responsive to any stress inducing the sE regulon, but only to a subclass of stresses. A common feature of the two lptAp1-inducing conditions is covalent modification of LPS in either the lipid or saccharide moieties. It thus appears that LPS modification may signal envelope stress and modulate expression of LptA and LptB. Since lptAp1 is sE-dependent but is not activated by several specific extracytoplasmic stresses inducing sE activity, this implies that additional LPS stress-specific factors control lptAp1. Two mechanisms could be postulated: i) lptAp1 is negatively controlled by a repressor inactivated (for example by proteolysis) upon LPS stress; alternatively, ii) a positive regulator, in addition to sE, is required for lptAp1 transcriptional activation. This putative transcription factor might be activated either transcriptionally or post-transcriptionally (e.g. by proteolysis of its specific inhibitor) by an LPS stressdependent pathway. Our data further suggest that the cisacting determinants for such regulation are present in the pAM8 construct and are thus either overlapping or downstream of the sE-dependent lptAp1 promoter. Identifying the mechanisms specifically controlling lptAp1 will shed new light on extracytoplasmic stress signaling pathways. In conclusion, the yrbG locus encodes genes for LPS synthesis and transport that are co-expressed from an upstream promoter and several internal promoters that may differentially regulate expression of the encoded genes. Transcription may also extend to the downstream rpoN operon. In E. coli, RpoN (sN) is reported to regulate transcription of genes implicated in diverse functions such as nitrogen metabolism, pili and flagella biosynthesis, and quorum sensing. Therefore,

481

expression of the the yrbG-rpoN locus may coordinate OM biogenesis and RpoN-dependent functions. All these functions are regulated to maintain cell homeostasis, allow growth, division and development, and respond to internal and external insults that may damage cell integrity. Differential expression of these genes may thus be required under different physiological and environmental circumstances. Clustering of genes implicated in such functions in the yrbG locus may be functional for their coordinated action. Acknowledgments We are grateful to Sandro Zangrossi for technical support and advice. We also thank Carol Gross for providing pBA114 and pBAD33, and the National BioResource Project (NIG, Japan) for providing E. coli strains. This work was partially supported by joint grants from the “Ministero dell’Istruzione, dell’Universita’ e della Ricerca” and the “Universita´ degli Studi di Milano” (Programmi di Rilevante Interesse Nazionale 2007). References Ades, S.E., 2008. Regulation by destruction: design of the sE envelope stress response. Curr. Opin. Microbiol. 11, 535e540. Ades, S.E., Connolly, L.E., Alba, B.M., Gross, C.A., 1999. The Escherichia coli sE-dependent extracytoplasmic stress response is controlled by the regulated proteolysis of an anti-s factor. Genes Dev. 13, 2449e2461. Alba, B.M., Gross, C.A., 2004. Regulation of the Escherichia coli sEdependent envelope stress response. Mol. Microbiol. 52, 613e619. Baba, T., Ara, T., Hasegawa, M., Takai, Y., Okumura, Y., Baba, M., Datsenko, K.A., Tomita, M., Wanner, B.L., Mori, H., 2006. Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol. Syst. Biol. 2, 2006. Bartoleschi, C., Pardini, M.C., Scaringi, C., Martino, M.C., Pazzani, C., Bernardini, M.L., 2002. Selection of Shigella flexneri candidate virulence genes specifically induced in bacteria resident in host cell cytoplasm. Cell Microbiol. 4, 613e626. Bertani, G., 1951. Studies on lysogenesis. I. The mode of phage liberation by lysogenic Escherichia coli. J. Bacteriol. 62, 293e300. Briani, F., Curti, S., Rossi, F., Carzaniga, T., Mauri, P., Deho`, G., 2008. Polynucleotide phosphorylase hinders mRNA degradation upon ribosomal protein S1 overexpression in Escherichia coli. RNA 14, 2417e2429. Briani, F., Del Favero, M., Capizzuto, R., Consonni, C., Zangrossi, S., Greco, C., De Gioia, L., Tortora, P., Deho`, G., 2007. Genetic analysis of polynucleotide phosphorylase structure and functions. Biochimie 89, 145e157. Carzaniga, T., Briani, F., Zangrossi, S., Merlino, G., Marchi, P., Deho`, G., 2009. Autogenous regulation of Escherichia coli polynucleotide phosphorylase expression revisited. J. Bacteriol. 191, 1738e1748. Cherepanov, P.P., Wackernagel, W., 1995. Gene disruption in Escherichia coli: TcR and KmR cassettes with the option of Flp-catalyzed excision of the antibiotic-resistance determinant. Gene 158, 9e14. Costanzo, A., Ades, S.E., 2006. Growth phase-dependent regulation of the extracytoplasmic stress factor, sE, by guanosine 3’,5’-bispyrophosphate (ppGpp). J. Bacteriol. 188, 4627e4634. Costanzo, A., Nicoloff, H., Barchinger, S.E., Banta, A.B., Gourse, R.L., Ades, S.E., 2008. ppGpp and DksA likely regulate the activity of the extracytoplasmic stress factor sE in Escherichia coli by both direct and indirect mechanisms. Mol. Microbiol. 67, 619e632. Datsenko, K.A., Wanner, B.L., 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97, 6640e6645.

482

A.M. Martorana et al. / Research in Microbiology 162 (2011) 470e482

Deho`, G., Zangrossi, S., Sabbattini, P., Sironi, G., Ghisotti, D., 1992. Bacteriophage P4 immunity controlled by small RNAs via transcription termination. Mol. Microbiol. 6, 3415e3425. Ghisotti, D., Chiaramonte, R., Forti, F., Zangrossi, S., Sironi, G., Deho`, G., 1992. Genetic analysis of the immunity region of phage-plasmid P4. Mol. Microbiol. 6, 3405e3413. Ghora, B.K., Apirion, D., 1978. Structural analysis and in vitro processing to p5 rRNA of a 9S RNA molecule isolated from an rne mutant of E. coli. Cell 15, 1055e1066. Guzman, L.M., Belin, D., Carson, M.J., Beckwith, J., 1995. Tight regulation, modulation, and high-level expression by vectors containing the arabinose pBAD promoter. J. Bacteriol. 177, 4121e4130. Hanahan, D., 1983. Studies on transformation of Escherichia coli with plasmids. J. Mol. Biol. 166, 557e580. Jones, D.H., Franklin, F.C., Thomas, C.M., 1994. Molecular analysis of the operon which encodes the RNA polymerase s factor s54 of Escherichia coli. Microbiology 140, 1035e1043. Kunz, D.A., Chapman, P.J., 1981. Catabolism of pseudocumene and 3-ethyltoluene by Pseudomonas putida (arvilla) mt-2: evidence for new functions of the TOL (pWWO) plasmid. J. Bacteriol. 146, 179e191. Lupski, J.R., Smiley, B.L., Godson, G.N., 1983. Regulation of the rpsU-dnaGrpoD macromolecular synthesis operon and the initiation of DNA replication in Escherichia coli K-12. Mol. Gen. Genet. 189, 48e57. Ma, B., Reynolds, C.M., Raetz, C.R., 2008. Periplasmic orientation of nascent lipid A in the inner membrane of an Escherichia coli LptA mutant. Proc. Natl. Acad. Sci. U. S. A 105, 13823e13828. Miller, J.H., 1992. A Short Course in Bacterial Genetics: A Laboratory Manual and Handbook for Escherichia Coli and Related Bacteria. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Missiakas, D., Mayer, M.P., Lemaire, M., Georgopoulos, C., Raina, S., 1997. Modulation of the Escherichia coli sE (RpoE) heat-shock transcription-factor activity by the RseA, RseB and RseC proteins. Mol. Microbiol. 24, 355e371. Mitchell, J.E., Zheng, D., Busby, S.J., Minchin, S.D., 2003. Identification and analysis of ’extended -10’ promoters in Escherichia coli. Nucleic Acids Res. 31, 4689e4695. Mutalik, V.K., Nonaka, G., Ades, S.E., Rhodius, V.A., Gross, C.A., 2009. Promoter strength properties of the complete sE regulon of Escherichia coli and Salmonella enterica. J. Bacteriol. 191, 7279e7287. Powell, B.S., Court, D.L., Inada, T., Nakamura, Y., Michotey, V., Cui, X., Reizer, A., Saier Jr., M.H., Reizer, J., 1995. Novel proteins of the phosphotransferase system encoded within the rpoN operon of Escherichia coli. Enzyme IIANtr affects growth on organic nitrogen and the conditional lethality of an erats mutant. J. Biol. Chem. 270, 4822e4839. Rhodius, V.A., Mutalik, V.K., 2010. Predicting strength and function for promoters of the Escherichia coli alternative s factor, sE. Proc. Natl. Acad. Sci. U. S. A 107, 2854e2859.

Rhodius, V.A., Suh, W.C., Nonaka, G., West, J., Gross, C.A., 2006. Conserved and variable functions of the sE stress response in related genomes. PLoS. Biol. 4, e2. Rowley, G., Spector, M., Kormanec, J., Roberts, M., 2006. Pushing the envelope: extracytoplasmic stress responses in bacterial pathogens. Nat. Rev. Microbiol. 4, 383e394. Ruiz, N., Gronenberg, L.S., Kahne, D., Silhavy, T.J., 2008. Identification of two inner-membrane proteins required for the transport of lipopolysaccharide to the outer membrane of Escherichia coli. Proc. Natl. Acad. Sci. U. S. A 105, 5537e5542. Saaf, A., Baars, L., von Heijne, G., 2001. The internal repeats in the Naþ/Ca2þ exchanger-related Escherichia coli protein YrbG have opposite membrane topologies. J. Biol. Chem. 276, 18905e18907. Simons, R.W., Houman, F., Kleckner, N., 1987. Improved single and multicopy lac-based cloning vectors for protein and operon fusions. Gene 53, 85e96. Sperandeo, P., Cescutti, R., Villa, R., Di Benedetto, C., Candia, D., Deho`, G., Polissi, A., 2007. Characterization of lptA and lptB, two essential genes implicated in lipopolysaccharide transport to the outer membrane of Escherichia coli. J. Bacteriol. 189, 244e253. Sperandeo, P., Deho, G., Polissi, A., 2009. The lipopolysaccharide transport system of Gram-negative bacteria. Biochim. Biophys. Acta 1791, 594e602. Sperandeo, P., Lau, F.K., Carpentieri, A., De Castro, C., Molinaro, A., Deho`, G., Silhavy, T.J., Polissi, A., 2008. Functional analysis of the protein machinery required for transport of lipopolysaccharide to the outer membrane of Escherichia coli. J. Bacteriol. 190, 4460e4469. Sperandeo, P., Pozzi, C., Deho`, G., Polissi, A., 2006. Non-essential KDO biosynthesis and new essential cell envelope biogenesis genes in the Escherichia coli yrbG-yhbG locus. Res. Microbiol. 157, 547e558. Sukhodolets, M.V., Garges, S., 2003. Interaction of Escherichia coli RNA polymerase with the ribosomal protein S1 and the Sm-like ATPase Hfq. Biochemistry 42, 8022e8034. Tam, C., Missiakas, D., 2005. Changes in lipopolysaccharide structure induce the sE-dependent response of Escherichia coli. Mol. Microbiol. 55, 1403e1412. Udekwu, K.I., Wagner, E.G., 2007. sE controls biogenesis of the antisense RNA MicA. Nucleic Acids Res. 35, 1279e1288. Wu, T., McCandlish, A.C., Gronenberg, L.S., Chng, S.S., Silhavy, T.J., Kahne, D., 2006. Identification of a protein complex that assembles lipopolysaccharide in the outer membrane of Escherichia coli. Proc. Natl. Acad. Sci. U. S. A 103, 11754e11759. Zhou, Z., Lin, S., Cotter, R.J., Raetz, C.R., 1999. Lipid A modifications characteristic of Salmonella typhimurium are induced by NH4VO3 in Escherichia coli K12. Detection of 4-amino-4-deoxy-L-arabinose, phosphoethanolamine and palmitate. J. Biol. Chem. 274, 18503e18514.