Microbiology (2014), 160, 2650–2660

DOI 10.1099/mic.0.082263-0

LdrP, a cAMP receptor protein/FNR family transcriptional regulator, serves as a positive regulator for the light-inducible gene cluster in the megaplasmid of Thermus thermophilus Hideaki Takano,1 Yoshihiro Agari,2 Kenta Hagiwara,1 Ren Watanabe,1 Ryuta Yamazaki,1 Teruhiko Beppu,1 Akeo Shinkai2,3 and Kenji Ueda1 Correspondence Hideaki Takano [email protected]

1

Life Science Research Center, College of Bioresource Sciences, Nihon University, 1866 Kameino, Fujisawa, Kanagawa 252-0880, Japan

2

RIKEN SPring-8 Center, 1-1-1 Kouto, Sayo, Hyogo 679-5148, Japan

3

RIKEN Structural Biology Laboratory, 1-7-22 Suehiro-cho, Tsurumi-ku, Yokohama, Kanagawa 230-0045, Japan

Received 31 July 2014 Accepted 3 October 2014

LdrP (TT_P0055) (LitR-dependent regulatory protein) is one of the four cAMP receptor protein (CRP)/FNR family transcriptional regulators retained by the extremely thermophilic bacterium Thermus thermophilus. Previously, we reported that LdrP served as a positive regulator for the light-induced transcription of crtB, a carotenoid biosynthesis gene encoded on the megaplasmid of this organism. Here, we showed that LdrP also functions as an activator of the expression of genes clustered around the crtB gene under the control of LitR, an adenosyl B12-bound lightsensitive regulator. Transcriptome analysis revealed the existence of 19 LitR-dependent genes on the megaplasmid. S1 nuclease protection assay confirmed that the promoters preceding TT_P0044 (P44), TT_P0049 (P49) and TT_P0070 (P70) were activated upon illumination in the WT strain. An ldrP mutant lost the ability to activate P44, P49 and P70, whilst disruption of litR resulted in constitutive transcription from these promoters irrespective of illumination, indicating that these genes were photo-dependently regulated by LdrP and LitR. An in vitro transcription experiment demonstrated that LdrP directly activated mRNA synthesis from P44 and P70 by the Thermus RNA polymerase holocomplex. The present evidence indicated that LdrP was the positive regulator essential for the transcription of the T. thermophilus light-inducible cluster encoded on the megaplasmid.

INTRODUCTION LitR (light-induced transcription regulator) is a MerR family transcriptional regulatory protein that contains an N-terminal helix–turn–helix domain and a C-terminal cobalaminbinding domain. Originally, we discovered that Streptomyces coelicolor A3(2), a Gram-positive soil bacterium renowned for its ability to produce a variety of secondary metabolites, produced carotenoid (Crt) in response to illumination. Our study revealed that LitR was involved in the light-induced transcription of an extracytoplasmic function sigma factor litS Abbreviations: CDD, Conserved Domain Database; CRP, cAMP receptor protein; GST, glutathione S-transferase; OHB12, hydroxocobalamin. Original microarray data have been deposited in the NCBI Gene Expression Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/), accession number GSE31463. Two supplementary tables are available with the online Supplementary Material.

2650

that directed the transcription of crt operons (Takano et al., 2005). Interestingly, LitR homologues are widely distributed in phylogenetically divergent genera of non-phototrophic bacteria (Takano et al., 2005, 2006a, b, 2011, Ortiz-Guerrero et al., 2011), frequently flanked by genes for Crt biosynthesis and DNA photolyase (Phr). Previously, we revealed that a LitR homologue played a role as a central regulator of light-inducible Crt production in Thermus thermophilus HB27 (Takano et al., 2011). In this organism, the light-dependent transcriptional regulation of crtB (the phytoene synthase gene) is regulated primarily by two transcriptional regulators: LitR and TT_P0055. TT_P0055 encoded downstream of the litR gene [hereafter designated LdrP (LitR-dependent regulatory protein)] belongs to the cAMP receptor protein (CRP)/FNR family. Genetic and biochemical evidence indicated that LitR was bound by cobalamin and associated with the intergenic promoter region between litR and crtB, repressing the bidirectional 082263 G 2014 The Authors

Printed in Great Britain

Light-inducible gene cluster in T. thermophilus

transcription of litR and crtB. It is probable that the inactivation of LitR caused by a photo-dependent mechanism induces the expression of the LdrP protein, which serves as a transcriptional activator for the crtB operon, resulting in expression of Crt biosynthesis and DNA repair system genes under light conditions (Takano et al., 2011). Pe´rez-Marı´n et al. (2008) showed that CarH of Myxococcus xanthus, a homologue of LitR, was involved in light-inducible Crt production. Ortiz-Guerrero et al. (2011) reported that a chimera protein, CTt2, which is composed of the N-terminal DNA-binding domain of CarH and the C-terminal cobalaminbinding domain of LitR of T. thermophilus HB8, was a coenzyme B12 (59-deoxyadenosylcobalamin, AdoB12)-binding transcriptional regulator with light-sensitive DNA-binding activity. The photolysis of the AdoB12-bound CTt2 leads to the conversion of AdoB12 into hydroxocobalamin (OHB12), resulting in the loss of its DNA-binding activity. It was also reported that the full-length native protein TtCarH (a LitR homologue with a single-residue substitution) from T. thermophilus HB8 bound to its operator DNA in an AdoB12and light-dependent manner (Ortiz-Guerrero et al., 2011, Dı´ez et al., 2013). We previously revealed that litR and crtB were under the regulation of LitR in T. thermophilus (Takano et al., 2011); however, it was not clear whether the transcription of the crt gene cluster as a whole and of other genes were regulated in a LitR-dependent manner. We expected that the identification of LitR-dependent genes would provide a new insight into the protection from light stress in this bacterium. In order to identify the LitR-dependent genes and elucidate the regulatory mechanism, we performed a transcriptome analysis and examined the regulatory mechanism. The evidence obtained in this study indicated that 19 genes encoded on the megaplasmid, including those in the crt gene cluster and its flanking genes, were under the regulation of LitR. In addition, LdrP, which is photo-dependently controlled by LitR, served as a transcriptional activator of the operons of these genes.

METHODS Bacterial strains, plasmids and culture media. The WT strain of

T. thermophilus used in this study was HB8 (Oshima, 1974), together with HB27 TH104 (proC) (pTT8) (Hoshino et al., 1994). Escherichia coli JM109 and BL21(DE3)pLysS (Takara-Shuzo) were used as hosts for DNA manipulation and protein expression, respectively. pUC19 (Takara-Shuzo) was used for general DNA manipulation. pT7Blue and pMD19 (Takara-Shuzo) were used for TA cloning of PCR-generated DNA fragments. pGEX-6P-2 (GE Healthcare Bio-Sciences) was used for the overexpression of T. thermophilus HB27 LitR and LdrP proteins in BL21(DE3)pLysS. Enzymes used for DNA manipulation were purchased from Takara-Shuzo. The conditions for the culture and genetic manipulation for E. coli and Thermus spp. were as described by Maniatis et al. (1982) and Koyama et al. (1986), respectively. T. thermophilus was grown at 60 uC in TM medium [containing (1–1): yeast extract (Difco), 2 g; polypeptone, 4 g; NaCl, 1 g; and Castenholz basal salt solution, 10 ml]; all chemicals therein were purchased from Wako Pure Chemicals unless otherwise indicated. Castenholz basal salt solution [containing (1–1): nitrilotriacetic acid, 1 g; CaSO4.2H2O, 0.6 g; MgSO4.7H2O, 1 g; NaCl, 0.08 g; KNO3, 1.03 g; NaNO3, 6.89 g; http://mic.sgmjournals.org

Na2HPO4, 1.11 g; aqueous 0.03 % FeCl3 solution, 10 ml; and Nitsch’s trace elements, 10 ml (pH 8.2)] and Nitsch’s trace elements [containing (1–1): H2SO4, 5 ml; MgSO4.5H2O, 2.2 g; ZnSO4.7H2O, 0.5 g; H3BO3, 0.5 g; CuSO4, 0.016 g; Na2MoO4.2H2O, 0.025 g; and CoCl2 6H2O, 0.046 g] were used to prepare TM broth. E. coli was grown in Luria– Bertani (LB) medium (Maniatis et al., 1982) and 1.0–1.5 % agar (Kokusan) was added to prepare solid media. To enable the selection of transformants of E. coli and T. thermophilus, ampicillin, kanamycin and hygromycin B were added at 50 mg ml21. Gene disruption. Kanamycin-resistant mutants of T. thermophilus HB8 were generated by the standard homologous recombination technique, using disruption plasmids. Each disruption plasmid contained a promoterless thermostable kanamycin resistance (htk) gene cassette to avoid polar effects (Hoseki et al., 1999). To construct the disruption plasmid for litR (TTHB100) of T. thermophilus HB8, two flanking fragments were amplified by PCR using the primer sets P01/P02 and P03/P04 (oligonucleotide primers used in this study are shown in Table S1, available in the online Supplementary Material) and cloned onto pUC19 by three-fragment ligation. Each resulting plasmid was digested with BamHI and ligated with a promoterless htk cassette amplified by PCR using the primer sets P05/P06 to generate the disruption plasmid. Disruption plasmids were linearized by digestion with DraI and introduced into T. thermophilus WT cells (Hashimoto et al., 2001). Subsequently, kanamycin-resistant mutants were screened and true recombination was verified by PCR using the appropriate primers. DNA microarray. The T. thermophilus HB8 strain and the litR mutant was cultured at 60 uC for 8 h in TM broth under dark conditions. The crude RNA was extracted from each cell and after the cDNA was synthesized, it was fragmented and labelled with biotin-dideoxy UTP, as described previously (Agari et al., 2008). The 39-terminally labelled cDNA was hybridized to a TTHB8401a520105F GeneChip (Affymetrix), and the array was then washed, stained and scanned as described previously (Agari et al., 2008). The raw intensities for three independent cultures of the WT strain and the litR mutant were each summarized to 2266 ORFs, using the GeneChip Operating Software, version 1.2 (Affymetrix). The datasets were then normalized through the following normalization steps using the Subio Platform, including shifting of low signals ,1.0 to 1.0, log-based transformation of the data and global normalization [normalized to 75th percentile (third quartile)]. We excluded several genes with detection levels labelled ‘Absent’ (Pepper et al., 2007) from the data for three independent cultures of the WT strain and the litR mutant. The remaining data of 2061 ORFs were used for the following analysis. The t-test P values of the observed differences in the normalized intensities between the WT and litR mutants were calculated using the Subio Platform and then their false-discovery rates (q value) (Storey & Tibshirani, 2003) were calculated from these values using R software (http://www.R-project.org).

The microarray data discussed in this study are accessible through GEO accession number GSE31463 (http://www.ncbi.nlm.nih.gov/geo/ query/acc.cgi?token=phupzkkmekwyipy&acc=GSE31463). S1 nuclease mapping. S1 nuclease mapping was performed using a

previously described method (Takano et al., 2011). The transcriptional activities of the promoters preceding litR (PlitR), crtB (PcrtB), TT_P0044 (P44), TT_P0049 (P49) and TT_P0070 (P70) of T. thermophilus HB27 were evaluated by S1 protection analysis. Hybridization probes were first generated by PCR using the primers P07/P08 (P44), P09/P10 (P49) and P11/P12 (P70) (see Table S1), and cloned onto pT7Blue by TA cloning. For the low- and high-resolution analysis of P44, P49 and P70, probes were prepared by PCR using the primer sets M13-RV/P08* (P44), M13RV/P10* (P49) and M13-M4/P12* (P70) (primers labelled at their 59 end with [c-32P]ATP using T4 polynucleotide kinase are denoted with asterisks), respectively. Probes for litR and crtB were prepared as described previously (Takano et al., 2011). The primers M13-M4 and 2651

H. Takano and others M13-RV (M13 sequencing primers) were purchased from Takara-Shuzo. Radioactivity was detected by exposing dried gels to a Fuji imaging plate (Fuji Film) and images were scanned with a Typhoon 9410 image analyser (GE Healthcare). Marker 10 (pBR322/MspI digest; Nippon Gene) labelled with [c-32P]ATP using T4 polynucleotide kinase was used as a standard to estimate the transcript sizes in the low-resolution assay. To determine the transcription start sites in the high-resolution analysis, Maxam–Gilbert sequencing ladders (A+G and C+T reactions) derived from the 32P-labelled probe DNA were used as the reference. The quality of the RNA was verified via a control assay for sigA, encoding the major sigma factor (Nishiyama et al., 1999). The probe for sigA was amplified by PCR using the primers P13/P14* (see Table S1). In vitro run-off transcription. The in vitro run-off transcrip-

tional assay and the preparation of LdrP recombinant protein of T. thermophilus HB27 were performed following methods described previously (Nishiyama et al., 1999; Shinkai et al., 2007; Takano et al., 2011). The DNA templates containing the transcriptional start sites of P44 were generated by PCR, using the primers P07/P08 for Template 2186 (276 bp), P15/P08 for Template 250 (142 bp), P16/P08 for Template 251 (143 bp), P17/P08 for Template 252 (144 bp), P18/P08 for Template 253 (145 bp), P19/P08 for Template 254 (146 bp), P20/ P08 for Template 255 (147 bp), P21/P08 for Template 256 (148 bp) and P22/P08 for Template 257 (149 bp). For P70, the following primer sets were used: P23/P12 for Template 242 (120 bp), P24/P12 for Template 247 (125 bp), P25/P12 for Template 252 (130 bp), P26/P12 for Template 257 (135 bp) and P27/P12 for Template 262 (140 bp) (see Table S1). The DNA templates (172 bp) containing mutated PcrtB were generated by PCR, using the primers P31/P30 for WT, P32/P30 for G to T substitution at position 250, P33/P30 for T to G at 249, P34/P30 for G to A at 248, P35/P30 for T to G at 247, P36/P30 for G to T at 245, P37/P30 for C to A at 244, P38/P30 for C to A at-43, P39/ P30 for G to A at 241, P40/P30 for G to T at 239, P41/P30 for C to A at 238 and P42/P30 for A to C at 235. The DNA templates (157 bp) containing mutated P44 were generated by PCR, using the primers P43/ P44 for WT, P45/P44 for A to T substitution at position 249, P46/P44 for T to G at 248, P47/P44 for G to A at 247, P48/P44 for T to G at 246, P49/P44 for G to T at 244, P50/P44 for C to A at 243, P51/P44 for C to A at 242, P52/P44 for G to A at 240, P53/P44 for G to T at 238, P54/P44 for C to A at –37 and P55/P44 for A to C at 234. A total of 0.5 pmol template DNA was mixed with 2 pmol commercial RNA polymerase holoenzyme of T. thermophilus (AR Brown), 100 nmol ribonucleotides, including [a-32P]CTP, and 0 or 10 pmol LdrP protein of T. thermophilus HB27. Transcripts were analysed by PAGE. Marker 10 (pBR322/MspI digest; Nippon Gene) labelled with [c-32P]ATP was used as a standard. Preparation of AdoB12-bound LitR. Recombinant Apo–LitR protein

with a glutathione S-transferase (GST)-tag at its N terminus was produced in E. coli and purified with a GST affinity chromatography as described previously (Takano et al., 2011). The GST-tag was removed by treatment with PreScission protease (GE Healthcare). The following experiment was performed in the dark or under dim light conditions. To prepare AdoB12-treated recombinant protein, the purified LitR was added to a 10-fold molar excess of AdoB12 (Sigma-Aldrich) and incubated for 1 h at 37 uC. The mixture was then dialysed against PBS buffer (containing 140 mM NaCl, 2.7 mM KCl, 10 mM Na2PO4 and 1.8 mM KH2PO4) to remove unbound AdoB12. Absorption spectra of the resultant LitR recombinants were recorded by using a UVmini-1240 spectrometer (Shimadzu) and NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific). The protein concentration was measured with a Protein Assay kit (Bio-Rad). Gel-shift assay. DNA-binding determinations by gel-shift assay

followed the method described previously (Takano et al., 2011). Probe DNA fragments were generated by PCR using primer sets P28/P29 (PlitR), P07/P08 (P44), P09/P10 (P49), P11/P12 (P70) and P13/P14 2652

(PsigA) (see Table S1), and labelled at the 59 end with [c-32P]ATP, using T4 polynucleotide kinase. A total of 0.5–5.0 ng 32P-labelled probe DNA (10 000–20 000 c.p.m.) was mixed with 30–120 pmol recombinant LitR prepared as above. The mixture was incubated at 55 uC for 30 min in 50 ml binding buffer [containing 10 mM Tris/HCl (pH 7.0), 50 mM KCl, 1 mM EDTA, 1 mM dithiothreitol, 10 % (v/v) glycerol, 1 mg poly(dI–dC) and 50 mg BSA ml21] under light or dark conditions in a darkroom equipped with white light fluorescent lamps (Toshiba; 15 W). In the light condition, samples were illuminated at ~2.4 mmol s21 m22 for 10 min at room temperature. Specific DNA–protein complexes were separated from free probe on non-denaturing polyacrylamide gel containing 6 % acrylamide. The gels were dried and radioactive signals were detected by exposing dried gels to a Fuji imaging plate (Fuji Film). Images were scanned with a Typhoon 9410 image analyser (GE Healthcare).

RESULTS Identification of LitR-dependent genes in T. thermophilus HB8 This study was performed using two strains, T. thermophilus HB8 and HB27, whose complete genome sequences have been determined (Bru¨ggemann & Chen, 2006; Henne et al., 2004). Each genome consists of a 1.85 Mb chromosome and a large (0.26 Mb) plasmid, pTT27 (Takayama et al., 2004). HB8, but not HB27, harbours two (9.32 and 81 kb) plasmids (Ohtani et al., 2012). The genome and megaplasmid sequences of HB8 and HB27 exhibit marked similarity to each other. Despite the availability of their genomic information, a protocol and DNA chip for the DNA microarray analysis had been established only for T. thermophilus HB8 (Agari et al., 2008) and not for the HB27 strain when we started this study. For this reason, we decided to use T. thermophilus HB8 for the identification of the LitRdependent genes, even though we used the HB27 strain for our previous analysis of LitR and LdrP (Takano et al., 2011). We used T. thermophilus HB27 for further in vivo and in vitro analysis of LdrP- and LitR-mediated transcriptional regulation. We first carried out a transcriptome analysis using a T. thermophilus HB8 GeneChip custom array to identify genes regulated by LitR. Of the 2266 genes analysed, we selected the down- and upregulated genes with q values ,0.05 in the litR mutant compared with the WT under dark conditions. The analysis identified the LitR-dependent genes characterized by 2.0- to 27.5-fold higher transcriptional activity in the litR mutant than in the WT strain under dark conditions – many of these genes were located in the neighbourhood of litR (Table 1) (the other LitR-dependent genes are shown in Table S2). This result suggested that the expression of these genes was negatively regulated by LitR. As shown in Fig. 1, the LitR-dependent genes clustered around ldrP form five putative operons: TTHB089–090–091, TTHB094–095–096, litR operon (TTHB097–098–ldrP–litR), crt biosynthesis gene operon (crtB–phr–TTHB103–104–109– 110) and TTHB112–113 (see also Table 1). TTHB089–090– 091 had an operon structure and encoded a hypothetical Microbiology 160

http://mic.sgmjournals.org

Table 1. LitR-dependent genes located in the flanking region of the litR operon of T. thermophilus Coding sequence ID HB8

Expression DlitR/WT dark (q value)*

Domain name

E value

Annotation for productD

Possible cellular role

Zinc-dependent hydrolase Hypothetical protein Methyltransferase domain a/b Hydrolase fold Pyridine nucleotide disulphide oxidoreductase Putative short-chain oxidoreductase Flavin reductase-like domain YceI-like domain Hypothetical protein NAD-dependent epimerase/ dehydratase family CRP/FNR family regulator (LdrP)

Unknown Unknown Unknown Unknown Unknown

Phytoene synthase (CrtB) DNA photolyase (Phr)

Crt biosynthesis DNA repair

Cytochrome P450 (Cyp175a1) Phytoene synthase-related protein Phytoene dehydrogenase Isopentenyl-diphosphate d-isomerase (IDI-2) Conserved hypothetical protein

Crt Crt Crt Crt

Hypothetical protein

Unknown

HB27 TT_P0043 TT_P0044 TT_P0045 TT_P0046 TT_P0047

2.7 23.1 27.5 6.1 2.3

(0.009) (0.009) (0.009) (0.009) (0.024)

cl00446 – cl12011 Abhydrolase_6 cl14644

1.67e–14 – 6.63e–11 3.3e–05 6.18e–13

TTHB094 TTHB095 TTHB096 TTHB097 TTHB098

TT_P0049 TT_P0050 TT_P0051 TT_P0052 TT_P0054

5.4 5.0 8.2 4.2 16.5

(0.009) (0.009) (0.009) (0.009) (0.009)

adh_short cl00801 cl01001 – cd05271

3.2e–21 7.88e–14 2.19e–10 – 2.46e–45

TTHB099

TT_P0055

15.5 (0.009)

TTHB101 TTHB102

TT_P0057 TT_P0058

4.7 (0.010) 9.7 (0.009)

TTHB103 TTHB104 TTHB109 TTHB110

TT_P0059 TT_P0060 TT_P0066 TT_P0067

2.8 6.2 5.7 5.6

cl00047 cl00088 cd00683 cl04074 cl02999 cl12078 cl11889 cl14644 cd02811

1.29e–06 5.32e–05 4.03e–64 3.12e–67 3.52e–25 2.22e–26 4.75e–09 2.52e–139 5.24e–82

TTHB112

TT_P0069

5.4 (0.009)

TTHB113

TT_P0070

6.7 (0.009)

cl00733 cl01284 –

2.37e–29 1.38e–22 –

(0.009) (0.009) (0.009) (0.009)

Unknown Unknown Unknown Unknown Unknown Activator for light-inducible genes

biosynthesis biosynthesis biosynthesis biosynthesis

Unknown

*Intensity of the DlitR strain normalized to that of the WT strain cultured under dark conditions. DA product that was not found in the Conserved Domain Database (CDD) search was annotated as a ‘Hypothetical protein’. The possible cellular role of the gene product is shown together with the name of the most closely related domain, E value and annotation of the gene product, which were obtained from CDD or Pfam searches.

2653

Light-inducible gene cluster in T. thermophilus

TTHB088 TTHB089 TTHB090 TTHB091 TTHB092

H. Takano and others

ldrP 89–91

Thermus thermophilus HB8 (TTHB088→114)

litR 94–96

44–46

ldrP

P49

crt biosynthesis 112–113

49–51

69–70

Thermus thermophilus HB27 (TT_P0043→0071) P44

phr

PlitR

PcrtB

P70

Meiothermus silvanus (Mesil_1472→1496) Mesil_1482 Meiothermus ruber (Mrub_368→389) Mrub_0374 Deinococcus maricopensis (Deima_2490→2468) Deima_2478 Truepera radiovictrix (Trad_2056→2670) Trad_2663

Fig. 1. Schematic representation of the light-induced gene cluster in T. thermophilus and related-bacteria. The coding sequence numbers from the genome sequence database are shown. The light-inducible genes cluster in T. thermophilus HB8 and HB27 were included in the large plasmid pTT27, in a region known to contain genes involved in DNA damage repair (Bru¨ggemann & Chen, 2006). The numbers used here prefixed by TTHB for the HB8 strain and TT_P for the HB27 strain. The transcription start site, as determined by high-resolution analysis, is indicated by the bent arrow. The genes flanking ldrP (TTHB099/TT_P0055) and litR (TTHB100/TT_P0056) encode methyltransferase-like protein (TTHB090/TT_P0045), a/b hydrolase-like protein (TTHB091/TT_P0046), putative short-chain oxidoreductase (TTHB094/TT_P0049), putative flavincontaining oxidoreductase (TTHB095/TT_P0050) and YceI-like protein (TTHB096/TT_P0051). TTHB089/TT_P0044, TTHB112/TT_P0069 and TTHB113/TT_P0070 encode hypothetical proteins. pTT27 also carries a cobalamin biosynthesis gene cluster that encompasses the 20 kb region corresponding to TTHB043–066/TT_P0001–0023.

protein (TTHB089), a methyltransferase (TTHB090) and the a/b-hydrolase fold protein (TTHB091), respectively. The litR operon was flanked by the TTHB094–095–096 operon, which contained the coding sequences for short-chain dehydrogenases/reductases (TTHB094), flavin-containing oxidoreductase (TTHB095) and YceI-like protein from E. coli (TTHB096). TTHB112–113 was located downstream of the crt biosynthesis gene cluster as a divergent operon and encoded proteins with no homology to proteins of known function.

Transcriptional analysis by S1 nuclease mapping in T. thermophilus HB27

TT_P0044, TT_P0049 and TT_P0070 of the HB27 strain, which are orthologous to the first genes of the operons identified by transcriptome analysis in the HB8 strain (Fig. 1, Table 1). As shown in Fig. 2(a), the transcriptional start sites for TT_P0044 and TT_P0070 were 15 and 2 bp upstream from the ATG and GTG translational initiation codons of TT_P0044 and TT_P0070, respectively. The transcription of TT_P0049 was started from the A of ATG of the translation initiation codon of TT_P0049. Based on the position of the transcriptional start sites, the translation initiation codons of TT_P0049 and TT_P0070 were located 207 and 9 bp, respectively, downstream from the respective positions assigned in the genome sequence database (Fig. 2a). The evidence suggested that both TT_P0049 and TT_P0070 were translated by a leaderless mechanism (Moll et al., 2002). We previously observed a similar situation with respect to the transcriptional start site of litR and crtB in this organism (Takano et al., 2011). These promoters were designated P44 (TT_P0044), P49 (TT_P0049) and P70 (TT_P0070). The putative 235 and 210 sequences were CCCATT...TAGGCT (P44), CAAACA... TAACGT (P49) and GGTACA...TAGAAT (P70); the 210 hexamer of these promoters was similar to the consensus of sA-dependent promoters of T. thermophilus (TTGACA... TANCCT) described by Sevostyanova et al. (2007) (Fig. 3b).

We performed a high-resolution analysis of S1 nuclease mapping to determine the transcriptional start sites of

The transcription level of the light-inducible promoters P44, P49 and P70 was analysed by low-resolution S1 nuclease

We suppose that the six coding sequences in the ldrP region (TTHB093, TTHB105–106–107–108 and TTHB111) are also litR-dependent although they were not included in the list (Table 1). TTHB105–106–107–108 and TTHB111 appear to be included in crtB operon because each intergenic region is very short or even does not exist. TTHB105 and TTHB106 are hypothesized to be Crt biosynthesis enzymes. It is not clear why the analysis did not show the litRdependence of these genes, but it could be due to the low quality of the corresponding probes on the DNA chip.

2654

Microbiology 160

Light-inducible gene cluster in T. thermophilus (a)

P44

P49

P70

G/A T/C

G/A T/C

G/A T/C

fM fM

R

K

V

A

L

fM

(b)

DldrP

WT D

L

Time (h) 24 40 24 40

D

DlitR L

D

24 40 24 40 24 40 24 40

P44 P49 PlitR PcrtB P70 PsigA

mapping with total RNA prepared from cells grown in liquid culture under dark or light conditions (Fig. 2b). The photoinducible promoters PcrtB and PlitR (Takano et al., 2011) were used as positive controls. The sigA gene encoding s70 was used as a constitutively expressed internal control to measure the amount and purity of the purified RNA. In the WT strain, the activity of P44, P49 and P70 were induced by light similar to PcrtB and PlitR, whilst the activity of PsigA was not significantly affected by illumination. In the ldrP mutant, transcription of P44, P49 and P70 was abolished under both light and dark conditions, whilst PlitR activity was lightdependent at a level comparable with that of the WT strain. In the litR mutant, P44, P49 and P70 were all active irrespective of illumination. The result for the litR mutant is in good agreement with the data of DNA microarray analysis that the transcription of PcrtB, P44, P49 and P70 was upregulated by the deletion of litR under dark conditions. These results indicated that LdrP protein served as an activator essential for the transcription initiated from the promoters P44, P49 and P70, and that the light-dependent expression of LdrP due to the function of LitR was fundamental for the light-induced transcription from the three promoters. http://mic.sgmjournals.org

L

Fig. 2. S1 protection analyses of the lightinduced promoters of T. thermophilus HB27. (a) High-resolution analysis for the determination of the transcription start site. Maxam– Gilbert sequencing ladders (G+A and T+C reactions) were generated using the 32Plabelled DNA fragment derived from the probe DNA. The positions of the S1-protected fragments are shown by arrowheads and the residues indicated by the bent arrows represent the transcription start sites. It is known that the fragments generated by chemical sequencing reactions migrate 1.5 nt further than the corresponding fragments generated by S1 nuclease digestion of the DNA–RNA hybrids (half a residue from the presence of the 39terminal phosphate group and one residue from the elimination of the 39-terminal nucleotide) (Sollner-Webb & Reeder, 1979). RNA prepared from the WT cells grown for 24 h and cultured in TM liquid medium under light conditions was used for hybridization. (b) Lowresolution analysis. The activities of the lightinduced promoters were estimated by the intensity of the hybridization signal. RNA was isolated from the cells of each T. thermophilus strain, and cultured in TM liquid medium for 24 and 40 h under dark (D) and light (L) conditions. Transcription of sigA was analysed to confirm the quality and quantity of RNA.

In vitro run-off transcription To clarify whether LdrP directly activated transcription from the light-inducible promoters by RNA polymerase holoenzyme, we performed an in vitro run-off transcriptional assay as previously carried out with respect to PcrtB (Fig. 3a). An RNA polymerase holocomplex and a recombinant protein of LdrP successfully generated transcripts whose length corresponded to those of P44 (92 bases) and P70 (73 and 74 bases), respectively. The transcription from P70 occurred at a low level even in the absence of LdrP. This result was consistent with that from S1 nuclease mapping (Fig. 2b): the measurement of signal intensity at 40 h indicated that the fold change of transcription (light/dark) at P70 (8.6) was lower than at P44 (53.3) and P49 (12.2). The low-level transcription at P70 was also observed in the ldrP mutant irrespective of illumination (Fig. 2b). These results indicated that the RNA polymerase holoenzyme inefficiently generated the transcript from P70 without LdrP. We performed the same transcriptional assay with trimmed templates: 250 to 257 for P44 and 242 to 262 for P70 (the number indicates the position of the 59 end of the antisense 2655

H. Takano and others

Template

(a)

LdrP

–186 –

+

–57

–56

–55

–54

–53

–52

–51

–50

+

+

+

+

+

+

+

+

123 110

P44

mRNA (92 bases)

90 76 67 (nt)

–62

Template

+

LdrP

–57 –

+

–52 –

+

–47 –

+

–42 –

+

–

90 P70

76

mRNA (73, 74 bases)

67 (nt) (b) –60

–50

-40

–30

–20

+1

–10

PcrtB 5′-

-3′

P44

5 5′-

-3′

P70

5′-

-3′

P49

5′-

-3′

Consensus

5′-

-3′

Consensus

(c)

A

T

T

G

C

G

-3′

5′-

T

G

C

C

T G

G

C

A

G

Template WT A-49T T-48G G-48A T-46G G-44T C-43A C-42A G-40A G-38T C-37A A-34C LdrP – + – + – + – + – + – + – + – + – + – + – + – + 110 P44

PcrtB

90 76 Template WT G-50T T-49G G-48A T-47G G-45T C-44A C-43A G-41A G-39T C-38A A-35C LdrP – + – + – + – + – + – + – + – + – + – + – + – + 123 110 90 (nt)

mRNA (92 bases)

mRNA (117 bases)

Fig. 3. In vitro run-off transcription assay. (a) T. thermophilus RNA polymerase holoenzyme was added to the reaction mixture containing the promoter DNA fragments in the presence (+) or absence (”) of 2 pmol LdrP recombinant protein to generate transcripts specific for P44 (TT_P0044) and P70 (TT_P0070). Closed triangles denote the positions of the expected transcripts for P44 and P70. The template number indicates the position of the 59 end of each template with respect to the transcriptional start site

2656

Microbiology 160

Light-inducible gene cluster in T. thermophilus

[see (b)]. The 39 end is commonly +92 for P44 and +74 for P70. Thus, ”57 for P44 means that the template contains the region from ”57 to +92. (b) Nucleotide sequence of LdrP-dependent promoters. The promoter regions required for the transcriptional activation by LdrP in vitro run-off assay are underlined. The transcriptional start site is indicated by +1. A putative ”10 hexamer sequence of each promoter is indicated by bold italic letters. The predicted consensus sequence based on the three promoters is shown. (c) The conserved nucleotides in the predicted LdrP consensus sequence were substituted for other nucleotides, e.g. A-49T indicates the A at position ”49 was mutated to T. The position was with respect to the transcriptional start site in each promoter. T. thermophilus RNA polymerase (2 pmol) was added to the reaction mixture in the presence (+) or absence (”) of LdrP recombinant protein.

strand of the template DNA when the transcriptional start site is numbered at +1). The assay showed that LdrP-dependent transcription occurred with templates 253 to 257 for P44 and 251 and 261 for P70 (Fig. 3a). When template 242 and 247 for P70 was used, the addition of LdrP partially inhibited the transcription initiated from each template compared with the control lane. This weak inhibition by LdrP may have been due to its incomplete binding to the template. These results indicated that the LdrP-dependent transcriptional initiation at P44 and P70 required the region encompassing positions 253 to +92 and 252 to +74, respectively (Fig. 3a, b).

transcription of P49 by the ldrP mutation and that a similar sequence to the LdrP consensus sequence was found in P49 (Fig. 3b). This result suggested that transcriptional initiation at P49 required an additional transcription factor. In all promoters tested in this assay, the addition of cAMP did not affect the efficiency of the transcriptional activation by LdrP (data not shown). Fig. 3(b) shows a predicted consensus sequence, 59-(A/G) (T/C)(T/G)TNGCCN(T/G)NG(G/C)NNA-39, for the LdrP recognition site, which was proposed on the basis of the nucleotide sequence alignment of the three LdrP-dependent promoters, PcrtB, P44 and P70. The sequence for P49 was not considered for the prediction due to the aforementioned possibility that other transcription factors may be involved in its regulation. In E. coli, the CRP–cAMP complex binds to

However, RNA polymerase holoenzyme and LdrP did not generate a specific transcript on the probe DNA for the P49 region in vitro, despite the fact that the in vivo transcriptional analysis showed the abolishment of light-inducible

PlitR–PcrtB

(a) Dark 1

2

3

Light 4

5

1

2

3

4

5

Probe–protein complex Probe

(b) 1

2

P44 3 4

5

1

2

P49 3 4

5

1

2

P70 3

4

5

1

2

PsigA 3 4

5

Probe

Fig. 4. Gel-shift assay of AdoB12–LitR. The amounts of AdoB12–LitR used were 0 (lane 1), 20 (lane 2), 40 (lane 3), 80 (lane 4) and 160 pmol (lane 5). The various amounts of AdoB12–LitR were mixed with the probes for (a) PlitR, and (b) P44, P49, P70 and PsigA, and applied to a non-denaturing polyacrylamide gel. The AdoB12–LitR protein was incubated in the dark (10 min) or irradiated with white light (10 min) prior to the addition of 32P-labelled probes. http://mic.sgmjournals.org

2657

H. Takano and others

Light

AdoB12 LitR

P44 P TT_P0044

Carotenoids DNA repair –

P49 P TT_P0049

ldrP

litR

P P PlitR PcrtB

P70 crtB

TT_P0070 P

LdrP

Fig. 5. Model for the light-inducible transcriptional control of the genes clustered around the crt genes by LitR and LdrP. LitR is bound by AdoB12, and associates with the intergenic region of PcrtB and PlitR, repressing the transcription of both genes under dark conditions. Under light conditions, the adsorption of light by the LitR–AdoB12 complex may cause its inactivation followed by the conversion of AdoB12 to OHB12 by photolysis (Ortiz-Guerrero et al., 2011). The depression of LitR induces the expression of LdrP, which in turn activates PcrtB, P44, P70 and probably P49. PcrtB directs the transcription of genes that encode Crt biosynthesis enzymes and DNA photolyase. The transcription initiated from P44, P70 and P49 directs the transcription of putative operons of TT_P0044–0045–0046, TT_P0049–0050–0051 and TT_P0069–0070, respectively (see Table 1 for the annotated gene products).

the consensus sequence 59-AAATGTGATCTAGATCACATTT-39 (Ebright et al., 1989). The CRP-binding sites of E. coli are grouped into two classes based on the position of the CRP-binding site: class I and class II are centred at positions 261.5 and 241.5, respectively, with respect to the transcriptional start site (Busby & Ebright, 1999; Lawson et al., 2004). The predicted LdrP consensus sequence did not exhibit a distinctive similarity to that of E. coli CRP, although the position of the LdrP consensus [–35 to 250 (PcrtB), 234 to 249 (P44) and –37 to 252 (P70)] was similar to that of E. coli CRP class II. We could not confirm that the proposed consensus was the LdrP-binding sequence because direct binding of LdrP to P44, P70 and P49 was not observed (see Discussion). To examine whether the predicted consensus was involved in LdrP transcriptional activation, we performed an in vitro run-off assay using the mutated PcrtB and P44 promoters as templates, which contained mutations at conserved nucleotides around the 240 regions. As shown in Fig. 3(c), similar nucleotide mutations caused low-level LdrP-mediated transcriptional activation of both promoters; mutation of the underlined nucleotides in 59-(A/G)(T/C)(T/ G)TNGCCN(T/G)NG(G/C)NNA-39 reduced mRNA synthesis from PcrtB and P44 promoters. These results suggested that LdrP recognized a part of the predicted LdrP consensus sequence to regulate the targeted promoters. In vitro DNA-binding activity of AdoB12-bound LitR (AdoB12–LitR) To examine whether the transcription of P44, P49 and P70 was regulated only by LdrP or LitR was also involved directly in the regulation, we prepared AdoB12–LitR of T. thermophilus HB27 and performed a gel-shift assay. As shown in Fig. 2658

4(a), AdoB12–LitR showed specific binding to the intergenic promoter region of litR and crtB under dark conditions, whilst AdoB12–LitR exposed to white light for 10 min did not bind to the promoter region. In contrast, AdoB12–LitR did not bind to the promoter regions of P44, P49, P70 and PsigA (control) under dark conditions (Fig. 4b). These results indicated that AdoB12–LitR of T. thermophilus HB27 was also a light-sensitive DNA-binding protein, as reported in TtCarH of HB8 (Ortiz-Guerrero et al., 2011), and suggested that only LdrP affected the transcription of P44, P49 and P70.

DISCUSSION Our transcriptome analysis revealed the existence of 19 LitR-dependent genes encoded on the megaplasmid of T. thermophilus. A correlation between many of the newly identified genes and light has not been reported in nonphototrophic bacteria, which indicates that they are novel light-inducible genes. Our data indicate that LdrP is a positive regulator essential for transcriptional initiation of the gene cluster and that LitR confers photo-dependency to LdrP-dependent transcription. The finding that 45 out of 251 genes encoded on the megaplasmid, including cobalamin biosynthesis genes, are associated with light indicates that the megaplasmid of T. thermophilus plays an important role in the protection of the cell from light stress. Fig. 5 shows the current working model for the transcriptional control of the light-inducible gene cluster by LitR and LdrP. This model is supported by the results of our previous study (Takano et al., 2011) and the present study, and other recent findings (Ortiz-Guerrero et al., 2011). The evidence strongly reinforces our view that the members of the CarH/LitR family of proteins [we previously described the family to be CarA/ Microbiology 160

Light-inducible gene cluster in T. thermophilus

LitR (Takano et al., 2011), but changed this based on the above description on the CarH of Myxococcus xanthus] are novel photosensors distributed in divergent bacterial genera (Takano et al., 2011). In our current working model, AdoB12– LitR binds PlitR to negatively control the expression of LdrP under dark conditions. The inactivation of AdoB12–LitR by illumination induces the production of the LdrP protein, which in turn directly activates PcrtB, P44, P70 and probably P49. The activation of P49 by LdrP was not reproduced in the in vitro transcription assay, suggesting the possibility that an additional element is required for the transcriptional control at P49. The finding that LdrP did not require additional effector molecules to activate transcriptional initiation in vitro was supported by the 3D structure of TTHB099 (Agari et al., 2012), which has a single amino acid substitution, E77D, compared with LdrP. The structure of TTHB099 was close to that of T. thermophilus SdrP, a member of the CRP/FNR family, that serves as an activator in the absence of an effector molecule such as cAMP (Agari et al., 2008). As the AdoB12–LitR protein bound to PlitR but not to P44, P70 and P49 in the in vitro DNA-binding assay (Fig. 4), we suppose that LitR regulates only the bidirectional promoter region between litR and crtB amongst the promoters examined in this study. LdrP showed marked activity in the activation of transcriptional initiation from P44, P70 (this study) and PcrtB (Takano et al., 2011) in the in vitro transcription experiment (Fig. 3a). However, the LdrP protein did not stably bind to the DNA fragment containing these promoter regions in a gel-shift assay. Furthermore, we could not confirm the binding of LdrP to PcrtB in the DNase I footprinting assay and BIAcore analysis (our unpublished observation). These experiments were performed carefully and repeatedly under varying conditions, including temperature, buffer composition and both, with or without the Thermus RNA polymerase holoenzyme. Hence, we assume that LdrP is able to recruit RNA polymerase holoenzyme to promoter regions through its weak contact with target promoters. What is the role of light-inducible genes in this organism? Upon illumination, photosensitizer molecules such as flavin and tetrapyrroles, which are present in both phototrophic and non-phototrophic bacteria, generate highly reactive species such as 1O2, which damages biomolecules (Glaeser et al., 2011). Several light-inducible genes identified in this study appear to be associated with 1O2 production. Carotenoids possess the ability to quench 1O2 (Edge et al., 1997). In Rhodobacter sphaeroides, the expression of a DNA photolyase that repairs pyrimidine dimers is induced by 1O2 (Hendrischk et al., 2007), although its role in protecting against 1O2 damage remains to be resolved. Both TT_P0049 and TT_P0050 are homologous with proteins having oxidation/reduction activity (short-chain dehydrogenases/ reductases and flavin-containing oxidoreductases, respectively), and could therefore function to detoxify peroxidation products of fatty acids or replenish an unknown molecule(s) that is critical for survival in the presence of 1O2 (Hayes & McLellan, 1999). These observations suggest that many of http://mic.sgmjournals.org

the light-inducible genes found in this study play a role in protecting cells from 1O2. The transcriptome analysis also identified the genes showing a lower transcriptional level in the litR mutant than that in the WT (Table S2). The downregulated genes existed on the chromosome and included functionally unknown genes forming an operon-like structure (TTHA0771–0774). This result suggests that LitR acts as a positive regulator to control the expression of these chromosomally encoded genes. Whether LitR binds directly to the promoter region of these genes is not known, and because the difference in the expression level of these downregulated genes between the litR mutant and the WT strain is relatively small, the direct involvement of LitR in their regulation is not clear. The light-inducible gene cluster, including LitR and LdrP, is also conserved in some species of Deinococcus-Thermus extremophiles (Fig. 1). Many of these genes, including the litR–ldrP operon, are distributed in the chromosome of Meiothermus ruber (Tindall et al., 2010), Meiothermus silvanus (Sikorski et al., 2010) and Deinococcus maricopensis (http://www.genome.ad.jp) – species that were isolated either from hot springs or from soil. This gene conservation suggests that the LitR-dependent genetic regulation also occurs in these bacteria. Meanwhile, the other genomesequenced species belonging to the phylum DeinococcusThermus, including Deinococcus radiodurans, Deinococcus geothermalis, Oceanithermus profundus and Marinithermus hydrothermalis, do not retain the litR and the neighbouring region (http://www.genome.ad.jp). O. profundus and Marinithermus hydrothermalis were isolated from deep-sea hydrothermal vents, where the environment is dark. This implies that the gene cluster has been acquired due to an advantage specifically developed within some group of the bacterial phylum, which suggests that this regulatory system is more required for their survival in extreme environments. The biochemical characterization of individual proteins encoded by the light-inducible gene cluster presented here will provide the necessary information to understand the biological responses and adaptation to illumination in this group of bacteria.

ACKNOWLEDGEMENTS This study was supported by a Grant-in-Aid for the Encouragement of Young Scientists (23780093) to H. T. and Scientific Research (C) (22510208) to A. S. This study was also supported by the High-Tech Research Center Project of the Ministry of Education, Culture, Sports, Science and Technology, Japan, the Noda Institute for Scientific Research, the Foundation NAGASE Science Technology Development and Charitable Trust Araki Medical and Biochemical Research Memorial Fund.

REFERENCES Agari, Y., Kashihara, A., Yokoyama, S., Kuramitsu, S. & Shinkai, A. (2008). Global gene expression mediated by Thermus thermophilus SdrP,

a CRP/FNR family transcriptional regulator. Mol Microbiol 70, 60–75. 2659

H. Takano and others Agari, Y., Kuramitsu, S. & Shinkai, A. (2012). X-ray crystal structure

Ohtani, N., Tomita, M. & Itaya, M. (2012). The third plasmid pVV8

of TTHB099, a CRP/FNR superfamily transcriptional regulator from Thermus thermophilus HB8, reveals a DNA-binding protein with no required allosteric effector molecule. Proteins 80, 1490–1494.

from Thermus thermophilus HB8: isolation, characterization, and sequence determination. Extremophiles 16, 237–244.

Bru¨ggemann, H. & Chen, C. (2006). Comparative genomics of

Ortiz-Guerrero, J. M., Polanco, M. C., Murillo, F. J., Padmanabhan, S. & Elı´as-Arnanz, M. (2011). Light-dependent gene regulation by a

Thermus thermophilus: plasticity of the megaplasmid and its contribution to a thermophilic lifestyle. J Biotechnol 124, 654–661.

coenzyme B12-based photoreceptor. Proc Natl Acad Sci U S A 108, 7565–7570.

Busby, S. & Ebright, R. H. (1999). Transcription activation by

Oshima, T. (1974). [Comparative studies on biochemical properties of

catabolite activator protein (CAP). J Mol Biol 293, 199–213.

an extreme thermophile, Thermus thermophilus HB 8 (author’s transl)]. Seikagaku 46, 887–907.

Dı´ez, A. I., Ortiz-Guerrero, J. M., Ortega, A., Elı´as-Arnanz, M., Padmanabhan, S. & Garcı´a de la Torre, J. (2013). Analytical

ultracentrifugation studies of oligomerization and DNA-binding of TtCarH, a Thermus thermophilus coenzyme B12-based photosensory regulator. Eur Biophys J 42, 463–476. Ebright, R. H., Ebright, Y. W. & Gunasekera, A. (1989). Consensus DNA

site for the Escherichia coli catabolite gene activator protein (CAP): CAP exhibits a 450-fold higher affinity for the consensus DNA site than for the E. coli lac DNA site. Nucleic Acids Res 17, 10295–10305. Edge, R., McGarvey, D. J. & Truscott, T. G. (1997). The carotenoids as

anti-oxidants – a review. J Photochem Photobiol B 41, 189–200. Glaeser, J., Nuss, A. M., Berghoff, B. A. & Klug, G. (2011). Singlet

oxygen stress in microorganisms. Adv Microb Physiol 58, 141–173.

Pepper, S. D., Saunders, E. K., Edwards, L. E., Wilson, C. L. & Miller, C. J. (2007). The utility of MAS5 expression summary and detection

call algorithms. BMC Bioinformatics 8, 273. Pe´rez-Marı´n, M. C., Padmanabhan, S., Polanco, M. C., Murillo, F. J. & Elı´as-Arnanz, M. (2008). Vitamin B12 partners the CarH repressor to

downregulate a photoinducible promoter in Myxococcus xanthus. Mol Microbiol 67, 804–819. Sevostyanova, A., Djordjevic, M., Kuznedelov, K., Naryshkina, T., Gelfand, M. S., Severinov, K. & Minakhin, L. (2007). Temporal

regulation of viral transcription during development of Thermus thermophilus bacteriophage phiYS40. J Mol Biol 366, 420–435.

Hashimoto, Y., Yano, T., Kuramitsu, S. & Kagamiyama, H. (2001).

Shinkai, A., Kira, S., Nakagawa, N., Kashihara, A., Kuramitsu, S. & Yokoyama, S. (2007). Transcription activation mediated by a cyclic

Disruption of Thermus thermophilus genes by homologous recombination using a thermostable kanamycin-resistant marker. FEBS Lett 506, 231–234.

AMP receptor protein from Thermus thermophilus HB8. J Bacteriol 189, 3891–3901.

Hayes, J. D. & McLellan, L. I. (1999). Glutathione and glutathione-

Sikorski, J., Tindall, B. J., Lowry, S., Lucas, S., Nolan, M., Copeland, A., Glavina Del Rio, T., Tice, H., Cheng, J. F. & other authors (2010).

dependent enzymes represent a co-ordinately regulated defence against oxidative stress. Free Radic Res 31, 273–300. Hendrischk, A. K., Braatsch, S., Glaeser, J. & Klug, G. (2007). The

Complete genome sequence of Meiothermus silvanus type strain (VIR2). Stand Genomic Sci 3, 37–46.

phrA gene of Rhodobacter sphaeroides encodes a photolyase and is regulated by singlet oxygen and peroxide in a sE-dependent manner. Microbiology 153, 1842–1851.

Sollner-Webb, B. & Reeder, R. H. (1979). The nucleotide sequence of

Henne, A., Bru¨ggemann, H., Raasch, C., Wiezer, A., Hartsch, T., Liesegang, H., Johann, A., Lienard, T., Gohl, O. & other authors (2004). The genome sequence of the extreme thermophile Thermus

Storey, J. D. & Tibshirani, R. (2003). Statistical significance for

thermophilus. Nat Biotechnol 22, 547–553. Hoseki, J., Yano, T., Koyama, Y., Kuramitsu, S. & Kagamiyama, H. (1999). Directed evolution of thermostable kanamycin-resistance

gene: a convenient selection marker for Thermus thermophilus. J Biochem 126, 951–956. Hoshino, T., Kosuge, T., Hidaka, Y., Tabata, K. & Nakahara, T. (1994).

Molecular cloning and sequence analysis of the proC gene encoding delta 1-pyrroline-5-carboxylate reductase from an extremely thermophilic eubacterium Thermus thermophilus. Biochem Biophys Res Commun 199, 410–417. Koyama, Y., Hoshino, T., Tomizuka, N. & Furukawa, K. (1986).

Genetic transformation of the extreme thermophile Thermus thermophilus and of other Thermus spp. J Bacteriol 166, 338–340. Lawson, C. L., Swigon, D., Murakami, K. S., Darst, S. A., Berman, H. M. & Ebright, R. H. (2004). Catabolite activator protein: DNA

binding and transcription activation. Curr Opin Struct Biol 14, 10–20. Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982). Molecular Cloning: A

Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory. Moll, I., Grill, S., Gualerzi, C. O. & Bla¨si, U. (2002). Leaderless mRNAs

in bacteria: surprises in ribosomal recruitment and translational control. Mol Microbiol 43, 239–246. Nishiyama, M., Kobashi, N., Tanaka, K., Takahashi, H. & Tanokura, M. (1999). Cloning and characterization in Escherichia coli of the gene

encoding the principal sigma factor of an extreme thermophile, Thermus thermophilus. FEMS Microbiol Lett 172, 179–186.

2660

the initiation and termination sites for ribosomal RNA transcription in X. laevis. Cell 18, 485–499. genomewide studies. Proc Natl Acad Sci U S A 100, 9440–9445. Takano, H., Obitsu, S., Beppu, T. & Ueda, K. (2005). Light-induced

carotenogenesis in Streptomyces coelicolor A3(2): identification of an extracytoplasmic function sigma factor that directs photodependent transcription of the carotenoid biosynthesis gene cluster. J Bacteriol 187, 1825–1832. Takano, H., Asker, D., Beppu, T. & Ueda, K. (2006a). Genetic control for

light-induced carotenoid production in non-phototrophic bacteria. J Ind Microbiol Biotechnol 33, 88–93. Takano, H., Beppu, T. & Ueda, K. (2006b). The CarA/LitR-family

transcriptional regulator: its possible role as a photosensor and wide distribution in non-phototrophic bacteria. Biosci Biotechnol Biochem 70, 2320–2324. Takano, H., Kondo, M., Usui, N., Usui, T., Ohzeki, H., Yamazaki, R., Washioka, M., Nakamura, A., Hoshino, T. & other authors (2011).

Involvement of CarA/LitR and CRP/FNR family transcriptional regulators in light-induced carotenoid production in Thermus thermophilus. J Bacteriol 193, 2451–2459. Takayama, G., Kosuge, T., Maseda, H., Nakamura, A. & Hoshino, T. (2004). Nucleotide sequence of the cryptic plasmid pTT8 from

Thermus thermophilus HB8 and isolation and characterization of its high-copy-number mutant. Plasmid 51, 227–237. Tindall, B. J., Sikorski, J., Lucas, S., Goltsman, E., Copeland, A., Glavina Del Rio, T., Nolan, M., Tice, H., Cheng, J. F. & other authors (2010). Complete genome sequence of Meiothermus ruber type strain

(21). Stand Genomic Sci 3, 26–36. Edited by: Y. Ohnishi

Microbiology 160