organization and transcription of the myo-inositol operon ... · inositol as the sole carbon...

JOURNAL OF BACTERIOLOGY,0021-9193/97/$04.0010

July 1997, p. 4591–4598 Vol. 179, No. 14

Copyright © 1997, American Society for Microbiology

Organization and Transcription of the myo-Inositol Operon, iol,of Bacillus subtilis

KEN-ICHI YOSHIDA, DAIKI AOYAMA, IZUMI ISHIO, TSUKASA SHIBAYAMA,AND YASUTARO FUJITA*

Department of Biotechnology, Faculty of Engineering, Fukuyama University, Fukuyama 729-02, Japan

Received 6 February 1997/Accepted 5 May 1997

Previous determination of the nucleotide sequence of the iol region of the Bacillus subtilis genome allowed usto predict the structure of the iol operon for myo-inositol catabolism, consisting of 10 iol genes (iolA to iolJ); iolGcorresponds to idh, encoding myo-inositol 2-dehydrogenase (Idh). Primer extension analysis suggested that aninositol-inducible promoter for the iol operon (iol promoter) might be a promoter-like sequence in the 5* regionof iolA, which is probably recognized by sA. S1 nuclease analysis implied that a r-independent terminator-likestructure in the 3* region of iolJ might be a terminator for iol transcription. Disruption of the iol promoterprevented synthesis of the iol transcript as well as that of Idh, implying that the iol operon is most probablytranscribed as an 11.5-kb mRNA containing the 10 iol genes. Immediately upstream of the iol operon, two genes(iolR and iolS) with divergent orientations to the iol operon were found. Disruption of iolR (but not iolS) causedconstitutive synthesis of the iol transcript and Idh, indicating that the iolR gene encodes a transcription-negative regulator (presumably a repressor) for the iol operon. Northern and S1 nuclease analyses revealedthat the iolRS genes were cotranscribed from another inositol-inducible promoter, which is probably recognizedby sA. The promoter assignments of the iol and iolRS operons were confirmed in vivo with a lacZ fusionintegrated into the amyE locus.

myo-Inositol is abundant in nature, especially in soil. Vari-ous microorganisms including soil bacteria are able to grow oninositol as the sole carbon source. The pathway of inositolcatabolism in Klebsiella aerogenes has been extensively studied(1, 2, 4, 5). myo-Inositol 2-dehydrogenase (Idh; EC 1.1.1.18) isthe first enzyme responsible for inositol catabolism in thisorganism (4). Other microorganisms, such as Bacillus subtilis(17), Rhizobium leguminosarum bv. viciae (16), and Cryptococ-cus melibiosum (20), probably possess similar pathways involv-ing Idh. These findings suggested that inositol catabolismmight be conserved among these microorganisms. However,the molecular genetics of the genes involved in inositol catab-olism have not been well studied.

B. subtilis Idh has been purified and characterized (17). Idhsynthesis is induced upon the addition of inositol and is alsounder catabolite repression (13, 14, 17). We succeeded in thecloning and expression of the idh gene of B. subtilis in Esche-richia coli (11). The regions surrounding the idh gene werecloned by chromosome walking (23, 24). The nucleotide se-quences of these regions implied a putative operon structureconsisting of 10 genes, which are probably involved in inositolcatabolism (iol) (23–25).

As a first step toward elucidating the pathway and regulationof inositol catabolism of B. subtilis, we analyzed the iol tran-script. We present here the results regarding the organizationand transcription of the iol operon. Moreover, we identified atranscription-negative regulator gene (iolR) for this operon.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth media. The B. subtilis strains used arelisted in Table 1; all are derivatives of our standard strain, 60015, constructed asdescribed below. E. coli JM109 {recA1 supE44 endA1 hsdR17 gyrA96 relA1 thi

D(lac-proAB) F9[traD36 proAB1 lacIq lacZDM15]} (21) was used as a cloninghost for plasmid pUC19 (21) derivatives. Plasmid pBEST-4F, which carries achloramphenicol acetyltransferase gene (cat) cassette containing a transcrip-tional terminator for cat, was provided by M. Itaya (Mitsubishi Kasei Life Sci-ence Institute, Tokyo, Japan). Strain JM109 and plasmid pUC19 were obtainedfrom Takara Shuzo Co., Ltd. B. subtilis cells were grown on tryptose blood agarbase (Difco) supplemented with 0.18% glucose (TBABG plates) or in S6 me-dium (7) containing 0.5% Casamino Acids (Difco) as a carbon source andsupplemented with 50 mg each of tryptophan and methionine per ml. StrainJM109 was grown on Luria-Bertani agar plates or in 23 YT medium (18).

Gene and promoter disruption by insertion of a cat cassette. The iolR gene wasdisrupted by insertion of the cat cassette derived from plasmid pBEST-4F asfollows. A 1.65-kb EcoRI fragment containing the iolR gene, which was derivedfrom the insert of l DASH II recombinant phage EM11 (23), was cloned into theEcoRI site of plasmid pUC19 to produce plasmid pIOLR1. Plasmid pIOLR1was cleaved at an EcoRV site within the iolR gene and then ligated with thecat cassette trimmed with SmaI, resulting in plasmid pIOLR1::cat. PlasmidpIOLR1::cat was linearized with EcoRI and then mixed with competent cells ofstrain 60015 (8). Chloramphenicol-resistant transformants, resulting from a dou-ble crossover, were selected on TBABG plates containing 15 mg of chloram-phenicol per ml, and one of the transformants was confirmed to carry the correctinsertion of the cat cassette in iolR and was designated strain YF244.

The iolS gene was disrupted by insertion of the cat cassette as follows. A1.47-kb EcoRV fragment containing part of the iolS gene, derived from the insertof l DASH II recombinant phage EM11 (23), was cloned into the blunt-endedEcoRI site of plasmid pUC19 to produce plasmid pIOLS1. Plasmid pIOLS1 wascleaved at an EcoRI site within the iolS gene and then ligated with the cat cassettecleaved with EcoRI, resulting in plasmid pIOLS1::cat. Plasmid pIOLS1::cat waslinearized with KpnI and then used for the transformation of strain 60015. One of thechloramphenicol-resistant transformants was designated strain YF246 after the cor-rect construction was confirmed.

The iol promoter (Piol) was disrupted by insertion of the cat cassette into its210 region as follows. A 1.5-kb region containing Piol was amplified with thegeneration of a new BglII site in its 210 region and two flanking EcoRI sites bymeans of recombinant PCR (12) with chromosome DNA of strain 60015 as atemplate and four oligonucleotides as primers: upstream and downstreamprimers (59-GGAATTCACGTTAGGCCAGATGAATGC-39 and59-AAATAGAATTCTGGCGCGGC-39 [EcoRI sites are underlined]) and two overlap-ping primers for the introduction of the BglII site into the 210 region of Piol(59-CTTATGGGTATTATGCGATTAGATCTTAACCAAGAAATG-39 and59-GGTCATTTCTTGGTTAAGATCTAATCGCATAATACCCAT-39 [theBglIIsite is underlined]). The amplified 1.5-kb fragment was cleaved with EcoRI and thencloned into the EcoRI site of plasmid pUC19, resulting in plasmid pPIOL1. Thecat cassette from plasmid pBEST-4F was cleaved with BamHI and inserted intothe BglII site of plasmid pPIOL1. The resulting plasmid, pPIOL1::cat, was lin-earized with EcoRI and then used for the transformation of strain 60015. One of

* Corresponding author. Mailing address: Department of Biotech-nology, Faculty of Engineering, 985 Sanzo, Higashimura-cho,Fukuyama-shi, Hiroshima 729-02, Japan. Phone: (81) 849 36 2111. Fax:(81) 849 36 2023.

4591

on February 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

the chloramphenicol-resistant transformants was designated strain YF248 afterconfirmation that the construction was correct.

Integration of the iolA*-*lacZ and iolR*-*lacZ fusions under the control of therespective iol and iolRS promoters into the chromosomal amyE locus. To con-struct strains YF323, YF324 and YF325, carrying a series of deletions of Piol,which directs the expression of the iolA9-9lacZ fusion integrated in amyE, threeiol sequences (nucleotides 2107 to 1245, 245 to 1245, and 223 to 1245; 11 isthe putative transcription initiation nucleotide) were amplified by PCR withchromosomal DNA of strain 60015 as a template and with primers consisting ofa pair of synthetic oligonucleotides (38-mer) designed to create EcoRI andHindIII sites flanking the upstream and downstream ends, respectively. Theupstream primers carrying nucleotides 2107, 245, and 223 and the downstreamprimer carrying nucleotide 1245 were 59-GCGGGAATTCCTTCTCTTACTTCTCTTACTTG-39, 59-GCGGGAATTCGTGTTTATGATTGACTTATGGG-39,and 59-GCGGGAATTCTATTATGCGATTAGAATATAACC-39 (the EcoRIsites are underlined) (upstream) and 59-GCGGAAGCTTGCTTTCAACCCATTCACCGTTG-39 (the HindIII site is underlined) (downstream). The PCR prod-ucts were digested with EcoRI and HindIII, and the resulting EcoRI-HindIIIfragments were cloned into the EcoRI and HindIII sites of plasmid ptrpBGI (19),respectively. The three resulting plasmids, which had been linearized with PstI,were used for the transformation of strain 60015 to produce chloramphenicol-resistant clones YF323, YF324, and YF325, respectively.

Strains YF327, YF328, YF329, and YF330, carrying another series of deletionsof the iolRS promoter (PiolRS), which directs the expression of the iolR9-9lacZfusion integrated in amyE, were constructed as described above with the PCRproducts possessing iolRS sequences (nucleotides 294 to 162, 270 to 162, 243to 162, and 219 to 162, respectively). The upstream primers carrying nucleo-tides 294, 270, 243, and 219 and the downstream primer carrying nucleotide162 were 59-GCGGGAATTCAGAAGGCCGTCCAATCTTGC-39, 59-GCGGGAATTCGATATACACCGATGTTATCATTTT-39, 59-GCGGGAATTCGCAACTATTGATTAACTTTTGG-39, and 59-GCGGGAATTCTTTATTATATATTTATGTTACG-39 (the EcoRI sites are underlined) (upstream) and 59-GCGCAAGCTTCTCCTCCATTTCCTGAATCC-39 (the HindIII site is underlined)(downstream).

RNA preparation. B. subtilis cells were grown in S6 medium containing 0.5%Casamino Acids, and with 10 mM inositol present or absent. The cells wereharvested and then successively treated with lysozyme, sodium dodecyl sulfate,and phenol to extract total RNA as described previously (9).

Primer extension. A 50-mg portion of each total RNA sample was hybridizedwith an oligonucleotide (59-TCAATATCCACTTTTCAGCAAGCC-39) that hadbeen labeled at the 59 end with T4 polynucleotide kinase (Takara Shuzo Co.,Ltd.) and [g-32P]ATP (ICN Biomedicals, Inc.). Primer extension reactions wereperformed with Molony murine leukemia virus reverse transcriptase (GIBCOBRL Life Technologies, Inc.) in 20 ml of buffer comprising 50 mM Tris-HCl (pH7.5), 60 mM KCl, 10 mM MgCl2, 2 mM (each) deoxyribonucleoside triphos-phates, 1 mM dithiothreitol, and 50 mg of actinomycin D per ml. The resultantcDNAs were subjected to urea-polyacrylamide gel electrophoresis as describedpreviously (9).

S1 nuclease mapping. Probe DNAs for S1 nuclease mapping were prepared asfollows. To map the 59 end of the iolRS transcript, a PiolRS region (417 bp) wasamplified from the chromosome by PCR with a pair of primers (59-CTCATCTAAGGAAACAGTGCCATG-39 and 59-TATCCACTTTTCAGCAAGCC-39).The PCR product was labeled with T4 polynucleotide kinase and [g-32P]ATPand then digested with NsiI to obtain a probe DNA (354 bp) labeled only at the59 end of the coding strand. To map the 39 ends of the iol and iolRS transcripts,probe DNAs were prepared as follows. An iol transcriptional terminator (Tiol)region (391 bp) was amplified from the chromosome by PCR with a pair ofprimers (59-AACAGCGATCTGTACGAACC-39 and 59-TATCAAGCAGCACCACATCG-39) and then digested with Sau3AI. The iolRS transcriptional ter-minator (TiolRS) region (383 bp) was amplified by PCR with another pair ofprimers (59-ACAACGTGGATATCCCTCAC-39 and 59-AATAAAAGAAGCT

GGGTCAGC-39) and then digested with HpaII. These Sau3AI and HpaII di-gests were labeled with the Klenow fragment of DNA polymerase I, [a-32P]dCTP(ICN Biomedicals, Inc.), and the three other unlabeled deoxynucleoside triphos-phates. The 303- and 335-bp fragments labeled at their coding strands wereisolated and then used for mapping of the 39 ends of the iol and iolRS transcripts,respectively.

Each total-RNA sample (100 mg) was hybridized with the respective labeledprobe DNAs. The hybrids were treated with S1 nuclease (Takara Shuzo Co.,Ltd.) in a buffer comprising 250 mM NaCl, 50 mM sodium acetate (pH 4.4), 4.5mM ZnSO4, and 20 mg of calf thymus DNA per ml. The S1 nuclease-resistanthybrids were subjected to urea-polyacrylamide gel electrophoresis as describedpreviously (9).

Northern blot analysis. Total RNAs as well as RNA size markers (GIBCOBRL Life Technologies, Inc.) were electrophoresed in formaldehyde gels andblotted (18). The iolR region (789 bp) was amplified from the chromosome byPCR with a pair of primers (59-ACGTTAGGCCAGATGAATGC-39 and 59-TACGTAAAGATTCAAGAAGGAG-39). The 59 end of the iolS gene (497 bp)was also amplified with another pair of primers (59-CATTTTATTCTAGGAGGCAG-39 and 59-GAAGTTGGATACACCGATGG-39). These two PCR prod-ucts were labeled with [a-32P]dCTP and a BcaBEST labeling kit (Takara ShuzoCo., Ltd.) and used as probes. Hybridization was carried out under standardconditions, and the blots were washed three times for 15 min each at 42°C in 23SSC (13 SSC is 0.15 M NaCl plus 0.015 M sodium citrate) containing 0.1%sodium dodecyl sulfate (18).

RESULTS

Organization of the B. subtilis iol operon. The B. subtilis idhgene is one of the iol genes involved in inositol catabolism (11).Previous determination of the nucleotide sequence of the iolregion of the Bacillus subtilis genome and a homology search ofthe Iol proteins (23–25) allowed us to predict the structure ofthe iol operon for inositol catabolism, which consists of 10 iolgenes (iolA to iolJ; iolG corresponds to idh) and includes asA-dependent promoter-like sequence in the 59 region of iolA(Piol) and a r-independent terminator-like sequence in the 39region of iolJ (Tiol) (Fig. 1). As described below, the iol genesare transcribed in the presence of inositol, most probably fromPiol to Tiol, resulting in a 11.5-kb mRNA.

Upstream of the iol operon, two genes (iolR and iolS) ori-ented divergently to the iol operon were found (24, 25). Asdescribed below, these two genes were most probably cotran-scribed from another sA-dependent and inositol-induciblepromoter (PiolRS) to a r-independent terminator (TiolRS),and the iolR gene was found to encode a negative regulator ofthe iol operon (Fig. 1).

Identification of a promoter of the iol operon. To detect theiol transcript, we performed primer extension analysis withtotal RNAs prepared from strain 60015 cells grown with andwithout inositol (Fig. 2). One major (approximately 130-base)and several minor reverse transcripts were detected whenRNA prepared from cells grown with inositol was hybridizedwith a DNA primer complementary to a sequence in the 59region of iolA (Fig. 2A, lane 2), whereas no reverse transcriptwas detectable when RNA from cells grown without inositolwas primed in the same way (lane 1). The 59 end of this iol

FIG. 1. Organization and transcription of the iol genes of B. subtilis. The ioloperon consists of the iolABCDEFGHIJ genes and is most probably transcribedfrom the iol promoter (Piol) and ends at the iol terminator (Tiol). The iolRSgenes are most probably cotranscribed from the iolRS promoter (PiolRS) to theiolRS terminator (TiolRS). The iolG and iolR genes (solid arrows) encode Idh(11) and a negative regulator of the iol operon (see the text), respectively.

TABLE 1. B. subtilis strains constructed in this study andtheir genotypes

Strain Genotype

60015........................trpC2 metC7 (our standard)YF244 ......................trpC2 metC7 iolR::catYF246 ......................trpC2 metC7 iolS::catYF248 ......................trpC2 metC7 Piol::catYF323 ......................trpC2 metC7 amyE::[Piol(2107) iolA9-9lacZ cat]YF324 ......................trpC2 metC7 amyE::[Piol(245) iolA9-9lacZ cat]YF325 ......................trpC2 metC7 amyE::[Piol(223) iolA9-9lacZ cat]YF327 ......................trpC2 metC7 amyE::[PiolRS(294) iolR9-9lacZ cat]YF328 ......................trpC2 metC7 amyE::[PiolRS(270) iolR9-9lacZ cat]YF329 ......................trpC2 metC7 amyE::[PiolRS(243) iolR9-9lacZ cat]YF330 ......................trpC2 metC7 amyE::[PiolRS(219) iolR9-9lacZ cat]

4592 YOSHIDA ET AL. J. BACTERIOL.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

transcript was determined to be 191 bp upstream of iolA (Fig.2B). These results strongly suggested that Piol might be an inos-itol-inducible iol promoter because of its position at the expectedinterval upstream from the 59 end of the iol transcript (Fig. 2).

To determine whether Piol is only the promoter for tran-scription of the iolABCDEFGHIJ genes, we disrupted Piol byinsertion of a cat cassette into its 210 region and designatedthe resultant strain YF248 (Piol::cat); this strain could not growon inositol as the sole carbon source. We also examined theinduction of Idh encoded by iolG in strain YF248 (Table 2).Idh production by strain YF248 was almost completely abol-ished even in the presence of inositol, while it was induced46-fold in strain 60015 upon the addition of inositol (Table 2).These data indicate that expression of iolG is under Piol con-trol and suggest that the iolABCDEFG genes (possibly alsoiolHIJ) are probably transcribed only from Piol.

Identification of a negative regulator gene for the iol operon(iolR), and involvement of IolR in catabolite repression of Idh

synthesis. The IolR protein belongs to the DeoR family ofbacterial regulatory proteins (24). To determine whether theIolR protein is involved in the regulation of the iol operon, theiolR gene was disrupted by insertion of a cat cassette to pro-duce strain YF244 (iolR::cat). Strain 60015 induced Idh uponthe addition of inositol (Table 2). However, strain YF244 pro-duced the highest level of Idh in the absence of inositol (Table2), probably due to catabolite repression by inositol itself.Primer extension analysis with total RNAs prepared from cellsof strain YF244 grown with or without inositol revealed thattranscription from Piol had become constitutive in strainYF244 (Fig. 2A, lanes 3 and 4). Table 2 also shows that Idhsynthesis was more severely repressed in strain 60015 (175-fold) than in strain YF244 (3.6-fold) by the addition of glucose,suggesting that IolR might be partially involved in cataboliterepression of the iol operon.

We could not determine from the results of iolR disruptionwhether the above phenotype of strain YF244 is due to loss of

FIG. 2. Mapping of the 59 end of a transcript of the iol operon by primer extension analysis. Primer extension analysis was performed as described in the text. Thelower section of the figure shows the nucleotide sequence of the noncoding strand of the iol promoter region. The locations of the labeled primer used for extensionanalysis and the putative transcription start site (11) are indicated. The putative Shine-Dalgarno (SD) sequence for iolA, and the 235 and 210 sequences of Piol areunderlined. The amino acid sequence of the N-terminal part of the iolA gene product is shown beneath the sequence. (A) Primer extension was performed with totalRNAs prepared from cells of strains 60015 (lanes 1 and 2) and YF244 [iolR::cat] (lanes 3 and 4) grown in the absence (lanes 1 and 3) or presence (lanes 2 and 4) of10 mM inositol. The arrow to the right of the panel indicates the position of the main reverse transcript. The positions of the DNA size markers and the primer (p)are indicated to the left of the panel. The panel was prepared from two autoradiograms (lanes 1 and 2, and lanes 3 and 4) obtained in experiments performed at 1-weekintervals. (B) Fine mapping of the 59 end of the transcript. The primer extensions were performed with total RNAs prepared from cells of strain 60015 grown in theabsence (lane 1) or presence (lane 2) of 10 mM inositol. The dideoxy sequencing ladders (lanes G, A, T, and C) were created with the same primer. The part of thenucleotide sequence of the noncoding strand corresponding to these ladders is shown with the 235 and 210 regions of Piol (underlined) and the putative transcriptionstart site (11).

VOL. 179, 1997 B. SUBTILIS myo-INOSITOL OPERON 4593


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

iolR or the downstream iolS, because the cat cassette we usedhas a transcriptional terminator. Therefore, we also disruptediolS by insertion of the cat cassette to produce strain YF246(iolR::cat). As shown in Table 2, the phenotype of strain YF246was indistinguishable from that of strain 60015 with regard toinositol inducibility and catabolite repression of Idh synthesis.These overall results clearly indicate that the iolR gene is anegative regulator gene for transcription from Piol and that theIolR protein is also involved in catabolite repression at Piol,probably via inducer exclusion or inducer expulsion.

Identification of a transcriptional terminator of the ioloperon. To localize the 39 end of the iol transcript supposedlyin the region downstream of iolJ, we first performed S1 nucle-ase analysis with total RNA prepared from cells of strain 60015grown with and without inositol (Fig. 3A, lanes 1 and 2). Whena probe DNA complementary to the Tiol region (Fig. 3) washybridized with RNA (with inositol), a large amount of aprotected fragment (approximately 65 bases) was detected(Fig. 3A, lane 1). However, when the DNA was hybridized withRNA from cells grown without inositol, only a faint band ofthis protected fragment was observed (lane 2). The end of theprotected fragment protected by RNA from cells grown withinositol was finely mapped (Fig. 3B), and this work showed thatthe 39 end of the inositol-induced transcript was at a T-richregion following a stem-loop structure downstream of iolJ (Tiol[Fig. 1 and 3B]) (23). This suggests that Tiol is most probablya transcription terminator of the iol operon. To confirm thatthis inositol-induced transcript is synthesized from Piol, weperformed S1 nuclease analysis with total RNAs of cells ofstrain YF248 (Piol::cat) grown with and without inositol. Asshown in Fig. 3A, we could not detect the above protectedfragment on hybridization with YF248 RNA with inositol (lane3) or without inositol (lane 4). The fact that Piol disruptionresulted in the disappearance of the inositol-induced transcrip-tion terminating at Tiol implies that this transcription mightstart from Piol. Moreover, S1 nuclease analysis involvingRNAs of cells of strain YF244 (iolR::cat) grown with andwithout inositol (lanes 5 and 6) indicated that the iol transcriptterminating at Tiol was synthesized constitutively when theiolR gene was disrupted. Therefore, these results suggestedthat the 10 iol genes (iolA to iolJ) are most probably tran-scribed from Piol to Tiol as a single 11.5-kb mRNA (Fig. 1).

Transcription of the iolR and iolS genes. The iolRS genes aredivergently oriented with respect to the iol operon (Fig. 1).Transcription of iolR and iolS was analyzed by Northern blotanalysis (Fig. 4). A DNA probe containing the coding region ofiolR hybridized to a 2.0-kb transcript, whose synthesis wasinduced by inositol (Fig. 4, lanes 1 and 2). The distance from

PiolRS to TiolRS was estimated to be approximately 2.0 kb. Wecannot adequately explain the appearance of a hybridizingband at approximately 1.4 kb. Another DNA probe containingthe N-terminal part of the coding region of iolS also hybridizedwith a 2-kb transcript, whose synthesis was also induced byinositol (lanes 3 and 4). These results suggest that the iolR andiolS genes might be transcribed as a single 2.0-kb mRNA,which is induced by inositol. Fine S1 nuclease mapping (Fig.5A and B) with DNA probes containing PiolRS or TiolRS andtotal RNAs of cells of strain 60015 grown with (lane 1) orwithout (lane 2) inositol enabled us to identify the 59 end (Fig.

FIG. 3. Mapping of the 39 end of the iol transcript by S1 nuclease analysis.The lower section of the figure shows our strategy for S1 nuclease mapping of the39 end of the iol transcript. Solid and dotted lines indicate the regions of theprobe (303 bp) protected and not protected against S1 nuclease, respectively.The position of the 32P label is indicated by 39. The details of the experimentalprocedures are given in the text. (A) S1 nuclease analysis of the iol transcript.Total RNAs prepared from cells of strains 60015 (lanes 1 and 2), YF248[Piol::cat] (lanes 3 and 4), and YF244 [iolR::cat] (lanes 5 and 6) grown in thepresence (lanes 1, 3, and 5) or absence (lanes 2, 4, and 6) of 10 mM inositol werehybridized with the probe and then treated with S1 nuclease. The positions of theprotected DNA fragments and the probe are indicated by the large and smallarrows to the right, respectively. The positions of the DNA size standards areindicated to the left. (B) Fine mapping of the 39 end of the iol transcript. S1nuclease mapping was performed with total RNAs prepared from cells of strain60015 grown in the presence (lane 1) or absence (lane 2) of 10 mM inositol. Thechemical cleavage sequencing ladders specific to purine (GA) and pyrimidine(CT) nucleotides of the same probe are shown. The corresponding part of thenucleotide sequence of the noncoding strand is shown to the left of the panel,with a pair of convergent arrows representing the palindromic sequence. Theends of the fragment protected against S1 nuclease are also indicated by arrow-heads.

TABLE 2. Idh synthesis in B. subtilis iol disruptants

Strain(relevant genotype)

Idh activity (nmol/min/mg)a incells grown with:

Nothing Iol Iol and Glc

60015 (wild type) 19 876 5YF248 (Piol::cat) 4 11 0YF244 (iolR::cat) 2,579 1,576 434YF246 (iolS::cat) 28 732 8

a Cells of B. subtilis strains were grown to an A600 of 0.75 in S6 mediumcontaining 0.5% Casamino Acids with or without 10 mM inositol (Iol) or with 10mM each inositol and glucose (Glc). Chloramphenicol (5 mg/ml) was also addedto the cultures of strains YF244, YF246, and YF248. The cells (4.5 A600 units)were harvested and then treated with lysozyme and sonicated briefly to preparecrude extracts (14). The Idh activity in the crude extracts was measured asdescribed previously (17). Each value is the average of four determinations.



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

5A) and the 39 end (Fig. 5B) of this 2.0-kb transcript. This workindicated that PiolRS and TiolRS (Fig. 1 and 5) are mostprobably the transcription promoter and terminator of theiolRS operon, respectively. In addition, S1 mapping (Fig. 5C)and Northern blotting (data not shown) revealed that the in-ositol-induced transcription from PiolRS became almost con-stitutive when iolR was disrupted. As shown in Fig. 5C, thebands of the ;100-base fragment protected on hybridizationwith RNA of strain YF244 (iolR::cat) (with inositol [lane 3] orwithout inositol [lane 4]) were less dense than with that due toprotection with RNA of strain 60015 (with inositol [lane 1]).This might be due to the instability of the mRNA encoding thedisrupted iolR. From these results, we conclude that the iolRSgenes are most probably cotranscribed from PiolRS, which isnegatively autoregulated by IolR in the absence of inositol.

Confirmation of the iol and iolRS promoters by deletionanalysis. To confirm that Piol and PiolRS function as promot-ers in vivo, we performed deletion analysis of these promotersequences with iolA9-9lacZ or iolR9-9lacZ integrated into thechromosomal amyE locus as a reporter gene (Fig. 6; Table 3).When the Piol(2107) insert, including nucleotides 2107 to1245, was fused to lacZ and then integrated into amyE (Fig.6), efficient LacZ synthesis (12.2-fold) was induced upon theaddition of inositol to the medium (Table 3). For the Piol(245)insert with the deletion of a region upstream of the 235 regionof Piol, LacZ synthesis was induced to a similar extent (12.3-fold), but the activities with and without the addition of inositolwere 3.8 times lower than those directed by Piol(2107). Fur-ther deletion including the 235 region of Piol [Piol(223)]completely abolished the promoter activity of the iol operon.These results clearly indicated that Piol functions as a pro-moter in vivo, its activity being enhanced by the AT-rich regionupstream of its 235 region (nucleotides 2107 to 245), and

that cis-acting sequences responsible for inositol induction (orsome of them) are located between nucleotides 245 and 1245.

When the PiolRS(294) insert including nucleotides 294 to162 was fused to lacZ and then integrated into amyE (Fig. 6),LacZ synthesis was induced 16.3-fold upon the addition ofinositol, but the activity was 68 times lower than that directedby Piol(2107) (Table 3). This activity was decreased 2.5-foldon the deletion of nucleotides 294 to 270 and further de-creased 6.4-fold by a deletion of nucleotides 294 to 243, wheninositol was added to the medium. When nucleotides 294 to219 including the 235 region of PiolRS were deleted, the

FIG. 4. Northern analysis of the iolRS transcripts. The lower section of thefigure shows the locations of the iolS (497-bp) and iolR (789-bp) probes used forNorthern analysis. Northern blotting was performed as described in the text, withtotal-RNA samples prepared from cells of strain 60015 grown in the absence(lanes 1 and 3) or presence (lanes 2 and 4) of 10 mM inositol. The blots werehybridized with the iolR (lanes 1 and 2) or iolS (lanes 3 and 4) probe. Thepositions of a 2.0-kb transcript and an anomalous 1.4-kb band are indicated tothe right of the panel.

FIG. 5. Mapping of the 59 and 39 ends of the iolRS transcript. The lowerright-hand section of the figure shows the strategy for S1 nuclease mapping of the59 and 39 ends of the iolRS transcript. Solid and dotted lines represent theprotected and unprotected regions of the 354- and 335-bp probes, respectively.The position of the 32P label is indicated by 59 or 39. The details of the experi-mental procedures are given in the text. (A) Fine S1 nuclease mapping of the 59end of the transcript. The 354-bp probe was hybridized with total RNAs pre-pared from cells of strain 60015 grown in the presence (lane 1) or absence (lane2) of 10 mM inositol. The chemical cleavage sequencing ladders specific topurine (GA) and pyrimidine (CT) nucleotides of the same probe are also pre-sented. Part of the nucleotide sequence of the noncoding strand is shown to theleft of the panel; the 39 end of the fragment protected against S1 nuclease isindicated by asterisks. The 235 and 210 regions of PiolRS are underlined, andthe major putative transcription start site (11) is also indicated. (B) Fine S1nuclease mapping of the 39 end of the transcript. The 335-bp probe was hybrid-ized with total RNAs of strain 60015. The lane assignments are the same as inpanel A. Part of the nucleotide sequence of the noncoding strand correspondingto the sequencing ladders is shown, with a pair of convergent arrows representinga palindromic sequence; the 59 ends of the fragment protected against S1 nu-clease are indicated by arrowheads. (C) S1 nuclease analysis of transcriptionfrom PiolRS. The 354-bp probe was hybridized with total RNAs of strains 60015(lanes 1 and 2) and strain YF244 (lanes 3 and 4), grown in the presence (lanes1 and 3) or absence (lanes 2 and 4) of inositol. The positions of DNA sizemarkers are indicated to the left of the panel.



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

promoter activity was completely abolished. These resultsclearly indicated that PiolRS functions as a promoter in vivo,with its activity being enhanced by the AT-rich region up-stream of its 235 region (nucleotides 294 to 243). Further-more, deletion of nucleotides 294 to 270 decreased the in-ducibility of LacZ synthesis to 3.9-fold, and deletion ofnucleotides 294 to 243 caused the constitutive synthesis ofLacZ. The latter facts indicated that cis-acting sequences re-sponsible for inositol induction (or part of them) are includedin nucleotides 294 to 243.

DISCUSSION

The results in this communication imply that the transcrip-tion of the iol operon most probably starts from Piol andterminates at Tiol, resulting in an 11.5-kb mRNA encoding the10 iolABCDEFGHIJ genes (Fig. 1). All the iol genes are in-volved in inositol catabolism because point mutations or dis-ruption of each of these genes affected inositol utilization (22).We could eliminate the slight possibility that a gene tran-scribed from Piol was a positive regulator for one or moreother promoters upstream of iolJ, because the iol operon wasexpressed constitutively in the iolR background even in thepresence of any point mutation or disruption of each of the iolgenes (22).

The fact that the iol operon consists of 10 iol genes meansthat inositol catabolism requires many specific steps before

common pathways are entered. In addition, two more genes(iolRS), which are cotranscribed (Fig. 4), seem to be involvedin inositol catabolism. Among the products of these 12 iolgenes, only the protein encoded by iolG and iolR are known(Idh [11] and a negative regulator of the iol operon [see above],respectively). To further deduce the function of the iol geneproducts in inositol catabolism, we searched for homology to

FIG. 6. Deletion analysis of the iol and iolRS promoter regions. Piol and PiolRS inserts fused to lacZ for deletion analysis are shown in the upper part of the figure.The Piol (2107) insert contains nucleotides 2107 to 1245, including the 59 part of the iolA coding region (nucleotides 1192 to 1245). The Piol(245) and Piol(223)inserts contain nucleotides 245 to 1245 and nucleotides 223 to 1245, respectively. The PiolRS (294) insert contains nucleotides 294 to 162, including the 59 partof the iolR coding region (nucleotides 129 to 162). The PiolRS(270), PiolRS(243), and PiolRS(219) inserts contain nucleotides 270 to 162, 243 to 162, and 219to 162, respectively. Nucleotides 2107 to 2103 of the Piol region overlap nucleotides 290 to 294 of the PiolRS region. In the lower part of the figure, the sequencesof the Piol and PiolRS regions are presented together with the deletion endpoints. After cloning of each of the inserts into plasmid ptrpBGI (19), iolA9-9lacZ underPiol control or iolR9-9lacZ under PiolRS control was integrated into amyE by means of double crossover by selecting chloramphenicol-resistant transformants, asdescribed in the text.

TABLE 3. Chromosome-encoded LacZ synthesis directed by the iolor iolRS promoter

Strain Promoter

LacZ activity (nmol/min/mg)a 1Iol/2Iol

ratio (fold)2 Iol 1 Iol

YF323 Piol(2107) 1,140 13,900 12.2YF324 Piol(245) 299 3,670 12.3YF325 Piol(223) ,1.0 ,1.0YF327 PiolRS(294) 12.6 205 16.3YF328 PiolRS(270) 20.8 81.1 3.90YF329 PiolRS(243) 40.8 32.1 0.79YF330 PiolRS(219) ,1.0 ,1.0

a For the LacZ assay, cells carrying amyE::[Piol iolA9-9lacZ] or amyE::[PiolRSiolR9-9lacZ] were grown to an A600 of 0.6 in S6 medium containing 0.5%Casamino Acids and 5 mg of chloramphenicol per ml, with or without 10 mMinositol (Iol). The cells (3.6 A600 units) were harvested and then lysed by ly-sozyme treatment and brief sonication (14). LacZ was assayed spectrophoto-metrically (3).



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

known proteins in sequence libraries, using the FASTA pro-gram to obtain FASTA optimized scores (15). The results ofthe latest homology search (November 1996) are shown inTable 4. Of the 12 iol gene products, 9 exhibited significanthomology (FASTA optimized scores, .200). However, threeproducts (IolB, IolH, and IolI) seemed to be unique, exhibitingno significant homology to known proteins (FASTA optimizedscores, ,130).

Although the pathway of inositol catabolism in B. subtilis hasnot been well studied, outlines of this pathway in K. aerogeneshave been presented (1, 2, 4, 5). Assuming that the pathway in

B. subtilis is similar to that in K. aerogenes, we can suggest thefunction of several iol gene products in this pathway (Fig. 7)based on the results of the homology search (Table 4). Inositolis transported into the B. subtilis cell, presumably by the IolFprotein, because this protein showed significant homology tothe proline/betaine transporter and dicarboxylic acid transportprotein (Table 4). The transported inositol is then dehydroge-nated by Idh (encoded by iolG). The resulting 2-keto-myo-inositol is dehydrated to D-2,3-diketo-4-deoxy-epi-inositol,which is then hydrated to 2-deoxy-5-keto-D-gluconic acid, thefirst open-chain intermediate, although the genes involved in

FIG. 7. Hypothetical pathway of inositol catabolism and the putative functions of the Iol proteins. CoA, coenzyme A.

TABLE 4. Results of comparison with protein databases of the iol genes

Genea Homologous protein and organism(s) Database and entryb FASTAoptimized score

% Identity for amino acids(no. of amino acids)

iolA Metylmalonate-semialdehydedehydrogenasesPseudomonas aeruginosa SP, MMSA_PSEAE 1,448 44.2 (482)Rat SP, MMSA_RAT 1,448 44.2 (482)

iolC Fructokinase (Beta vulgaris) GP, BVU37838_1 454 27.7 (321)2-Dehydro-3-deoxygluconokinase (B. subtilis) SP, KDGK_BACSU 428 28.4 (320)

iolD Acetolactate synthasesSpirulina platensis SP, ILVB_SPIPL 353 26.4 (545)Synechococcus sp. PRF, 1611501A 341 26.0 (526)

iolE MocC protein (Rhizobium meliloti) SP, MOCC_RHIME 667 37.9 (285)

iolF Proline/betaine transporter (E. coli) SP, PROP_ECOLI 230 23.8 (432)Dicarboxylic acid transport protein

(Pseudomonas putida)GP, PPU48776_1 223 23.7 (389)

iolG Streptomycin biosynthesis protein StrI(Streptomyces griseus)

SP, STRI_STRGR 344 24.6 (345)

MocA protein (R. meliloti) SP, MOCA_RHIME 225 25.0 (316)

iolJ Fructose-bisphosphate aldolase (B. subtilis) SP, ALF1_BACSU 1,073 57.9 (285)Tagatose-bisphosphate aldolase (E. coli) SP, AGAY_ECOLI 659 38.4 (289)

iolR Glucitol operon repressor (E. coli) SP, SRLR_ECOLI 438 30.2 (245)Glp repressor (P. aeruginosa) GP, PAU49666_5 412 28.5 (249)

iolS MocA protein (Agrobacterium tumefaciens) PFR, 2214302F 482 31.3 (316)Auxin-induced protein (common tobacco) SP, A115_TOBAC 430 29.9 (298)

a The iol genes whose products exhibited significant homology to known proteins (FASTA optimized scores [15], .200) are listed.b Among the homologous proteins, the top two proteins are shown. Protein sequence databases are abbreviated as follows: PRF, Protein Research Foundation

protein database; SP, SwissProt protein sequence database; GP, translated protein sequence database from the NCBI-GenBank nucleotide sequence database. Thedatabase entries of GP are followed by an underlined space and an ordinal number for each of the coding sequences.



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

these reactions cannot be suggested at present. This open-chain intermediate is then phosphorylated, probably by theIolC protein, because this protein showed meaningful homol-ogy to fructokinase and 2-dehydro-3-deoxygluconokinase (Ta-ble 4). The resulting 2-deoxy-5-keto-D-gluconic acid 6-phos-phate is probably cleaved by the IolJ protein, which exhibitshigh homology to fructose-bisphosphate aldolase and tagatose-bisphosphate aldolase (Table 4), to yield dihydroxyacetonephosphate and malonic semialdehyde. The latter is convertedto acetyl coenzyme A and CO2, most probably by the iolAproduct, which exhibits very high homology to methylmalo-nate-semialdehyde dehydrogenases (Table 4). Thus, inositolcould be degraded finally to dihydroxyacetone and acetyl co-enzyme A.

In addition to three iol gene products (IolB, IolH, and IolI)which show no significant homology to known proteins, wewere unable to suggest any function for the IolD, IolE, andIolS proteins in inositol catabolism, although they exhibitedsignificant homology to known proteins (Table 4).

The iolR gene was found to encode a negative regulator forthe iol operon (Fig. 2 and 3; Table 2) and to be cotranscribedwith the iolS gene from PiolRS (Fig. 4 and 5; Table 3). TheIolR protein exhibited significant homology to DNA bindingproteins of the DeoR family of bacterial regulatory proteins(Table 4) (24). It was very recently found that the IolR proteinactually binds to an operator in the iol promoter region (22),indicating that this protein is a repressor for the iol operon.However, we do not think that the IolS protein is necessary forinositol catabolism because disruption of the iolS gene affectedneither growth on inositol (22) nor the inducibility and catab-olite repression of Idh synthesis (Table 2).

Currently, catabolite repression of B. subtilis is thought to bedue to negative regulation of gene expression involving theCcpA and HPr proteins (with the latter phosphorylated on aserine residue) and catabolite-responsive elements found invarious catabolic operons (10). It has been reported that CcpAand HPr are also involved in catabolite repression of Idh syn-thesis (6, 13). Catabolite repression of the iol operon was alsoattributed to its induction system, probably via inducer exclu-sion or inducer expulsion, because disruption of the iolR genepartially affected catabolite repression of Idh synthesis (Table2).

ACKNOWLEDGMENTS

We thank Y. Fujii, T. Fujiwara, R. Koga, M. Masuda, M. Nagaike,A. Noda, and Y. Ohshima for their help in the experiments.

This work was supported by a Grant-in-Aid for the Encouragementof Young Scientists from the Ministry of Education, Science and Cul-ture of Japan to K. Yoshida.

REFERENCES

1. Anderson, W. A., and B. Magasanik. 1971. The pathway of myo-inositoldegradation in Aerobacter aerogenes. Identification of the intermediate 2-de-oxy-5-keto-D-gluconic acid. J. Biol. Chem. 246:5653–5661.

2. Anderson, W. A., and B. Magasanik. 1971. The pathway of myo-inositol

degradation in Aerobacter aerogenes. Conversion of 2-deoxy-5-keto-D-glu-conic acid to glycolytic intermediates. J. Biol. Chem. 246:5662–5675.

3. Atkinson, M. R., L. V. Wray, Jr., and S. H. Fisher. 1990. Regulation ofhistidine and proline degradation enzymes by amino acid availability inBacillus subtilis. J. Bacteriol. 172:4758–4765.

4. Berman, T., and B. Magasanik. 1966. The pathway of myo-inositol degra-dation in Aerobacter aerogenes. Dehydrogenation and dehydration. J. Biol.Chem. 241:800–806.

5. Berman, T., and B. Magasanik. 1966. The pathway of myo-inositol degra-dation in Aerobacter aerogenes. Ring scission. J. Biol. Chem. 241:807–813.

6. Deutscher, J., J. Reizer, C. Fischer, A. Galinier, M. H. Saier, Jr., and M.Steinmetz. 1994. Loss of protein kinase-catalyzed phosphorylation of HPr, aphosphocarrier protein of the phosphotransferase system, by mutation of theptsH gene confers catabolite repression resistance to several catabolic genesof Bacillus subtilis. J. Bacteriol. 176:3336–3344.

7. Fujita, Y., and E. Freese. 1981. Isolation and properties of a Bacillus subtilismutant unable to produce fructose-bisphosphatase. J. Bacteriol. 145:760–767.

8. Fujita, Y., and T. Fujita. 1983. Genetic analysis of a pleiotropic deletionmutation (Digf) in Bacillus subtilis. J. Bacteriol. 154:864–869.

9. Fujita, Y., T. Fujita, Y. Miwa, J. Nihashi, and Y. Aratani. 1986. Organizationand transcription of the gluconate operon, gnt, of Bacillus subtilis. J. Biol.Chem. 261:13744–13753.

10. Fujita, Y., Y. Miwa, A. Galinier, and J. Deutscher. 1995. Specific recognitionof the Bacillus subtilis gnt cis-acting catabolite-responsive element by a pro-tein complex formed between CcpA and seryl-phosphorylated HPr. Mol.Microbiol. 17:953–960.

11. Fujita, Y., K. Shindo, Y. Miwa, and K. Yoshida. 1991. Bacillus subtilis inositoldehydrogenase-encoding gene (idh): sequence and expression in Escherichiacoli. Gene 108:121–125.

12. Higuchi, R. 1990. Recombinant PCR, p. 177–183. In M. A. Innis, D. H.Gelfand, J. J. Sninsky, and T. J. White (ed.), PCR protocols: a guide tomethods and applications. Academic Press, Inc., San Diego, Calif.

13. Miwa, Y., M. Saikawa, and Y. Fujita. 1994. Possible function and someproperties of the CcpA protein of Bacillus subtilis. Microbiology 140:2567–2575.

14. Nihashi, J., and Y. Fujita. 1984. Catabolite repression of inositol dehydro-genase and gluconate kinase syntheses in Bacillus subtilis. Biochim. Biophys.Acta 798:88–95.

15. Pearson, W. R., and D. J. Lipman. 1988. Improved tools for biologicalsequence comparison. Proc. Natl. Acad. Sci. USA 85:2444–2448.

16. Poole, P. S., A. Blyth, C. J. Reid, and K. Walters. 1994. myo-Inositol catab-olism and catabolite regulation in Rhizobium leguminosarum bv. viciae. Mi-crobiology 140:2787–2795.

17. Ramaley, R., Y. Fujita, and E. Freese. 1979. Purification and properties ofBacillus subtilis inositol dehydrogenase. J. Biol. Chem. 254:7684–7690.

18. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: alaboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, ColdSpring Harbor, N.Y.

19. Shimotsu, H., and D. J. Henner. 1986. Construction of a single-copy inte-gration vector and its use in analysis of regulation of the trp operon ofBacillus subtilis. Gene 43:85–94.

20. Vidal-Leiria, M., and N. van Uden. 1973. Inositol dehydrogenase from theyeast Cryptococcus melibiosum. Biochim. Biophys. Acta 293:295–303.

21. Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. Improved M13 phagecloning vectors and host strain: nucleotide sequences of the M13mp18 andpUC19 vectors. Gene 33:103–119.

22. Yoshida, K., and Y. Fujita. Unpublished data.23. Yoshida, K., H. Sano, Y. Miwa, N. Ogasawara, and Y. Fujita. 1994. Cloning

and nucleotide sequencing of a 15 kb region of the Bacillus subtilis genomecontaining the iol operon. Microbiology 140:2289–2298.

24. Yoshida, K., S. Seki, M. Fujimura, Y. Miwa, and Y. Fujita. 1995. Cloning andsequencing of a 36-kb region of the Bacillus subtilis genome between the gntand iol operons. DNA Res. 2:61–69.

25. Yoshida, K., K. Shindo, H. Sano, S. Seki, M. Fujimura, N. Yanai, Y. Miwa,and Y. Fujita. 1996. Sequencing of a 65 kb region of the Bacillus subtilisgenome containing the lic and cel loci, and creation of a 177 kb contigcovering the gnt-sacXY region. Microbiology 142:3113–3123.



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

organization and transcription of the myo-inositol operon ... · inositol as the sole carbon...

Documents