JOURNAL OF BACTERIOLOGY, May 1989, p. 2435-2442 Vol. 171, No.

5
0021-9193/89/052435-08$02.00/0
Copyright C) 1989, American Society for Microbiology

Bacillus licheniformis oL-Amylase Gene, amyL, Is Subject to

Promoter-Independent Catabolite Repression in Bacillus subtilis
BRID M. LAOIDE,lt* GLENN H. CHAMBLISS,2 AND DAVID J. McCONNELL'
Department of Genetics, Trinity College, University of Dublin, Dublin 2, Ireland,' and Department of Bacteriology,
University of Wisconsin, Madison, Wisconsin 537062
Received 4 November 1988/Accepted 1 February 1989

Expression of the Bacillus licheniformis a-amylase gene, amyL, was temporally activated and subject to
catabolite repression both in its natural host and when cloned on a 3.55-kilobase fragment in Bacillus subtilis.
A subclone from which the promoter region of amyL and sequences upstream from the promoter were deleted
had a low level of amylase activity. Expression of the promoterless gene was still subject to repression by glucose
when the gene was present either on a multicopy plasmid or integrated into the B. subtilis chromosome.
Catabolite repression occurred independently of the amylase promoter and irrespective of the distance of the
promoterless amyL gene from the promoter which transcribed it. The transcriptional start sites of amyL
activated by its own promoter and by a vector sequence promoter were determined by Si mapping.
a-Amylase-specific mRNA levels were measured in repressing and nonrepressing media, and catabolite
repression was found to act at the level of transcription.

Catabolite repression is the repression of enzyme synthe- 1983). Here we confirm these earlier findings and report
sis that occurs when bacterial cultures are grown in the evidence that catabolite repression of amyL occurs at the
presence of a readily metabolized carbon source, such as level of transcription, independently of the promoter from
glucose (13). This phenomenon, which was initially de- which the gene is transcribed and irrespective of the distance
scribed as the glucose effect (6, 19), is a feature of both between the structural amyL gene and the promoter which
gram-positive and gram-negative organisms. Early studies activates it.
on Escherichia coli demonstrated that the addition of glu-
cose to bacterial cultures resulted in a rapid loss of intracel- MATERIALS AND METHODS
lular cyclic AMP (cAMP) (14) and that the addition of high Bacterial strains and plasmids. The following strains were
levels of cAMP overcame, at least in part, the repressive used in this study: B. licheniformis FDO2, B. subtilis S0113
effects of glucose (35). From these and other studies (re- (trpC2 amy-3), B. subtilis S0116 [trpC2 amy-3(pSA33)] (25),
viewed in references 27 and 36), a model emerged in which and B. subtilis S0122 [trpC2 amy-3(pBS86-3)].
it was proposed that cAMP binds to a catabolite activator Plasmid pSA33 has been described previously (25). pSA33
protein (CAP) and that this complex in turn binds to catab- carries the B. licheniformis amyL gene on a 3.55-kilobase
olite-sensitive promoters exerting a positive control on cat- (kb) fragment cloned into the EcoRI site of pBD64 (8).
abolic operons. However, to associate catabolite repression Plasmid pBS86-3 was a generous gift from M. A. Stephens
solely with cAMP-CAP-dependent regulation of transcrip-

Downloaded from https://journals.asm.org/journal/jb on 07 August 2024 by 125.23.119.78.

tion is misleading. Although there is clear evidence that
and was constructed as follows. There is a unique Ndel site
present immediately 3' to the -10 promoter region of the
cAMP plays an important role in regulating the transcription amyL gene. The 3.55-kb fragment was cut at this site, and a
of a large number of operons, it has not been established as BamHI linker was added (5'-CGGATCCG-3'). A 1.9-kb
the sole mediator of catabolite repression. Indeed, mutation BamHI-HindIII fragment was then inserted into the multiple
studies have demonstrated that neither a functional CAP nor cloning site of pUC8, producing pSL5 (24). The promoter-
an active adenylate cyclase enzyme is essential to mediate less amyL gene was cut out on an EcoRI-HindlIl fragment
catabolite repression in E. coli (4, 9, 38). and inserted between the EcoRI-HindIII sites of pBD64
Many cellular and developmental processes including (after conversion of the PvuII site of pBD64 to a HindIII
sporulation, enzyme transport systems, and extracellular site), producing pBS86-3 (see Fig. 2a). pOK4 is an integrat-
enzyme synthesis are subject to catabolite repression in ing vector carrying the promoterless amyL gene and has
Bacillus species. Unlike E. coli, under physiological growth been described previously (24).
conditions Bacillus species do not contain cAMP (1, 10); Growth conditions. Overnight cultures were grown in
therefore, cAMP cannot be involved in mediating catabolite Luria broth (LB) plus 5 ,ug of the appropriate antibiotic per
repression in Bacillus species. ml. Before inoculation, the cells were pelleted in a bench
The Bacillus subtilis amyE gene is subject to catabolite centrifuge, washed, and suspended in CM medium (Spizizen
repression (3, 20). It was shown that when the Bacillus Salts supplemented with 0.1% sodium citrate, 0.05% casein
licheniformis a-amylase gene, amyL, is cloned in B. subtilis, hydrolysate, and 50 ,ug of L-tryptophan per ml) or in CM
its expression is also subject to catabolite repression (S. A. medium plus 1% glucose (1% G). All liquid cultures were
Ortlepp, Ph.D. thesis, University of Dublin, Dublin, Ireland, vigorously aerated at 30 or 37°C. Growth was monitored
spectrophotometrically (optical density at 600 nm [OD16.])
* Corresponding author. or with a Klett-Summerson colorimeter (filtered with a red
t Present address: Unite de Biochimie des Regulations Cellu- filter, no. 66). Throughout the growth cycle, 3-ml samples
laires, Institut Pasteur, 28 rue du Dr Roux, 75724 Paris Cedex 15, were removed, the cells were pelleted, and the supernatants
France. were stored at -20°C for up to 1 week. The following media
2435
2436 LAOIDE ET AL. J. BACTERIOL.

were also used to study growth curves and subsequently Copy number determination. The plasmid copy number of
ot-amylase assays: LB, LB plus 1% G, nutrient sporulation pSA33 in B. slibtilis S0113 grown in LB, CM, and CM plus
medium (NSM) (31), and NSM plus 1% G. Selective antibi- 1% G media was measured as described previously (32).
otic media contained 5 ,ug of chloramphenicol per ml and 50 RNA isolation. RNA was isolated from B. licheniformis
p.g of ampicillin per ml. FDO2, B. suibtilis S0113, B. siubtilis S0116, and B. sbibtilis
a-Amylase assays. (i) Plate test. Cells were plated on LB S0122 as follows. Each 200-ml culture was grown to the
(LBS) or CM (CMS) agar containing 0.2% soluble starch mid-exponential phase (60 Klett units) in NSM, and a 40-ml
with or without the addition of 1%G. After overnight incu- sample was removed for RNA isolation. The culture was
bation at 37°C, the plates were flooded with a 4 mM 12-160 then divided equally, and glucose was added to a 1% final
mM KI solution. Alternatively, the petri dishes were placed concentration to one half. Samples of 20 ml of cultures
over a warm dish containing 12 pellets. It is possible to detect growing in NSM and NSM plus 1%G were removed through-
haloes resulting from starch breakdown by this method out the growth cycle. To each sample, 200 [l of 1 M
without killing the cells. NaN3-0.1 M MgCl2 was added, and the culture was rapidly
(ii) Liquid assays. cx-Amylase activity was measured by a chilled in a dry ice-ethanol bath. After centrifugation, the
modification of the method of Smith and Roe (33). Linter cells were suspended in 10 ml of protoplasting buffer (15 mM
starch (0.1%) was prepared by boiling the solution and was Tris hydrochloride [pH 8], 450 mM sucrose, 8 mM disodium
kept at >80°C for the duration of the assay. Test tubes EDTA, 320 p.g of lysozyme per ml) and incubated at 0°C for
containing 1 ml of 0.1 M phosphate buffer (pH 6.9), 0.2 ml of 30 min. After another centrifugation, the protoplasts were
5 M NaCl, and 1 ml of 0.1% starch solution were preheated suspended in 1 ml of lysis buffer (10 mM Tris hydrochloride
to 93°C, and 0.5 ml of appropriately diluted culture superna- [pH 8], 10 mM NaCl, 1 mM sodium citrate, 1.5% sodium
tant was added. A control blank was treated identically dodecyl sulfate) and 30 Rl of diethyl pyrocarbonate and
except that 0.5 ml of growth medium was added instead of incubated at 37°C. After 5 min, the protoplasts were chilled
culture supernatant. After 20 min at 930C, the test tubes were on ice, 0.5 ml of cold saturated NaCl was added, and
removed and the reaction was stopped by the addition of incubation on ice was continued for a further 15 min. The
0.01% 12-0.1% KI in 1 N HCl. Activity (A) per milliliter of supernatant was retained after centrifugation at 10,000 rpm
supernatant per milligram of starch is defined as: (Sorvall SS34 rotor) for 20 min, and 7 volumes of 4 M LiCl
were added to precipitate the RNA at 4°C for 15 to 20 h. The
A = {[OD620 (blank) - OD620 (sample)]/OD620 (blank)} x 2 RNA was pelleted at 10,000 rpm in a Sorvall SS34 rotor for
x dilution factor 90 min, washed in 0.5 ml of 3 M LiCl, and dissolved in 0.4 ml
of solubilization buffer (1% sodium dodecyl sulfate, 1 mM
OD6. was found to correspond directly with cell number disodium EDTA, 10 mM Tris hydrochloride [pH 7.5]). After
and was used to determine the specific activity (activity per two phenol extractions and one phenol-chloroform extrac-
cell) of the enzyme: specific activity = A/OD600. tion, the RNA was precipitated in 2.5 volumes of 100%
Glucose assay. To test for the presence of glucose in ethanol, washed, suspended in 100 [LI of sterile dH20, and
culture supernatants, 1 ml of supernatant was added to 1 ml stored at -20 or -70°C. RNA was routinely checked on
of distilled H20 (dH2O) and 3 ml of DNS reagent (5 g of 1.1% agarose-ethidium bromide (0.5 mg/ml) gels, and the
dinitrosalicylic acid, 1 g of phenol, 0.25 g of sodium sulfate, RNA concentration was measured spectrophotometrically
100 g of sodium potassium tartate dissolved in 250 ml of 2% (OD260 and OD280).
[wt/vol] NaOH, made up to 500 ml with dH20, and stored at Northern (RNA) blotting. RNA samples were run on
4°C) and the tubes were placed in a bath of boiling H20 for vertical denaturing formaldehyde-agarose gels (15) and
transferred to a Biodyne nylon membrane (Pall Corp., Glen

Downloaded from https://journals.asm.org/journal/jb on 07 August 2024 by 125.23.119.78.

15 min. The tubes were cooled quickly, and OD64o was read.
dH2O (1 ml) was used as a blank, and 1 ml of LB or CM Cove, N.Y.). Hybridizations were performed as recom-
medium was used as a negative control. Known concentra- mended by the manufacturer.
tions of glucose were assayed to calibrate a standard curve: SI mapping. Cellular RNA (20 to 80 ,ug), 50 p.g of carrier
glucose concentration (milligrams per milliliter) versus tRNA, and 25 ng of DNA probe (30,000 to 50,000 cpm) were
AOD64O (AOD64o = OD64o [sample] - OD640 [blank]). precipitated in 3.4 volumes of 100% ethanol, and the pellet
Plasmid isolation. B. subtilis minipreparations were pre- was washed twice with 85% ethanol. The dried pellets were
pared as described previously (37), and E. coli miniprepara- suspended by occasional gentle vortexing in 35 Ill of S1
tions were prepared by the method of Birnboim and Doly (2). hybridization buffer (80% formamide, 400 mM PIPES [piper-
Large-scale plasmid DNAs were prepared similarly except azine-N,N'-bis(2-ethanesulfonic acid)] [pH 6.4], 400 mM
that the DNA was further purified by CsCI-ethidium bromide NaCl, 1 mM disodium EDTA) and heated at 85°C for 10 min
density gradient centrifugation (15). to denature the DNA, and the temperature was allowed to
DNA manipulations. Restriction enzymes, T4 DNA ligase, drop slowly to 46°C for overnight hybridization. Nuclease S1
E. coli Klenow fragment, polynucleotide kinase, nuclease (170 U) was added in 350 p.l of S1 digestion buffer (5%
S1, and polynucleotide linkers were obtained from Boehr- glycerol, 0.25 M NaCl, 0.03 M sodium acetate [pH 4.5], 1
inger Mannheim Biochemicals (Indianapolis, Ind.) and used mM ZnSO4), and the samples were incubated at 37°C for 30
as recommended by the manufacturer. Nuclease S1 was also min. The DNA-RNA hybrids were collected by ethanol
obtained as a gift from the McCardle Cancer Research precipitation, extracted with phenol-chloroform twice, re-
Laboratory, University of Wisconsin-Madison. Plasmid precipitated with ethanol, washed, dried, suspended in 5 pl
DNA was routinely analyzed by electrophoresis in 1% of loading buffer, and electrophoresed on a 6% polyacryl-
agarose gels cast and electrophoresed in 89 mM Tris-89 mM amide-8.3 M urea sequencing gel.
boric acid-2 mM disodium EDTA and stained in an ethidium
bromide (1 mg/ml) bath. DNA fragments for sequencing and RESULTS
subcloning were purified by centrifugation in a sucrose
density gradient or by the Gene Clean method (Bio 101, La Temporal activation and catabolite repression of B. licheni-
Jolla, Calif.) after isolation from an agarose gel. formis o-amylase gene. B. licheniformis FDO2 ot-amylase is a
VOL. 171, 1989 CATABOLITE REPRESSION OF B. LICHENIFORMIS amyL 2437

thermostable extracellular enzyme that is not induced by

starch. The specific activity of the enzyme was measured in 101 a -4
cultures grown in repressing (1%G) and nonrepressing (no
glucose) media at different stages throughout the growth
cycle. The activity of the enzyme in nonrepressed cultures
increased dramatically at the onset of the stationary phase 100. UF
and was subject to repression throughout the growth cycle in
cultures growing in medium containing glucose (Fig. la). -2c3
Temporal activation and catabolite repression of B. licheni- ODE iOO
formis a-amylase gene cloned in B. sublils. Plasmid pSA33 10 1. w
contains a 3.55-kb B. licheniformis fragment which includes p-U)
the amyL gene inserted at the EcoRI site of the vector
pBD64 (8). B. subtilis S0116 was made by transforming
pSA33 into the a-amylase-negative strain S0113 (25). S0116
was grown in CM medium with and without the addition of 1U I I I I I

1%G. Culture samples were removed, and the specific 0 10 20 30 35

amylase activity was measured. The expression of the TIME (hours)
cloned gene on pSA33 in B. subtilis is temporally activated
and catabolite repressed in a pattern similar to its expression 120
in its natural host, B. licheniformis FDO2 (Fig. lb). There
was a sharp rise in amylase activity as the cells entered the
stationary phase in nonrepressing medium. Amylase activity
was more than 10-fold higher than in an equivalent culture
grown in the presence of glucose. After 32 h of growth in
glucose-containing medium, amylase activity remained re-
pressed. When limiting amounts of glucose (<0.1%) were
added to the growth medium, amylase activity remained OD600
repressed until the glucose present in the medium was 10'
exhausted (as determined by the glucose test; see Materials w
and Methods), and then activity increased to nonrepressed
levels. The same results were obtained when LB and LB
plus 1%G were used as nonrepressing and repressing growth
media, respectively (data not shown).
To test whether this effect might simply be a copy number 20
effect, we measured the copy number of pSA33 in exponen- TIME (hours)
tially growing B. subtilis S0116 in CM, CM containing M%G, C.)
and LB media (as described in Materials and Methods). C.)
There were eight plasmid copies per genome equivalent in all
growth media, demonstrating that the low levels of amylase
activity in glucose-containing medium are not due to a
decrease in plasmid copy number. The 3.55-kb B. licheni-

Downloaded from https://journals.asm.org/journal/jb on 07 August 2024 by 125.23.119.78.

formis DNA fragment in pSA33 must therefore contain all
the cis sequences required for catabolite repression of amyL. OD600
The 5' end of the gene as well as regulatory and upstream
regions of the gene have been sequenced (34). The a- U)
amylase gene is preceded by a 393-base-pair (bp) open
reading frame flanked by possible regulatory sequences
(Ortlepp, Ph.D. thesis) (Fig. 2b). These sequences are con-
tained within the 3.55-kb fragment.
Catabolite repression of promoterless B. licheniformis a-
amylase gene in B. subtUis. B. subtilis S0122 carries a
subclone of pSA33, pBS86-3. This plasmid contains the 0 10 20 30 40
entire a-amylase structural gene and 30 bp upstream from TIME (hours)
the ATG initiation codon, but not its promoter (Fig. 2a). FIG. 1. Growth curves and corresponding oL-amylase activity
S0122 is also amylase positive. The expression of amyL is profiles of B. licheniformis FDO2 (a), B. subtilis S0113 (b), and B.
presumably activated by an upstream plasmid promoter. subtilis S0122 (c). Cultures were grown in CM (x) and CM plus
Amylase activity from S0122 was measured throughout the 1%G (+) media, and extracellular a-amylase activity was measured
growth cycle in cells grown in the presence or absence of throughout the growth cycle (see Materials and Methods). Symbols:
glucose. Synthesis of the enzyme was still strongly repressed L, specific activity in CM; *, specific activity in CM plus 1%G.
in the presence of glucose (Fig. lc). The deletion of the
upstream open reading frame sequences did not affect the found in S0116, but expression was still subject to catabolite
modulation of amyL expression in either a positive or repression.
negative manner. Thus, catabolite repression of amyL in B. O'Kane et al. (24) ligated random B. subtilis chromosomal
subtilis does not require its own promoter or the expression restriction fragments into sites upstream from the promoter-
of the upstream open reading frame. The level of amylase less amylase gene in pOK4, transformed S0113, and se-
activity was much lower (>50-fold) than the level of activity lected for a plasmid marker (Cmr). The transformants carried
2438 LAOIDE ET AL. J. BACTERIOL.

TABLE 1. oe-Amylase activities to B. subtilis S0113

a pBS86-3 transformants carrying an integrated B. licheniformis
promoterless amyL gene
A E S B P Ha Hindm
HindM So-Amylase sp act (U/mg per ml)"
I \NSII If/K Strain no.b
pBD6l CM CM + 1%G LB LB + 1%G
*4 promoterless amy L - 1 1.67 0.28
b (1.9kb) 2
3
0.76
6.55
0.11
0.41
TGATCATTUCTTGCGTGTCGTTAATCTCCTGATACGCITTTMCGmGACAGC 4 6.5 0.5
5 1.47 0.09
6 0.75 0.1
CTTGTCAATAACGGATGGTCCCAGGAATGGTTGCCGACTTCGTCCTTCCI 7 1.35 0.25
8 0.9 0.3
CAGCATCCG I I 1ACG1TTCGGGATMTATrGGGCTCTGCTTCCMGCACM 9 0.55 0.05
-35 -10 10 0.95 0.25
AGAAGGTCG 2 CCC U7C ATG CTC TGT AAA GCG mTT MT ATT TTA TTC "Stationary-phase culture supernatants from strains grown in nonrepress-
RBS
ing (CM or LB) media or in repressing (CM + 1%G or LB + 1%G) media were
assayed for a-amylase activity.
GTT GTA GCG GGA TTC GGA CCG TCA TCA AAT GTG AGG GCA ATC ACG b Randomly chosen o-amylase-positive integrants picked from LBS or
CMS plates.
TTT TTC ATC GGG ATT AAT m CGC TTG CTT CGG AAG CGG MC AGG CTC
CTG ATC AGT GAT TCC GTC CGC TCG CTT TCC AAT CTG MG GTT TCA TTG
integrating vector (pWD3; kind gift from H. Wood), and the
TGG GAT GTT GAT CCG GM GAT TGG MG TAC AAA MT MG CM MG ATU same results were obtained (data not shown).
GTC AAT CAT GTC ATG AGC CAT GCG GGA GAC GGA MA ATC GTC TTA Transcriptional start site of amyL. Plasmids pSA33 and
ATG CAC GAT ATT TAT GCA ACG TCC GCA GAT GCT GCT GM GAG ATT ATT pBS86-3 were linearized with PstI, labeled at the 5' end (Fig.
3), and used as probes for Si mapping to determine the in
AAA MG CTG AAA GCA AM GGC TAT CM TTG GTA ACT GTA TCT CAG CTT vivo start site of transcription of the amyL gene in S0116
GAA GAA GTG AAG MG CAG AGA GGC TAT TGA ATAAATGAGTAGAA&Q and of the promoterless gene in S0122, respectively. Total
CQ9TTTATATCQ3QITC1T1IGGMGMMTATAGGGAAATGGTACTTGTT RNA isolated from these strains and from the parental
strain, S0113 (amylase-negative control), was hybridized to
the appropriate denatured probe, and nuclease Si-resistant
AAAAAUCAGAATAUTTAAMIATCA TATGTTCACAUGAAAG2AMA hybrids of approximately 100 bp (SO113, Fig. 4a) and 600 bp
.10 RBS (SO112, Fig. 4b) were detected. A single band was protected
from each clone. The 100-bp protected fragment corre-
CAATC AOAAA CAA CAA AM CGG CTT TAC GCC CGA UTG CTG ACG sponds to the expected start site of amyL downstream from
CTG UTA mT GCG CTC ATC UTC TTG CTG CCT CAT TCT GCA GCA GCG GCG a c consensus sequence (34). The start site of transcription
E1i of the promoterless gene present on the plasmid pBS86-3 is
FIG. 2. (a) pBS86-3 (6.7 kb) carries the promoterless amyL gene approximately 600 bp upstream from the PstI site. Possible
inserted into the.EcoRI-HindIlI sites of pBD64 (PvuII site converted crA consensus sequences are present on the vector (pBD64)
to Hindlll). A, AccI; B, BamHI; E, EcoRI; P, PstI; and S, SmaI. (b) fragment 568 and 721 bp upstream from the PstI site, and

Downloaded from https://journals.asm.org/journal/jb on 07 August 2024 by 125.23.119.78.

Sequence of the 393-bp open reading frame present immediately readthrough from one of these promoters is therefore likely
upstream from the amyL gene and the 5' region of the amyL gene. to be responsible for the low level of activity found in the
The EiJA -10 and -35 consensus sequences and putative ribosome- promoterless construct. Promoter sequences recognized by
binding site (RBS) are underlined. A possible terminator IR se- other holoenzyme forms were not found from sequence
quence is shown with dotted underline. Asterisk (*) shows start of
the promoterless amyL sequence. The IR sequence is shown in analysis.
bold-face type. The ATG initiation codons are boxed. Regulation of amyL expression occurs at the level of tran-
scription. To determine whether regulation occurs at the
level of transcription, we grew B. subtilis S0113, S0116, and
the plasmid integrated in a site on the chromosome homol-
S0122 and B. licheniformis FDO2 in NSM in both the
presence and absence of excess glucose. At various times
ogous to the chromosomal fragment carried on the plasmid.
A total of 97 randomly chosen amylase-positive transfor-
throughout the growth cycle, total RNA was extracted from
mants were tested on LBS and LBS containing 0.5%G.
the bacterial cultures. Equal amounts of total RNA from
each sample were hybridized to 3'-end-labeled probes (Fig.
Qualitative differences in halo size between identical colo- 3) to determine the relative amounts of amyL-specific
nies were found for all insertions tested. All the haloes were mRNA which could protect the probe (present in excess)
significantly larger on nonrepressing LB medium compared from S1 digestion (Fig. 5) or when bound to nylon mem-
with those on glucose-containing medium (data not shown). branes would hybridize to excess labeled DNA (Fig. 6). The
To test this difference quantitatively, we measured amylase results showed that when a 3'-end-labeled probe was used,
activity of stationary-phase cultures of 10 randomly chosen the levels of S1-resistant mRNA and the amount of hybrid-
transformants in liquid cultures (Table 1). The activity of the ization to Northern blots correlated directly with the amy-
enzyme varied about sevenfold between the transformants,
lase activity profiles (Fig. 1).
presumably because they were being read from chromo-
somal promoters of different strengths which are at different DISCUSSION
distances from the gene. However, it was clear that in all
cases glucose exerted a repressive effect on amylase expres- We studied the expression of the B. licheniformis ox-
sion (Table 1). This experiment was repeated with a similar amylase gene, amyL, in its natural host and when cloned on
VOL. 171, 1989 CATABOLITE REPRESSION OF B. LICHENIFORMIS amyL 2439

pSA33

E E N' P C H E
I , I I I

4 amy L

1.05 kb 1.15 kb

pBS86-3

H N B P C H
I . . I I I

*4 promoterless amyL -.

1.09 kb
FIG. 3. The 5'-end-labeled and 3'-end-labeled DNA probes used for Si protection experiments and for Northern blots. The 5' probes were
labeled on the antisense strand at the recessed 5' PstI site by using the T4 DNA polymerase exonuclease activity to create a 5' overhang and
labeling with [-y-32P]ATP. The 3' probes were labeled on the antisense strand at the ClaI site by using E. coli Klenow fragment polymerase
activity and [a-32P]dCTP to fill in the recessed 3' end. B, BamHI; C, ClaI; E, EcoRI; H, Hindll; N', NdeI; N, NsiI; and P, PstI.

a low-copy-number vector, pSA33, in B. subtilis. We utilized A plasmid-borne promoterless construct of the amyL gene
this system in an attempt to understand the mechanism of (pBS86-3) was also subject to catabolite repression. From
catabolite repression in Bacillus species, a phenomenon that the size of the Si transcripts (Fig. 4), it appears that the gene
is not fully understood in either gram-positive or gram- is transcribed by an upstream plasmid promoter which has a
negative bacteria. The amyL gene, present in eight copies consensus CA recognition sequence. This suggested that
per B. subtilis chromosome, is subject to both temporal regulation of the gene occurs independently of its own
activation and catabolite repression. amyL shows very little promoter or that the plasmid promoter itself is also subject to
homology at the DNA level with the corresponding B. temporal activation and catabolite repression. To decide
subtilis gene, amyRI-amyE (34), which is also temporally between these two possibilities, we looked at the expression
activated and subject to catabolite repression (3, 20). a- of the promoterless gene when it was randomly integrated
Amylase production is repressed 10-fold in glucose-con- into the B. subtilis chromosome (24). Amylase-positive
taining cultures in the early stationary phase, similar to the transformants, which arise due to transcription of amyL by
levels of repression of the B. subtilis amyE gene (20; upstream chromosomal promoters, were tested, and all were
unpublished observations). This suggests that the plasmid- found to be subject to catabolite repression. Amylase activ-
borne heterologous gene is not titrating out a regulatory ity between the transformants varied (Table 1), presumably
factor present in B. subtilis cells. because the gene is being read by promoters of different

Downloaded from https://journals.asm.org/journal/jb on 07 August 2024 by 125.23.119.78.

strengths and/or because the gene is at various distances
from these promoters. In many cases, very low amounts of
a 1 2 3 amylase activity were detected, and it was not clear whether
the integrated promoterless gene was temporally activated in
the transformants. Many Bacillus promoters are switched on
at specific stages of the growth cycle by particular sigma
factors (12, 28), and the activation of these promoters may
overide any temporal activation of amyL in these integrants.
S2.4 What is clear is that all the integrants are subject to catabo-
lite repression. Glucose-mediated repression of the promot-
442 erless a-amylase gene when it is present on a multicopy
plasmid or when it is integrated in the B. subtilis genome
does not require its own promoter and occurs irrespective of
a b the distance between amyL and the promoter that is tran-
FIG. 4. Results of Si mapping experiment to determine the start scribing it. This suggests that the putative regulatory protein
sites of transcription of amyL (a) and of the promoterless gene in involved in mediating catabolite repression binds to a cis-
pBS86-3 (b). (a) A 30-p.g sample of total RNA isolated from S0116 acting site present on the promoterless amylase fragment,
was hybridized with a denatured 1.05-kb EcoRI-PstI DNA probe possibly at or adjacent to the transcriptional start site,
end labeled at the PstI site (Fig. 3). After digestion by Si nuclease, resulting in the attenuation of the amyL transcript in the
the protected hybrid was run on a 6% polyacrylamide gel (lane 2) absence of a repressing sugar. Alternatively, the regulatory
along with undigested probe (lane 1) and HpaII-digested pAT153 as
size standards. (b) A 50-p.g sample of total RNA isolated from protein contacts the RNA polymerase owing to looping or
S0122 was hybridized with a denatured 1.09-kb NsiI-PstI DNA bending of the DNA (17, 29) so that protein-protein interac-
probe end labeled at the PstI site (Fig. 3). The Si-protected fragment tion can occur even if the proteins are bound at widely
was run out on a 6% polyacrylamide gel (lane 3). Lane 1, Size separated sites. In the wild-type gene, the cis-acting se-
markers (pAT153/HpaII); lane 2, undigested probe. quence appears to be adjacent to the promoter (11 [accom-
2440 LAOIDE ET AL. J. BACTERIOL.

5 6 7

-i
-,.igi
26-z-&..

U_ S

2 3 4 5 6 7 8 .

..M

0 .:....:.

FIG. 5. Si protection experiments to determine the relative amounts of amyL-specific transcripts present in cultures grown in
nonrepressing and repressing media. (a) RNA isolated from stationary-phase cultures of S0122 (50 ,ug; lanes 3 and 4), FDO2 (80 ,ug; lanes 5
and 6), and S0116 (40 [Lg; lanes 9 and 10) grown in NSM (lanes 3, 5, and 9) and NSM plus 1%G (lanes 4, 6, and 10) (see Materials and Methods)
was hybridized with a 1.1-kb Clal-HindIll-denatured DNA probe labeled at the ClaI site (Fig. 3). A 50-,ug sample of RNA from S0113 grown
in NSM (lane 7) and 50 ,ug of tRNA (lane 8) were used as negative controls. Lane 2 shows the undigested denatured probe. Lanes 1 and 11
are pBR322/Taql size standards. The arrow shows the position of the protected hybrids. (b) Relative amounts of amyL-specific mRNA in
S0116 cultures entering the stationary phase (to) (lanes 4 and 7), in t, cultures (lanes 5 and 8), and in t2 5cultures (lanes 6 and 9) grown in NSM
(lanes 4, 5, and 6) and in NSM plus 1%G (lanes 7, 8, and 9). Total cellular RNA (40 ,ug) was hybridized with the 3'-end-labeled probe in each
case. TaqI-digested pBR322 was used as a size standard (lane 1). Undigested probe is in lane 2, and lane 3 contains tRNA.

panying report]), and in this case looping out of intervening

sequences may not be necessary. A third possibility is that

Downloaded from https://journals.asm.org/journal/jb on 07 August 2024 by 125.23.119.78.

the putative regulatory protein interacts in an indirect man-
ner with the RNA polymerase holoenzyme.
We searched for a possible cis-acting site downstream
from the promoters of catabolite-repressible B. subtilis
genes. The data are summarized in Table 2. There is a
candidate sequence, 5'-A/TTGTNA/T-3', in the vicinity of
the transcription initiation sites of the genes analyzed. In the - 2 32
case of amyL, this sequence is contained within an inverted
repeat (IR) sequence and is also present on the promoterless p.
amyL fragment. The IR overlaps with the probable start site
of transcription of amyL and also with the ATG translation
initiation codon (Fig. 2b). The IR sequence, 5'-TGTTTCAC-
3', is homologous to a consensus half site (5'-TNTNAN-3')
for DNA-binding proteins that utilize the cz-helix-turn-ot-
helix binding motif (26). An equivalent half-site consensus
sequence, 5'-TGTAAG-3', is present immediately down-
stream from the P1 promoter of the B. siubtilis amyRi locus, FIG. 6. Northern blot experiment to determine the amount
overlapping the transcriptional start site of the amyE gene amyL-specific transcripts present in S0116 grown in NSM (lanes 1,
(22). This motif is also present on an amyRI deletion 2, 3, and 4) and in NSM plus 1%G (lanes 5, 6, and 7). Total cellular
derivative which can still direct glucose-repressible expres- RNA (5 ,ug) from exponentially growing cells (60 Klett units; lane 1),
from cells entering the stationary phase (to; lanes 2 and 5), from
sion of a fused indicator cat gene (21). A point mutation early-stationary-phase cultures (tl; lanes 3 and 6), and from late-
(gra-10) in this region of dyad symmetry relieves catabolite stationary-phase cultures (t2.5; lanes 4 and 7) was run out on
repression of amyE (20, 22). formaldehyde-agarose gels, blotted onto nylon membranes (see
We determined that the start site of transcription of the Materials and Methods), and hybridized to a denatured, a-32P-
am-yL gene in pSA33 is approximately 110 bp from the PstI labeled arnyL Clal-HindII fragment (Fig. 3). Numbers to the right
site; there is a ua consensus sequence immediately upstream are size standards in kilobases.
VOL. 171, 1989 CATABOLITE REPRESSION OF B. LICHENIFORMIS amyL 2441

TABLE 2. Comparison of sequences downstream from the presence of glucose during growth on the enzyme activities of
promoters of B. subtilis genes subject to catabolite repression Escherichia coli: comparison of the effect with that produced by
fermentation acids. Biochem. J. 36:619-623.
Gene Sequencea Reference 7. Fujita, Y., and T. Fujita. 1986. Identification and nucleotide
amyL ATGTTT 34 sequence of the promoter region of the gluconate operon.
amyRI ATGTAA 39 Nucleic Acids Res. 14:1237-1252.
citB ATGTGA 5 8. Gryzan, T., S. Contente, and D. Dubnau. 1980. Molecular
gnt TTGTATA 7 cloning of heterologous chromosomal DNA by recombination
hut AGTTAATAGTTA 23 between a plasmid vector and a homologous resident plasmid in
sacC ATGTAC 16 Bacillus subtilis. Mol. Gen. Genet. 177:459-467.
sdh TTGTCA 18 9. Guidi-Rontani, C., A. Danchin, and A. Ulimann. 1980. Catabo-
lite repression in Escherichia coli mutants lacking cyclic AMP
Consensus sequence A/T T G T N A/T receptor protein. Proc. Natl. Acad. Sci. USA 77:5799-5801.
7 fi7 7 fi 10. Ide, M. 1971. Adenyl cyclase of bacteria. Arch. Biochem.
7 77 7 7 Biophys. 144:262-268.
11. Laoide, B. M., and D. J. McConnell. 1989. cis sequences
a
The known transcription start sites are shown in bold-face type (for amyL involved in modulating expression of Bacillus licheniformis
both possible sites are indicated). amyL in Bacillus subtilis: effect of sporulation mutations and
catabolite repression resistance mutations on expression. J.
Bacteriol. 171:2443-2450.
(34) (Fig. 2b). This agrees with the predicted start site of the 12. Losick, R., and J. Pero. 1981. Cascades of sigma factors. Cell
B. licheniformis 5A1 a-amylase gene (30). Either an A or a G 25:582-584.
13. Magasanik, B. 1%1. Catabolite repression. Cold Spring Harbor
nucleotide in an A+T-rich region immediately downstream Symp. Quant. Biol. 26:249-256.
from the oA -10 region 29 or 31 nucleotides, respectively, 14. Makman, R. S., and E. W. Sutherland. 1965. Adenosine 3',5'-
from the ATG translation start site is the most likely start phosphate in Escherichia coli. J. Biol. Chem. 240:1309-1314.
point of transcription. The start site of transcription of the B. 15. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular
subtilis amyE gene is 121 nucleotides upstream from the cloning, a laboratory manual. Cold Spring Harbor Laboratory,
ATG codon, which emphasizes the difference between the Cold Spring Harbor, N.Y.
two genes. We measured a-amylase mRNA levels by Si 16. Martin, I., M. Debarbouille, A. Klier, and G. Rapoport. 1989.
mapping and Northern blot experiments with 3'-end-labeled Induction and metabolite regulation of levanase synthesis in
probes (Fig. 3) in strains FDO2, S0113, S0116, and S0122 Bacillus subtilis. J. Bacteriol. 171:1885-1892.
grown in repressing and nonrepressing media (Fig. 5). The 17. Martin, K., L. Huo, and R. Scheif. 1986. The DNA loop model
for ara expression: AraC protein occupies the proposed loop
hybridization data showed clearly that in nonrepressed cul- sites in vivo and repression-negative mutants lie in these same
tures, amyL-specific mRNA is transcribed and the levels sites. Proc. Natl. Acad. Sci. USA 83:3654-3658.
increase throughout the growth phase in a pattern similar to 18. Melin, L., K. Magnusson, and L. Rutberg. 1987. Identification of
the observed a-amylase activity profiles (Fig. 6). In the the promoter of the Bacillus subtilis sdh operon. J. Bacteriol.
presence of glucose, amyL mRNA was barely detectable. 169:3232-3236.
We conclude that regulation of amyL occurs at the level of 19. Monod, J. 1947. The phenomenon of enzymatic adaptation.
transcription. Catabolite repression of the gene occurs inde- Growth Symp. XI:223-289.
pendently of its promoter and irrespective of the distance of 20. Nicholson, W. L., and G. H. Chambliss. 1985. Isolation and
the promoterless amyL gene from the promoter which acti- characterization of a cis-acting mutation conferring catabolite
vates it. repression resistance to a-amylase synthesis in Bacillus subtilis.
J. Bacteriol. 161:875-881.

Downloaded from https://journals.asm.org/journal/jb on 07 August 2024 by 125.23.119.78.

21. Nicholson, W. L., G. H. Chambliss, L. Buckbinder, N. P.
ACKNOWLEDGMENTS Ambulos, Jr., and P. S. Lovett. 1985. Isolation and expression of
a constitutive variant of the chloramphenicol-inducible plasmid
We are grateful to Frank Ollington of Biocon Ltd., Carrigaline, gene cat-86 under control of the Bacillus subtilis 168 amylase
County Cork, Ireland, for his interest in this project. We are grateful promoter. Gene 35:113-120.
to our colleagues in both Dublin and Madison, Wis., for critical 22. Nicholson, W. L., Y. Park, T. M. Henkin, M. Won, M. J.
reading of the manuscript. Weickert, J. A. Gaskell, and G. H. Chambliss. 1987. Catabolite
This project was supported by grants from the National Board of repression-resistant mutations of the Bacillus subtilis alpha-
Science and Technology and the Trinity College Trust. amylase promoter affect transcription levels and are in an
operator-like sequence. J. Mol. Biol. 198:609-618.
LITERATURE CITED 23. Oda, M., A. Sugishita, and K. Furukawa. 1988. Cloning and
1. Bernlohr, R. W., M. K. Maddock, and N. D. Goldberg. 1974. nucleotide sequences of histidase and regulatory genes in the
Cyclic guanosine 3':5' monophosphate in Escherichia coli and Bacillus subtilis hut operon and positive regulation of the
Bacillus licheniformis. J. Biol. Chem. 249:4329-4331. operon. J. Bacteriol. 170:3199-3205.
2. Birnboim, H., and J. Doly. 1979. A rapid alkaline extraction 24. O'Kane, C., M. A. Stephens, and D. J. McConnell. 1986.
procedure for screening recombinant plasmid DNA. Nucleic Integrable a-amylase plasmid for generating random transcrip-
Acids Res. 7:1513-1523. tional fusions in Bacillus subtilis. J. Bacteriol. 168:973-981.
3. Coleman, G., and M. A. Grant. 1966. Characteristics of alpha- 25. Ortlepp, S. A., J. F. Ollington, and D. J. McConnell. 1983.
amylase formation in Bacillus subtilis. Nature (London) 211: Molecular cloning in Bacillus subtilis of a Bacillus licheniformis
306-307. gene encoding a thermostable alpha amylase. Gene 23:267-276.
4. Dessein, A., M. Schwartz, and A. Ullmann. 1978. Catabolite 26. Pabo, C. O., and R. T. Sauer. 1984. Mechanism of autonomous
repression in Escherichia coli mutants lacking cyclic AMP. Mol. control of the Escherichia coli F plasmid; purification and
Gen. Genet. 162:83-87. characterization of the repE gene product. Annu. Rev. Bio-
5. Dingman, D. W., and A. L. Sonenshein. 1987. Purification of chem. 53:293-322.
aconitase from Bacillus subtilis and the correlation of its N- 27. Pastan, I., and S. Adhya. 1976. Cyclic adenosine 5'-monophos-
terminal amino acid sequence and the sequence of the citB gene. phate in Escherichia coli. Bacteriol. Rev. 40:527-551.
J. Bacteriol. 169:3062-3067. 28. Price, C. W., and R. H. Doi. 1985. Genetic mapping of rpoD
6. Epps, H. M. R., and E. F. Gale. 1942. The influence of the implicates the major sigma factor of Bacillus subtilis RNA
2442 LAOIDE ET AL. J. BACTERIOL.

polymerase in sporulation initiation. Mol. Gen. Genet. 201: 369-372.

88-95. 35. Ullmann, A., G. Contesse, M. Crepin, F. Gros, and J. Monod.
29. Ptashne, M. 1986. Gene regulation by proteins acting nearby and 1969. Cyclic AMP and catabolite repression in Escherichia coli,
at a distance. Nature (London) 322:697-701. p. 215-246. In T. W. Kall, M. Rothbell, and P. Condliffe (ed.),
30. Rothstein, D. M., P. E. Devlin, and R. L. Cate. 1986. Expression The role of adenyl cyclase and cyclic 3',5' AMP in biological
of ot-amylase in Bacillus licheniformis. J. Bacteriol. 168:839- systems Fogarty International Center Proceedings, no. 4. Na-
842. tional Institutes of Health, Bethesda, Md.
31. Schaeffer, P., J. Millet, and J.-P. Aubert. 1965. Catabolic 36. Ullmann, A., and A. Danchin. 1983. Role of cyclic AMP in
repression of bacterial sporulation. Proc. Natl. Acad. Sci. USA bacteria. Adv. Cyclic Nucleotide Res. 15:1-53.
54:704-711.
32. Scheer-Abramowitz, J., T. J. Gryczan, and D. Dubnau. 1981. 37. Vehmaanpera, J. O., and M. P. Korhola. 1986. Stability of the
Origin and mode of replication of plasmids pE194 and pUB110. recombinant plasmid carrying the Bacillus amyloliquefaciens
Plasmid 6:67-77. alpha amylase gene in B. subtilis. Appl. Microbiol. Biotechnol.
33. Smith, B. W., and J. H. Roe. 1949. A photometric method for 23:456-461.
the determination of alpha amylase in blood and urine with use 38. Wanner, B. L., R. Kodaira, and F. C. Neidhardt. 1978. Regula-
of the starch-iodine color. J. Biol. Chem. 179:53-59. tion of lac operon expression: reappraisal of the theory of
34. Stephens, M. A., S. A. Ortlepp, J. F. Ollington, and D. J. catabolite repression. J. Bacteriol. 136:947-954.
McConnell. 1984. Nucleotide sequence of the 5' regulatory 39. Yang, M., A. Galizzi, and D. Henner. 1983. Nucleotide sequence
region of the alpha amylase from Bacillus licheniformis: com- of the amylase gene from Bacillus subtilis. Nucleic Acids Res.
parison with the B. amyloliquefaciens gene. J. Bacteriol. 158: 11:237-249.

Downloaded from https://journals.asm.org/journal/jb on 07 August 2024 by 125.23.119.78.