![]()
|
|
||||||||
(Received for publication, December 23, 1994; and in revised form, June 5,
1995) From the
The gene encoding the
The basic features of the transcriptional machinery are
remarkably conserved in all organisms. In particular, the Although
the well characterized RNA polymerase from E. coli is
available to represent the Gram-negative lineage of eubacteria, there
has been less information regarding RNA polymerases from Gram-positive
bacteria. With the view that the transcriptional apparatus of a
genetically amenable Gram-positive bacterium could contribute to a
structure-function analysis of RNA polymerases, we began a study of the
genes encoding the core subunits from the spore-forming bacterium Bacillus subtilis. We earlier reported the isolation and
characterization of rpoA, the gene for the
Figure 1:
Genetic
organization of the B. subtilis
The promoter
activity manifested by pKB4 was further localized by subcloning
PCR-generated fragments into pDH32. pKB14 carried a fragment extending
from nt 829 to the SpeI site at nt 1101 ( Fig. 3and 5),
pKB15, a fragment from nt 935 to the SpeI site, and pKB16, a
fragment from nt 829 to 989. DNA sequencing confirmed that no
alterations had been introduced by the PCR reactions and that all three
fragments were oriented with the gene direction toward lacZ.
These three plasmids were then linearized and integrated into the amyE locus of strain PB2 to yield strains PB365, PB366, and
PB367.
Figure 3:
Nucleotide sequence of the orf23-rpoB (P23-
Figure 4:
Mapping the 5`-end of the rpoB message by primer extension. Wild type strain PB2 was grown in 2
Sequences necessary for the promoter activity manifested by
pKB16 were further defined by site-directed mutagenesis. A mutation in
the proposed -35 recognition sequence of the rpoB promoter carried by pKB16 was created using the primer
5`-GCAAAAAAAGTTAAACTCGGTATTTTAACTATG, the same as nt 829-861
except for the underlined residues, and extended to nt 989 using the
same second primer employed to amplify the pKB16 insert. DNA sequencing
confirmed that this mutagenized fragment was identical to that carried
by pKB16 except for the alteration of the proposed -35 sequence
from TTGACT to TAAACT. The mutagenized fragment was cloned into pDH32
to yield pKB17, which was linearized and integrated into the amyE locus of PB2 to yield strain PB368. Initial screening for
promoter activity of the integrated fusions was on tryptose blood agar
base plates (Difco Laboratories) containing
5-bromo-4-chloro-3-indolyl-
Figure 2:
Alignment of the primary sequences of the B. subtilis and E. coli
Rifampicin is an antibiotic that traps prokaryotic RNA polymerases
in the initial transcribing complex, preventing elongation of the
nascent transcript beyond three or four
nucleotides(40, 41) . All known mutations resulting in
rifampicin resistance map to the gene encoding the Conserved blocks 3 and 4 are
also thought to comprise part of the active center for On the basis of the strong
similarity of its predicted product to E. coli
To confirm the presence of a promoter within this 1.1-kb
fragment, we moved this region into the single-copy transcriptional
fusion vector pDH32(26) . Upon introduction of the resulting
fusion into the B. subtilis chromosome at the amyE locus, this fragment showed clear promoter activity (Fig. 1). To more precisely locate this activity, we first
subcloned two portions of the 1.1 kb-fragment into the pDH32 vector and
introduced these two fusions into the amyE locus. As shown in Fig. 1, the upstream 0.7-kb EcoRI-HindIII
fragment, which included the rplL-orf23 intercistronic region,
had no detectable promoter activity. In contrast, the downstream 0.4-kb HindIII-SpeI fragment, which included the orf23-rpoB intercistronic region, had strong promoter
activity. The nucleotide sequence of this region is shown in Fig. 3. Primer extension experiments (Fig. 4) located two
5`-ends of the rpoB message within the region. One signal was
centered near nt 876 (labeled A in Fig. 3, 4, and 5)
and the other near nucleotide 1012 (labeled B in Fig. 3-5). To establish whether the 5`-ends detected in
the primer extension experiment correlated with regions that contained
promoter activity, we made a series of additional transcriptional
fusions in the pDH32 vector, as shown in Fig. 5. From the
activities manifested by these fusions, we concluded that whereas the
5`-end centered around nt 876 lay near sequences conferring promoter
activity, the 5`-end centered around nt 1012 did not. Instead, it seems
likely that this latter 5`-end represented a site at which the rpoB message was processed. Because we detected this same signal near
nt 1012 using two different primers, we think it less likely that it
was an artifact of the extension reactions.
Figure 5:
Activity of lacZ reporter genes
transcriptionally fused with fragments from the P23-
Inspection of the region
necessary for promoter activity in the fusion experiments revealed the
sequence TTGACT-(17 base pairs)-TAATAT just upstream from
the 5`-end near nt 876 (Fig. 3). This sequence and spacing
closely match the consensus recognized by RNA polymerase holoenzyme
containing the major On the basis of promoter activity
measurements, primer extension experiments, and mutational analysis, we
conclude that a major rpoB promoter partly overlaps the
3`-coding region for the preceding orf23 gene. Our experiments
further suggest that the estimated 213-nt leader of the message
originating from this promoter is processed near nt 1012. In E.
coli, the rpoB message is processed at an RNaseIII site
located within a region of secondary structure in the 321-base pair rplL-rpoB interval(53, 54) . In B.
subtilis, there is no stable secondary structure apparent in the
sequence of the orf23-rpoB interval, and the mechanism of the
inferred processing remains to be established. Because there is no
obvious factor-independent terminator sequence separating orf23 and rpoB, transcription originating upstream from the
gene for r-protein L7/L12 may also contribute to rpoB expression. However, as shown by the plasmid integration
experiments, this upstream transcription is not required for normal
growth rate and sporulation frequency. Thus, the rpoB promoter
we identified is sufficiently strong to provide adequate levels of
We have identified rpoB, the gene encoding the The transcriptional organization of the B. subtilis rpoB region stands in sharp contrast to that of E. coli, in
which the primary transcripts originate from the promoters of the
upstream ribosomal protein operons L11 and
L10(54, 55, 56, 57) . In E.
coli, differential regulation of r-protein and RNA polymerase
subunit expression results partly from the action of a transcriptional
attenuator that lies between the gene for r-protein L7/L12 and rpoB(53, 58) and partly from specific translational
feedback mechanisms operating on the r-protein and polymerase subunit
messages(59, 60, 61, 62) . We do not
yet know whether B. subtilis rpoB expression is subject to the
same translational regulation as has been proposed for the enteric
system. However, because B. subtilis has a strong promoter
immediately upstream from rpoB, for which no counterpart
exists in E. coli, it is possible that differential regulation
of r-protein and RNA polymerase gene expression could be accomplished
largely at the transcriptional level. The E. coli
In E. coli, this 69-residue
segment contains the paf32 alteration (39) and is
therefore thought to define part of a contact site with the Alc
protein, a site-specific termination factor encoded by bacteriophage T4
which acts as a block to the transcription of host
genes(63, 64) . On the basis of these results,
Severinov et al.(39) proposed that non-essential
regulatory proteins specific to the E. coli system may have
evolved to target this dispensable region. However, the conservation of
the 69-residue segment within variable region 1 of B. subtilis This notion is supported by work of Landick et al.(13) , who identified a series of substitutions
in E. coli
The nucleotide
sequence(s) reported in this paper has been submitted to the
GenBank[GenBank].
Volume 270,
Number 35,
Issue of September 01, pp. 20329-20336, 1995
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
Subunit of Bacillus subtilis RNA Polymerase (*)
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
subunit of Bacillus subtilis RNA polymerase was isolated from a gt11 expression library
using an antibody probe. Gene identity was confirmed by the similarity
of its predicted product to the Escherichia coli
subunit
and by mapping an alteration conferring rifampicin resistance within
the conserved rif coding region. Including the rif region, four colinear blocks of sequence similarity were shared
between the B. subtilis and E. coli
subunits.
In E. coli, these conserved blocks are separated by three
regions that either were not conserved or were entirely absent from the B. subtilis protein. The B. subtilis
gene was
part of a cluster with the order rplL (encoding ribosomal
protein L7/L12), orf23 (encoding a 22,513-dalton protein that
is apparently essential for growth), rpoB (
), and rpoC (
`). This organization differs from the
corresponding region in E. coli by the inclusion of orf23. Experiments using promoter probe vectors and
site-directed mutagenesis located a major rpoB promoter
overlapping the 3`-coding region of orf23, 250 nucleotides
upstream from the
initiation codon. Thus, the B. subtilis
rpoB region differs from its E. coli counterpart in both
genetic and transcriptional organization.
,
`, and
subunits that comprise the catalytic core of the
eubacterial RNA polymerase are similar to subunits of the three nuclear
RNA polymerases of eukaryotes (see (1) and (2) for
reviews). The
` subunit of Escherichia coli shares eight
regions of sequence similarity with the largest subunit of the
eukaryotic enzymes(3) , and the
subunit of E. coli was initially reported to share nine regions with the second
largest subunit(4, 5) . Additional sequence alignment
of
homologues has further refined the boundaries of these nine
regions into 12 colinear segments, which are conserved in both
eubacteria and eukaryotes(6) .
, the third largest subunit
of the eubacterial core enzyme, also has similarity to the third
subunit of RNA polymerase II and to the fourth subunits of RNA
polymerases I and III, although this similarity is less striking than
that found among the
or
` homologues (reviewed in (1) and (2) ). This conservation of primary amino acid
sequence suggests common functions for the shared regions.
subunit(7, 8) . Here, we describe the genetic and
transcriptional organization of the region containing rpoB,
the gene encoding the
subunit, and show that this organization
differs substantially from that of the corresponding region in E.
coli. Because
is involved in most of the catalytic functions
of RNA polymerase, including nucleotide
binding(9, 10) , transcription initiation, elongation,
and
termination(10, 11, 12, 13, 14, 15) ,
and interactions with both the subunit (16, 17) and the NusA
protein(18, 19, 20) , we also compared the
likely functional domains of the B. subtilis and E. coli subunits. This comparison revealed two regions of E. coli
that were entirely absent from the B. subtilis protein and thus are likely dispensable for function.
Significantly, another region of 186 residues, which is known to be
dispensable in E. coli, shared little overall sequence
similarity with B. subtilis
but nonetheless contained a
common, 69-residue segment. We infer from these results that the common
segment has a more fundamental role in
function than previously
believed.
Bacterial Strains, Phage, and Genetic
Methods
E. coli Y1090 was used as host for the
gt11 expression vector, grown as previously
described(21) . B. subtilis strains used are shown in Table 1. For strain constructions, B. subtilis was made
competent for natural transformation as described by Dubnau and
Davidoff-Abelson (23) . Standard recombinant DNA methods were
performed as described by Davis et al.(24) ,
polymerase chain reactions were done according to published
protocols(25) , and DNA sequencing was done by the
dideoxynucleotide chain termination method with reactions primed on
double-stranded DNA templates, using the Sequenase enzyme and protocols
from U. S. Biochemical Corp. We sequenced 6130 base pairs of the rpoB region on both strands, using either nested deletions,
made as previously described(8) , or custom oligonucleotide
primers (Operon Technologies, Alameda, CA) as necessary.
Location of the Alteration Caused by the rfm2103
Allele
A fragment of rpoB bearing strong sequence
similarity to the rif region of E. coli rpoB was
amplified from the chromosome of B. subtilis strain PB355 (rfm2103), together with a control fragment from the PB2 wild
type strain. These PCR (
)products were cut at the PvuII and StyI sites, which flank the rif region, yielding 817 nt fragments that were subcloned into pUC19
for further analysis. The resulting plasmids were linearized and
transformed into PB2 with selection for resistance to rifampicin (200
µg/ml). The fragment amplified from strain PB355 conferred
rifampicin resistance at a frequency 7-fold higher than the appearance
of spontaneous resistance, determined from transformations using the
control fragment isolated from strain PB2. The rfm2103 alteration was confirmed by directly sequencing the rif chromosomal region of PB300, the parent of PB355, using the fmol system (Promega).
Construction of Transcriptional
Fusions
Promoter activity was detected using single-copy
transcriptional fusions to the lacZ reporter gene of the pDH32
vector(26) . As shown in Fig. 1, we first used three
fragments to locate activity in the L7/L12-P23-
interval. pKB3
contained a 1.1-kb EcoRI-SpeI fragment that spanned
the entire interval. pKB13 contained the upstream portion of this
interval on a 0.7-kb EcoRI-HindIII fragment, and pKB4
contained the downstream portion on a 0.4-kb HindIII-SpeI fragment. All three plasmids were
linearized and transformed into B. subtilis strain PB2,
whereupon they integrated into the amyE chromosomal locus by
means of the amy ``front'' and ``back''
sequences carried by the pDH32 vector(26) .
-
` region. The
chromosome in the rpoB region is represented by the heavyline and kilobase scale. The rectanglesabove the physical map indicate the open reading frames
encoding ribosomal protein L7/L12, the 22,513-dalton protein P23, and
the
and
` subunits of RNA polymerase, all of which are
transcribed from left to right; the arrows adjacent to the L7/L12 and
` reading frames indicate that
they extend beyond the cloned region. The restriction map shows the
sites used in the recombinant DNA constructions described under
``Materials and Methods.'' (R), EcoRI
linkers derived from the construction of the gt11
library(7) ; H, HindIII; S, SpeI; P, PvuII; and Y, StyI. The regions of the chromosome used in some of the
constructions are indicated by the horizontallinesbeneath the restriction map. pKB10 was made by subcloning
the indicated 1.1-kb EcoRI-SpeI fragment into the
pCP115 integration vector(27) . Upon integration into the B. subtilis chromosome, pKB10 would prevent any transcription
initiating upstream of the L7/L12 gene from entering the
and
` genes. pKB3, pKB4, and pKB13 were made by sucloning the
indicated fragments into the single-copy transcriptional fusion vector
pDH32 (26) . The two fragments conferring promoter activity are
labeled +. The physical map of the
-
` region was derived
from the restriction map and DNA sequence of the chromosomal inserts
carried by the two gt11 phages.
) interval. The sequence shown represents the C-terminal
coding region for P23, the P23-
intercistronic region, and the
N-terminal coding region for
. This includes the 0.4-kb HindIII-SpeI fragment that contains rpoB promoter activity (Fig. 1). The two 5`-ends of rpoB message detected in the primer extension experiment described in Fig. 4are indicated by the verticallines marked A at nt 876 and B at nt 1012. The
locations of the three complementary primers used to analyze the rpoB message are indicated by <P973, <P1065, and
<P1150, with (<) denoting the 3`-end of the primer. The proposed
-35 and -10 recognition sequences of the rpoB promoter are doubleunderlined at nt
840-845 and 863-868, respectively; the indicated mutations
change the -35 sequence from TTGACT to TAAACT in pKB17. The
proposed ribosomal binding site for the rpoB message is underlined at nt 1076-1083.
SG sporulation medium and harvested during logarithmic growth;
the RNA was then extracted. Primer extensions were done using a molar
excess of three different synthetic primers, P973, P1065, and P1150
(described under ``Materials and Methods''). These primers
are complementary to message synthesized using the orf23-rpoB (P23-
) intercistronic region as template; their locations are
shown in Fig. 3. For each experiment, samples containing 75
µg of RNA were loaded onto lane1 of a sequencing
gel. A sequencing ladder was run in parallel using the same primer
employed in the extension reactions; the letters A, C, G, and T indicate the dideoxynucleotide
used to terminate the reaction. The sequences indicated on the right are from the non-transcribed strand and are the
complement of the sequences that can be read from the ladder. Reactions
using primers P973, P1065, and P1150 all gave a 5`-signal centered
around the thymidine complementary to adenosine 876 (Fig. 3);
the experiment using primer P1065 is shown in panel A.
Reactions using primers P1065 and P1150 both gave a 5`-signal centered
around the thymidine complementary to adenosine 1012 (Fig. 3);
the experiment using primer P1150 is shown in panel
B.
-D-galactoside. For
quantitative estimates of promoter activity, we performed
-galactosidase assays essentially as described by
Miller(28) . B. subtilis cells were grown to late
logarithmic stage in 2 SG sporulation medium (29) and
then diluted 1:25 into fresh medium. Samples were taken throughout the
logarithmic and stationary phases of growth. Activity was expressed in
Miller units, defined as 1000
A/min/ml/unit of optical density at 600 nm.
Mapping the 5`-Ends of rpoB Message
RNA
was prepared essentially by the method of Igo and Losick(30) ,
using the modifications described by Varón et
al.(31) . Strain PB2 was grown in 50 ml of 2 SG
sporulation medium and harvested during exponential growth. RNA was
extracted as described (30, 31) and precipitated by
overnight incubation at -20 °C with 2 volumes of ethanol. For
primer extension reactions, three different oligonucleotide primers
were used to analyze the orf23-rpoB intercistronic region:
P973, a 17-mer (5`-TTAAGAAAACCACATCC-3`) complementary to nt
973-989 in Fig. 3; P1065, a 21-mer
(5`-ATTCACCCCTCAAATCATGCG-3`) complementary to nt 1065-1085; and
P1150, a 19-mer (5`-GGTAATTCTAACACTTCGC-3`) complementary to nt
1150-1168. Each primer was 5`-end labeled with
[
-
P]ATP (3000 Ci/mmol, Amersham Corp.) and
T4 polynucleotide kinase (Promega). Annealing and primer extension were
done using the Promega primer extension kit according to the
manufacturer's instructions, except that 75 µg of RNA and 5
ng of primer were used in a 20-µl reaction volume.
Computer Analysis
The statistical
significance of protein sequence comparisons was evaluated with the
FASTA and RDF programs of Pearson and Lipman(32) , using the
National Biomedical Research Foundation data bases. Highly related
sequences usually have an optimized alignment score exceeding 100 and a z value greater than 10 standard deviations above the mean
alignment of a shuffled sequence.
Isolation of
Antibody raised against B. subtilis RNA polymerase holoenzyme was used to screen gt11 Bacteriophages Encoding the
Subunitgt11 libraries
for phages that might carry core subunit genes(21) . From a
random library constructed with chromosomal DNA cut with AluI, HaeIII, or RsaI(7) , we found nine positive
clones, one of which was subsequently shown to encode the alternative
transcription factor
(21) . Among the
remaining eight phages, epitope selection (33) identified five
that might encode either of the two largest core subunits,
or
`. The restriction maps of the five presumptive
and
`
clones indicated that they fell into two classes and carried
overlapping regions of the B. subtilis chromosome. These two
classes are represented by the phages gt11-15 and
gt11-17, shown in Fig. 1. Subsequent characterization
established that the chromosomal inserts carried by these phages
contained the entire
coding region.Genetic Organization of the
We determined the DNA sequence on both strands of
the 6.1-kb region isolated on the
-
`
Regiongt11 clones. We then identified
the genes within this region by aligning the predicted B. subtilis gene products with their E. coli counterparts, using the
FASTA program of Pearson and Lipman(32) . As shown in Fig. 1, we found that B. subtilis had a gene order
similar to E. coli from the rplL homologue (encoding
ribosomal protein L7/L12) through the rpoC homologue (encoding
the
` subunit), confirming the results of previous genetic mapping
experiments(34) . However, our analysis revealed that the gene
order in B. subtilis differed from E. coli by the
inclusion of an open reading frame (orf23) that could code for
a protein of 22,513 daltons (P23). The predicted sequence of P23 was
significantly similar to the products of two unidentified open reading
frames sequenced as part of the E. coli genome project
directed by F. R. Blattner (University of Wisconsin, Madison). (
)The first of these (GenBank accession number U14003 283)
is located at 99 min on the E. coli chromosome and encodes a
predicted product of 343 residues (optimized alignment score with P23
is 217; z value is 26.1 S.D. above the mean). The second
(GenBank accession number U18997 12) is located at 76 min and encodes a
predicted product of 388 residues (optimized alignment score is 214; z value is 24.5). Both E. coli products also
significantly resemble each other (optimized alignment score is 282; z value is 29.9), and both reading frames lie distant from the E. coli rpoB region at 88 min. Notably, our attempts to
disrupt B. subtilis orf23 by plasmid integration failed to
yield viable transformants (data not shown). Because we demonstrate
below that B. subtilis rpoB is expressed from a promoter
immediately downstream from orf23, an orf23 disruption would not significantly affect the integrity of the rpoB transcriptional unit. Thus, the observation that orf23 could not be readily disrupted suggests that it encodes
an essential product.Identification of B. subtilis rpoB
The
proposed
reading frame encodes a predicted 1,193-residue product
that is 56% identical to the E. coli
subunit (Fig. 2). The two proteins share their highest sequence
conservation in four large blocks, and these are separated by three
regions that are either variable or entirely absent in B. subtilis
(see Table 2). Within the four large blocks, the 12
segments that are highly conserved among the second largest subunits of
eubacterial and eukaryotic RNA polymerases (6) are found in the
expected order (see Fig. 2). Moreover, individual residues that
are known to be important for E. coli
function are found
in the corresponding locations in the B. subtilis protein. As
shown in Fig. 2, these include two key residues that lie near
the active center that binds the transcript-initiating nucleotide (10) , as well as 15 of the 16 residues at which single
substitutions confer rifampicin
resistance(12, 13, 18, 37, 38) .
subunits. The predicted
sequence of B. subtilis
(upper) is from this
work and that of E. coli
(lower) is from
Ovchinnikov et al.(35) . Each of the four conserved
blocks was separately aligned by means of the FASTA program of Pearson
and Lipman(32) , with identical residues indicated by a colon (:) and conserved substitutions (36) by a period (.). The underlinedsegments denote
the 12 regions that are highly conserved in the second largest subunits
of RNA polymerases from eubacteria and eukaryotes(6) . The four
conserved blocks are separated by three variable regions that either
are not conserved or are absent entirely from the B. subtilis protein. Key residues of E. coli
that have been
altered by mutation are shown below the E. coli sequence. These include residues at which single substitutions
cause either rifampicin resistance (*) or an altered termination
phenotype (ˆ) (see (12) , (13) , (18) , (37) , (38) ). Arrows indicate one particular
substitution that confers rifampicin resistance on both organisms; this
is the site of the E. coli rpoB2 alteration (H526Y) (18) and the corresponding B. subtilis rfm2103 alteration (H482Y) (this work). Other arrows indicate E. coli substitutions, which render RNA polymerase insensitive
to the Alc termination factor of bacteriophage T4 (R368H (paf32)(39) ) or which alter promoter clearance
capabilities of the mutant polymerase (K1065R (12) and
H1237A(10) ). Affinity labeling studies have shown that
Lys-1065 and His-1237 lie near the binding site for the
transcript-initiating nucleotide(10) .
subunit(42, 43, 44) . In E. coli
, most alterations conferring rifampicin resistance lie
between residues 512 and 573 in conserved block
2(12, 13, 18, 37, 38) . We
used PCR to isolate this region from the chromosome of B. subtilis strain PB355 (rfm2103) and found that it contained only
one alteration from the wild type sequence: a C to T transition that
would change residue 482 of B. subtilis
from a histidine
to a tyrosine. Transformation of strain PB2 with the fragment bearing
the rfm2103 allele conferred the ability to grow in the
presence of high levels of rifampicin, indicating that the H482Y
alteration alone is sufficient for resistance. As shown in Fig. 2, this is the identical change in primary sequence caused
by the E. coli rpoB2 allele(18) , which elicits severe
defects in transcription termination and is incompatible with the rho-15, nusA10, nusA11, and dnaA46 mutations (11, 19, 45, 46) .
Indeed, most rifampicin-resistant mutations are highly pleiotropic,
causing altered initiation, elongation, and termination
phenotypes(9, 11, 15, 19, 46, 47, 48) .
Because rifampicin-resistant mutants in conserved block 1 have the same in vivo and in vitro phenotypes as mutants in
conserved block 2, Jin and Gross (18) and Severinov et al.(49) have suggested that these two blocks perform a common
catalytic function in the core enzyme.
catalytic
function. Affinity labeling studies indicate that both Lys-1065 and
His-1237 of E. coli
lie near the site that binds the
transcript-initiating nucleotide, although neither residue appears to
be directly involved in subsequent phosphodiester bond
formation(10) . Notably, Lys-1065 and His-1237 are conserved in
all prokaryotic and eukaryotic
homologues characterized to
date(10, 12) , and these residues are found at the
expected positions in B. subtilis
(Fig. 2).
Genetic evidence underscores the importance of these residues in
function. Alteration of E. coli Lys-1065 to arginine (K1065R)
results in a dominant lethal mutant enzyme that is blocked in promoter
clearance(12) , and alteration of His-1237 to alanine (H1237A)
also results in a mutant polymerase with impaired promoter
clearance(10) . Other single residues surrounding E. coli Lys-1065 are also important for
function, because the
R1069A, G1071A, and K1073A substitutions result in dominant lethal
mutants with decreased promoter clearance capabilities and aberrant
response to pause sites(50) . All three of these residues are
conserved in B. subtilis.
, we
conclude that we have isolated the gene encoding the B. subtilis
subunit. This conclusion is reinforced by two independent
criteria. First, the initial antibody screening and epitope selection
indicated that the gt11 clones encoded a protein with the
antigenic properties of B. subtilis
. Second, as an
additional biological criterion, we mapped within the B. subtilis
coding sequence an alteration conferring rifampicin
resistance. In keeping with the standard genetic nomenclature, we will
refer to the gene for the
subunit as rpoB.Transcriptional Organization of the
We used a combination of plasmid integration
experiments, primer extension studies, and fusion constructions to
determine the location of both the rpoB promoter and a
possible site of processing for the rpoB message. Plasmid
integration (51) allowed us to determine that the cloned region
contained a promoter capable of sustaining normal cell growth. We first
constructed the integration plasmid pKB10, which contained the 1.1-kb EcoRI-SpeI fragment from the rpoB region
shown in Fig. 1, then transformed B. subtilis wild-type
strain PB2 with selection for the chloramphenicol resistance encoded by
the plasmid. In all the resulting transformants tested, PCR confirmed
that pKB10 had in fact integrated into the rpoB region of the
genome via a single homologous crossover event, leading to a tandem
duplication of the 1.1-kb EcoRI-SpeI fragment
separated by plasmid sequences (data not shown). Such an integration
event would prevent any transcription that originates upstream of rplL from entering the rpoB gene(8, 31, 51) . Because many viable
transformants were obtained from this experiment and because the
transformants that carried the integrated pKB10 plasmid in the expected
configuration were indistinguishable from wild type cells in terms of
growth rate and sporulation frequency (data not shown), we conclude
that the 1.1-kb EcoRI-SpeI fragment contains a
promoter that can supply the cells with a functional level of
-
`
Region
protein.
interval. Panel A depicts the HindIII-SpeI fragment
that contains rpoB promoter activity (Fig. 1). The
C-terminal coding region of P23 and the N-terminal coding region of
are indicated by the rectanglesabove the
nucleotide scale. The locations of the 5`-ends of rpoB message
detected in the primer extension experiments of Fig. 4are shown
by the arrows, with the signal centered at nt 876 marked A and the signal centered at nt 1012 marked B. The
fragments shown beneath the nucleotide scale were subcloned into the
transcriptional fusion vector pDH32 and integrated in single copy at
the amyE loci of the indicated strains. The fragments carried
by strains PB367 and PB368 are identical except for the mutation of the
proposed -35 recognition sequence (TTGACT to TAAACT), indicated
by the filledtriangle. Panel B shows the
-galactosidase activities directed by the four transcriptional
fusions depicted in panelA. The strains carrying
these fusions were grown and assayed for
-galactosidase activity
as described under ``Materials and Methods.'' The fusion
carried by strain PB367 had strong promoter activity, and this activity
was abolished by the -35 mutation in the otherwise identical
fusion carried by strain PB368. We therefore conclude that the A signal centered at nt 876 represents the likely site of initiation
for the rpoB message. In contrast, the fusion carried by
strain PB366 had no promoter activity, suggesting that the B signal represents a processing site.
factor of B. subtilis,
(52) . To determine whether this sequence was
required for rpoB promoter activity, we made a double mutation
in the proposed -35 recognition sequence, changing it from TTGACT
to TAAACT (Fig. 3). When the fragment containing this mutation
was fused to the lacZ reporter gene of the pDH32 vector and
integrated into the chromosome, promoter activity was completely
abolished (Fig. 5).
subunit under most growth conditions.
subunit of B. subtilis RNA polymerase. The evidence for this
identification includes (i) the antigenic properties of the rpoB product, (ii) the high similarity of the predicted rpoB product with E. coli
, and (iii) the presence within
the rpoB coding region of an alteration conferring rifampicin
resistance. We have further shown that B. subtilis rpoB is
transcribed from a promoter that overlaps the 3`-end of the preceding orf23 coding region and that this promoter is sufficiently
active to support wild type growth rate and sporulation frequency.
subunit plays a key role in transcriptional initiation, elongation, and
termination. What information regarding structure-function
relationships can be derived from comparison of the predicted sequences
of B. subtilis and E. coli
? As shown in Fig. 2, the two proteins share highest sequence conservation in
four large blocks. This degree of conservation between such highly
divergent organisms suggests that the four large blocks mediate common
functions, and there is ample biochemical and genetic evidence that key
residues within these blocks are critical for
function in E.
coli. However, it is the differences between the two
subunits that provide our most significant results. On the one hand,
the four conserved blocks are separated by three regions that are
either variable or absent in B. subtilis
. On the other
hand, we find that the B. subtilis protein has a C-terminal,
46-residue extension of the fourth conserved block that is not present
in E. coli. These differences appear to be characteristic of
the Gram-positive and Gram-negative lineages of eubacteria.Variable Region 1 (Amino Acid Residues 211-368 in B.
subtilis and 224-409 in E. coli)
The deletions
(166-328) and
(186-433) created within E.
coli
by Severinov et al.(39) had no
obvious effect on
function in vitro, and the
(166-328) alteration was non-lethal in vivo,
leading these authors to conclude that much of this region was
dispensable for minimal
function. As shown in Fig. 2,
there is little overall conservation between these regions in E.
coli and B. subtilis
, and B. subtilis variable region 1 also has 43 fewer residues, supporting the
notion that this region is not critical for function. However, as first
pointed out to us by Benjamin Hall, this region does contain a common,
69-residue segment that is located near the C-terminal part of variable
region 1 in E. coli (corresponding to residues 339-409)
and near the N-terminal part in B. subtilis (corresponding to
residues 211-279). This common segment shares 48% identity
between the two organisms.
, albeit at a different location, suggests that the segment
is widespread among prokaryotes and therefore has a more fundamental
role in
function.
that affects transcription termination in
vivo. As shown in Fig. 2, four of these substitutions map
within the common 69-residue segment, and a fifth maps in the region
immediately adjacent in the E. coli protein. Furthermore, we
note that the common segment is also conserved in the
subunits of Pseudomonas putida(65) , Buchnera
aphidicola(66) , Mycobacterium
leprae(67) , and Mycobacterium
tuberculosis(44) , lending further support to the
suggestion of a more universal role. In Pseudomonas and Buchnera, the
segment lies toward the C-terminal portion of variable region 1, as is
the case in E. coli. In contrast, the mycobacterial segment
lies toward the N-terminal portion of the region, as is the case in B. subtilis. Thus, it appears that the difference in location
of this segment within variable region 1 reflects a difference in
domain organization between Gram-positive and Gram-negative bacteria.
The importance of this common, 69-residue segment may also extend to
the eukaryotic lineage. Shaaban et al.(6) found that
alterations mapping in or near the corresponding region of the second
largest subunit of yeast RNA polymerase III also affected termination
properties of the enzyme.Variable Region 2 (Amino Acid Residues 897-898
in B. subtilis and 939-1038 in E. coli)
The phenotypes of
deletion and insertion mutations have shown that amino acid residues
940-1040 are dispensable for normal
function in E.
coli(68, 69, 70) . This region is
entirely missing from B. subtilis
, as well as from the
homologous RNA polymerase subunits from mycobacteria, chloroplasts,
eukaryotes, and archaebacteria(44, 67, 69) ,
supporting the notion that these residues are not important for
function. Because variable region 2 is retained but not highly
conserved in the
subunits of P. putida and B.
aphidicola(65, 66) , the possibility remains that
some of these residues mediate a dispensable function unique to
Gram-negative bacteria.Variable Region 3 (Amino Acid Residues 984-985
in B. subtilis and 1126-1179 in E. coli)
The complete
absence of these 54 amino acid residues in the B. subtilis protein suggests that this region is not critical for
function. This view is supported by the complete absence of the
corresponding region in the
homologues of chloroplasts and
eukaryotes (6, 71) and its significant reduction in
the
subunits of mycobacteria(44, 67) . However,
as was the case for variable region 2, variable region 3 is fully
retained but not highly conserved in other Gram-negative bacteria (65, 66) .C-terminal Extension of Conserved Block
4
In vitro reconstitution studies (16, 17) suggest that sequences within conserved block
4 (residues 985-1144 in B. subtilis and 1180-1339 in E. coli) are necessary for holoenzyme formation. When the
C-terminal portion of E. coli
was truncated by as few as
42 residues, it retained the ability to associate with the
` and
subunits to form complexes of the size expected for RNA
polymerase core enzyme. However, these complexes did not include the
subunit, which confers promoter recognition specificity on the
core(52) , leading to the proposal that the C-terminal region
of E. coli
is important for binding
(16, 17) . In B. subtilis
, the
corresponding C-terminal region is highly conserved but extends an
additional 46 residues beyond the terminus of the E. coli protein (see Fig. 2). This 46-residue, C-terminal extension
is highly charged, with 21 acidic residues and 4 basic residues.
Notably, the C-terminal regions of the
subunits from M.
leprae(67) and M. tuberculosis(44) also
extend an additional 38 residues beyond the terminus of the E. coli protein, and these mycobacterial extensions share 20% identity
with the B. subtilis extension (data not shown). Thus, a
C-terminal extension of
is likely characteristic of the
Gram-positive lineage rather than specific to B. subtilis.
Whether this C-terminal extension has any influence on the binding of
factors to core has yet to be determined.
)
)
We thank Benjamin Hall and Konstantin Severinov for
helpful discussions and for communicating results prior to publication,
David Dubnau for providing strain DR1010, Carl Batt for the gift of
oligonucleotide primers, and Susan Thomas for technical assistance.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
![]()
CiteULike
Complore
Connotea
Del.icio.us
Digg
Reddit
Technorati What's this?
This article has been cited by other articles:
![]() |
A. E. Perkins and W. L. Nicholson Uncovering New Metabolic Capabilities of Bacillus subtilis Using Phenotype Profiling of Rifampin-Resistant rpoB Mutants J. Bacteriol., February 1, 2008; 190(3): 807 - 814. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. P. Rooney, J. L. Swezey, R. Friedman, D. W. Hecht, and C. W. Maddox Analysis of Core Housekeeping and Virulence Genes Reveals Cryptic Lineages of Clostridium perfringens That Are Associated With Distinct Disease Presentations Genetics, April 1, 2006; 172(4): 2081 - 2092. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y.-J. Yun, K.-H. Lee, L. Haihua, Y.-J. Ryu, B.-J. Kim, Y.-H. Lee, G.-H. Baek, H.-J. Kim, M.-S. Chung, M.-C. Lee, et al. Detection and Identification of Mycobacterium tuberculosis in Joint Biopsy Specimens by rpoB PCR Cloning and Sequencing J. Clin. Microbiol., January 1, 2005; 43(1): 174 - 178. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Maughan, B. Galeano, and W. L. Nicholson Novel rpoB Mutations Conferring Rifampin Resistance on Bacillus subtilis: Global Effects on Growth, Competence, Sporulation, and Germination J. Bacteriol., April 15, 2004; 186(8): 2481 - 2486. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Adekambi, P. Colson, and M. Drancourt rpoB-Based Identification of Nonpigmented and Late-Pigmenting Rapidly Growing Mycobacteria J. Clin. Microbiol., December 1, 2003; 41(12): 5699 - 5708. [Abstract] [Full Text] [PDF] |
||||
![]() |
C.-Y. Lim, K.-H. Lee, M.-J. Cho, M.-W. Chang, S.-Y. Kim, N.-H. Myong, W.-K. Lee, K.-H. Rhee, and Y.-H. Kook Detection of Helicobacter pylori in Gastric Mucosa of Patients with Gastroduodenal Diseases by PCR-Restriction Analysis Using the RNA Polymerase Gene (rpoB) J. Clin. Microbiol., July 1, 2003; 41(7): 3387 - 3391. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. L. Nicholson and H. Maughan The Spectrum of Spontaneous Rifampin Resistance Mutations in the rpoB Gene of Bacillussubtilis 168 Spores Differs from That of Vegetative Cells and Resembles That of Mycobacterium tuberculosis J. Bacteriol., September 1, 2002; 184(17): 4936 - 4940. [Abstract] [Full Text] [PDF] |
||||
![]() |