Previous Article | Next Article ![]()
Applied and Environmental Microbiology, November 2003, p. 6888-6898, Vol. 69, No. 11
0099-2240/03/$08.00+0 DOI: 10.1128/AEM.69.11.6888-6898.2003
Copyright © 2003, American
Society for
Microbiology. All Rights Reserved.
Christen A. White, Muthusamy Kunnimalaiyaan,
and Patricia S. Vary*
Present address: Argonne National Laboratory, Bioscience Division, Argonne, IL 60439.,
Present address: Department of Surgery, K4/638 Comprehensive Cancer Center, University of WisconsinMadison, Madison, WI 53792.,
Department of Biological Sciences, Northern Illinois University, DeKalb, Illinois 60115,
Received 26 June 2003/ Accepted 18 August 2003
|
|
|---|
|
|
|---|
Many plasmids in Bacillus and other gram-positive bacteria are cryptic in nature. While the small rolling circle replicating plasmids of gram-positive bacteria have been extensively studied (20, 22), only a few of the plasmid genes or replicons of the large theta plasmids have been characterized (for reviews, see references 10 and 16). Plasmids usually provide the host cell with a selective advantage, such as antibiotic resistance (17), resistance to toxic heavy metals (49), or degradation of unusual compounds (14, 44). In addition, some plasmids produce bacteriocins (for review, see reference 19). Bacteriocin (megacin) genes have been reported to be carried by a 47-kb plasmid of B. megaterium 216 (43), a 45-kb plasmid of ATCC 19213 (59), and a plasmid in strain QM B1551 (23). Gram-positive bacteria may also harbor plasmids that carry virulence genes, i.e., the insecticidal toxins of B. thuringiensis and B. sphaericus (46) and the anthrax toxins produced by B. anthracis from plasmids pXO1 and pXO2 (36, 42).
B. megaterium QM B1551 carries over 11% of its cellular DNA as stable plasmid DNA under laboratory conditions without known selection (23). The seven indigenous plasmids are 5.4, 9.1, 26.3, 54, 71, 108, and 165 kb (23) and are designated pBM100 to pBM700, respectively. A plasmidless derivative of QM B1551, strain PV361, has been isolated (58). The complexity of the array of plasmids in B. megaterium offers a unique opportunity to study plasmid interactions, replication, compatibility, plasmid-borne genes, and possible horizontal transfer. It also presents a challenge for the isolation and sequencing of plasmids with no known selectable markers and widely varying copy numbers within a multiplasmid array. When strain QM B1551 of B. megaterium is cured of its seven indigenous plasmids, there are few obvious phenotypic effects under laboratory conditions besides a defect in germination (52) and production of a megacin (23). QM B1551 is known to be very efficient at sporulation and germination (12). The plasmidless strain PV361 can grow normally on rich media, sporulate normally, and germinate on rich media but cannot germinate on single germination triggers because of a plasmid-borne germination gene (52). The plasmids are dispensable under laboratory conditions, yet they are fairly stable upon subculturing.
To begin to understand plasmid interactions and the role of plasmid genes in B. megaterium plasmid DNA, we cloned and sequenced fragments containing six of seven compatible replication origins. The two smallest plasmids have been completely sequenced and identified as rolling circle replicating plasmids with replicon similarities to the pTX14-3 plasmid of B. thuringiensis and the pXO1-89 plasmid gene of B. anthracis (G. Baisa, Y. Zhou, and P. Vary, unpublished data). The remaining four sequenced replicons on clones I, II, III, and IV are iteron-type theta replicons from the larger plasmids and are very similar in both Rep proteins and iterons (all five large plasmid replicons cross-hybridize) (24, 51, 57; M. Kunnimalaiyaan and P. Vary, unpublished data). The 12.3-kb replicating clone II fragment and its adjacent rRNA operon (6.3 kb), located just upstream from this replication clone, have been sequenced and characterized from pBM400 (54 kb) (24). To date, this is the only rRNA operon found in nonessential DNA, and its presence has considerable evolutionary significance. Since B. megaterium produces many unusual enzymes and is often found in contaminated environments along with degradative-plasmid-bearing pseudomonads (56), we are greatly interested to see if any similar metabolic genes are plasmid-borne. The complete sequencing of pBM400 and the larger cryptic plasmids of B. megaterium will enable comparative plasmid genomics and the study of plasmid horizontal transfer while expanding our understanding of plasmid genetics in gram-positive bacteria.
This paper presents the recombinational labeling, isolation, and sequencing of the remaining 35 kb of pBM400. It reports the initial description of several interesting plasmid-borne genes, including a possible integrase, an insertion sequence (IS) element, unusual metabolic genes, and several large inverted and direct repeats (DRs).
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Strains
and plasmids used in this study
|
55 kb). Large quantities of
plasmid DNA were isolated by the method of Lovett and Keggins
(29) after growth to an
optical density at 600 nm of 1.0 in MC broth. Purification of plasmid
DNA by CsCl-ethidium bromide gradient centrifugation and library
construction were performed according to the methods of Sambrook et al.
(45). Three restriction
enzyme libraries as follows were constructed from the pBM400 plasmid:
(i) a partial Sau3A library in pGEM3-Z f(+) (Promega,
Madison, Wis.) was used for shotgun cloning and sequencing in E.
coli DH10B, and (ii) EcoRI and (iii) PstI
fragment libraries in pJM103
(39) were used for
closing sequence gaps. B. megaterium was transformed by
protoplast fusion based on the method of Von Tersch and Carlton
(59) and regenerated in
regeneration medium
(43,
59) as described by
English and Vary
(11).
PCR.
PCR was performed by using Promega
Taq polymerase (Promega) or cDNA polymerase (Clontech, Palo
Alto, Calif.) according to the manufacturer's specifications with
a Perkin-Elmer (Foster City, Calif.) 480 thermocycler. Primers derived
from the sequence of clone II that were specific for pBM400 and used
for the detection of integration of the kanamycin gene insertion were
as follows: 5'-CGGGAAGATGGCAAAT-3'
and
5'-GCCTGGCAACGTCCTACTC-3'.
Southern hybridization.
Southern hybridization and probe
labeling was done by use of the NEBlot phototope kit (New England
Biolabs, Beverly, Mass.) according to the manufacturer's
instructions. The resulting membrane was processed and exposed to X-ray
film for 3 to 12 min.
DNA
sequencing.
DNA sequencing
was performed in the core DNA sequencing facilities at Northern
Illinois University (DeKalb, Ill.) and Integrated Genomics (Chicago,
Ill.) by use of the DYEnamic ET terminator cycle sequencing kit
(Amersham Biosciences, Piscataway, N.J.) and a 373 DNA sequencer
(Applied Biosystems, Foster City, Calif.) or AB3700 sequencer
(Perkin-Elmer). Universal primers were utilized for end sequencing of
Sau3A, EcoRI, and PstI library clones.
Sequence gaps were closed by primer walking and PCR, with custom
primers from MWG Biotech (High Point, N.C.). Sequences were assembled
by use of Sequencher (Gene Codes, Ann Arbor, Mich.). Approximately 120
oligonucleotides were designed for PCR and primer walking of pBM400 to
close gaps and increase the overall coverage to an average of
x3.
Construction of the
integrative vector and integration.
A list of all plasmids utilized in
vector construction is summarized in Table
1. An integrative plasmid
was constructed that carried a kanamycin gene flanked by DNA homologous
to pBM400, as shown in Fig.
1. The plasmid pKM31 harbors a 2.1-kb XbaI-to-KpnI
subclone from the distal 3' end of clone II that is specific
for pBM400 (24). This
plasmid was cleaved with EcoRI to remove one of two
ClaI sites, ligated, and transformed into E. coli
DH5
cells. The resultant plasmid, pKM58, was cleaved with
ClaI, and a kanamycin cassette from pDG792
(15), also digested with
ClaI, was inserted (pKM59). Plasmid pLTV1
(7) was cut with
EcoRI, and the 4-kb fragment harboring the
temperature-sensitive replication origin (pE194ts) was ligated into
pKM59 at the EcoRI site. The resulting temperature-sensitive
pBM400-specific integrative plasmid for B. megaterium was
designated pSM1 (Fig. 1).
QM B1551 containing pSM1 was grown at 46°C in SNB-kanamycin,
with subculturing at 6 and 14 h. Cultures were serially
diluted, plated on SNB plates, grown overnight at 37°C, and
replica plated on SNB-chloramphenicol and SNB-kanamycin plates to test
for integration of the labeled fragment by double crossover and loss of
pSM1.
![]() View larger version (21K): [in a new window] |
FIG. 1. Construction
of an integrative-replicative vector. The figure shows the parental
clone containing an EcoRI-KpnI clone II fragment
specific to pBM400 (A), removal of a 1-kb fragment containing a
ClaI site (in bold) (B), incorporation of a 1.5-kb kanamycin
gene from pDG792 into the remaining ClaI site (C), and
insertion of the 4-kb fragment with the origin of replication from
pE194ts to yield pSM1
(D).
|
Nucleotide
sequence accession number.
Sequence data for the entire 53,903
bp of pBM400 have been submitted to GenBank databases under the
accession number
AF142677.
The previously reported 18.6 kb corresponds to 8,003 bp before the
BglII start site (nucleotide 45,900) to nucleotide 10,630 in
the completed plasmid
sequence.
|
|
|---|
Isolation of a B.
megaterium strain containing only
pBM400::kan.
Plasmid DNA from QM
B1551/pBM400::kan was separated by
electrophoresis, and a band corresponding to a size of approximately 54
kb was purified from agarose gels. Plasmidless strain PV361 was
transformed with pBM400::kan DNA and
selected for Kmr. Plasmid DNA was purified by CsCl-ethidium
bromide gradient centrifugation from one of the transformants,
designated PV627, and the presence of only one plasmid and its correct
size were verified by agarose gel electrophoresis and hybridization.
Plasmid DNA was probed with both a pBM400-specific probe
(KpnI-XbaI fragment from pKM31) and a kanamycin gene
probe (pDG792 1.5-kb ClaI fragment), as shown in Fig.
2. As expected, the kanamycin gene probe hybridized to the bands
corresponding to a 54-kb plasmid in PV627 and QM B1551/pSM1, but not to
QM B1551 plasmid DNA or chromosomal DNA from plasmidless PV361. The
pBM400 probe from clone II produced a signal, as expected, for plasmid
bands corresponding to approximately 54 kb, but not for an E.
coli V517 plasmid standard or B. megaterium chromosomal
DNA. These results verified pBM400 as the source plasmid for clone II
and showed that double-crossover integration of the kanamycin gene had
occurred. It was interesting that in QM B1551/pSM1, a band
approximately 10 kb above 54 kb suggested that it contained a
single-crossover integration as well as pSM1, which was confirmed by a
chloramphenicol gene probe (data not shown). However, it is clear from
the hybridization reactions with PV627 and QM
B1551/pBM400::kan that the double
crossover did occur during subculturing at
46°C.
![]() View larger version (36K): [in a new window] |
FIG. 2. Southern
hybridization with pBM400 and kanamycin gene probes to detect the
integrated kanamycin gene. (A) Southern hybridization with a
kanamycin gene DNA probe. Lane 1, E. coli V517 supercoiled
standard; lane 2, QM B1551; lane 3, 54-kb plasmid
pBM400::kan (from PV627); lane 4, QM B1551
containing pSM1 and a possible single crossover of pSM1 into pBM400;
lane 5, pBM400::kan in the QM B1551
plasmid array following growth at 46°C; lane 6, PV361
chromosomal DNA. (B) Agarose gel electrophoresis of the same
gel. (C) Southern hybridization with the pBM400-specific
probe.
|
|
View this table: [in a new window] |
TABLE 2. Plasmid
pBM400-carried genes
|
![]() View larger version (38K): [in a new window] |
FIG. 3. Circular
map of pBM400. The positions of genes are shown relative to a
BglII start point of pBM400. Putative sigma A promoters (bent
arrows), IRs (blue arrows), DRs (gray arrows), and possible
rho-independent terminators are shown. The rRNA operon is shown in
blue; the replicon region is shown in red. Names for gene products with
similarity to proteins with known functions in the databases are in
bold, and the ORFs are green. ORFs for gene products with similarity to
known proteins but with unknown functions are yellow. Those with no
similarity in the databases are white. Large bent arrows are sites of
five tandem putative promoters. The 39-bp IRs flanking the transposase
(ORF18) are red and are labeled IS. The large 127-bp IRs flanking the
Int/Rec ORF (14) and 79-bp repeats (1) near the replicon are also red.
ORFs shown on the outside of the circle are read in the forward
direction (clockwise) and those on the inside are read in the reverse
direction. All of the restriction sites of EcoRI and
PstI and relevant BglII, KpnI,
ClaI, and XbaI restriction sites are
shown.
|
The predicted ORF22 protein was highly similar to a cadmium-copper ATPase of the oral bacterium Fusobacterium nucleatum that is a component of heavy metal transport across the bacterial membrane. The adjacent ORF (ORF20) product was also similar to a hypothetical protein in the same position in the same organism. The cadmium-copper ATPase encoded by pBM400 was also similar to the first cadmium ATPase found for gram-positive organisms, encoded by Staphylococcus plasmid pI258 (49). Cadmium resistance genes have been reported for other gram-positive bacterial plasmids, including pXU5 in Staphylococcus aureus (55), pND302 in Lactococcus lactis (26), and other soil bacilli (18). However, initial tests for copper and cadmium resistance in B. megaterium QM B1551 yielded ambiguous results (data not shown). Another cation-transporting ATPase gene (ORF19) is located just upstream on pBM400. Plasmid pBM400 also carries a possible zinc-binding dehydrogenase gene (ORF24) in the same region.
Interestingly, the predicted ORF25 protein showed significant similarity to a styrene monooxygenase from Pseudomonas that is involved in the biodegradation of styrenes. In Pseudomonas fluorescens ST, four genes, designated styA, styB, styC, and styD, are required for the oxidation of styrene to phenylacetic acid (5, 32). The protein encoded by pBM400 is similar to the styA gene product, which forms epoxystyrene from styrene (5). Chiral epoxides are valuable intermediates in several optically active drugs. Homologues to other genes for styrene degradation were not found on pBM400, but there seem to be several pathways for styrene utilization found in different organisms, including one that has only the monooxygenase with oxidation to ß-phenylethanol (50). Styrene utilization often requires a consortium of organisms. To our knowledge, there has been only one report of styrene utilization in Bacillus, by thermophilic isolates that were not further identified (50). The role of the monooxygenase is also of interest since three other P450 cytochrome oxygenases are present in strains of B. megaterium and have been intensely studied (40, 47).
Possible cell division proteins FtsZ and FtsK (SpoIIIE), encoded by ORFs 44 and 27, respectively, were not expected to be carried by a plasmid. FtsZ is a key division protein that has been shown to form the signature Z-ring at the septum during cell division in E. coli (1). In B. subtilis, FtsZ is required for both vegetative septation and asymmetric septation during sporulation (4). Although FtsZ is essential for normal cell division, excess FtsZ has the same effect as its absence, so that its presence on a multicopy plasmid is puzzling. The ORF44 product is similar to FtsZ from two Archaea, Pyrococcus horikoshii and Methanocaldococcus jannaschii, but not to B. subtilis or E. coli FtsZ. However, a motif for GTPase activity of FtsZ was found in pBM400 FtsZ (GGGSGGG) that was similar to the functional motif GGGTGTG reported for E. coli (9). A 1.1-kb PCR probe was therefore made to recognize plasmid FtsZ and was hybridized to DNA from B. megaterium, E. coli, and B. subtilis as shown in Fig. 4. This probe hybridized only to plasmid DNA from QM B1551 and not to B. subtilis, E. coli, or genomic B. megaterium DNA (PV361). Since the plasmidless strain PV361 divides normally, there must be an unrelated gene for FtsZ on the chromosome as well. To test this, a 1.1-kb PCR probe was also made for the B. subtilis FtsZ gene (25) with the following primers: forward, 5'-CATTCGGCAGATTAGGAGG-3', and reverse, 5'-GATTTTGTCCTTTACATTAGC-3'. There was evidence of hybridization to appropriate PstI fragments (11.4 and 5.8 kb) and an EcoRI fragment (10.8 kb) in the B. subtilis genomic lanes. There was also faint binding to PV361 and QM B1551 in the chromosomal DNA lanes at 4.8 kb, but not to plasmid DNA (data not shown).
![]() View larger version (52K): [in a new window] |
FIG. 4. Hybridization
of pBM400 FtsZ (ORF44) with genomic and plasmid DNA of E.
coli, B. subtilis, and B. megaterium. A 1.1-kb
PCR product from ORF44 (FtsZ) of pBM400 was used as a probe to test for
similarity to genes on E. coli, B. subtilis, and
B. megaterium chromosomes. The primers were forward,
5'-AGGTTGAGGTACCATGATAGGTA-3', and
reverse, 5'-ATACTAAAAAGAATTCACGATTCTG-3'.
Lanes 1, 11, and 19, biotinylated 1-kb standard; lanes 2, 3,
and 4, E. coli DH5 total DNA cut with PstI,
EcoRI, and ClaI, respectively; lanes 5, 6, and 7,
B. subtilis total DNA cut with PstI, EcoRI,
and ClaI, respectively; lanes 8, 9, and 10, PV627 plasmid DNA
cut with PstI, EcoRI, and ClaI,
respectively; lanes 12, 13, and 14, QM B1551 total DNA cut with
PstI, EcoRI, and ClaI, respectively; lanes
15, 16, and 17, PV361 (plasmidless) total DNA cut with PstI,
EcoRI, and ClaI, respectively; lane 18, pSM3 cut with
EcoRI.
|
Other genes carried by pBM400 include those for ORF26, coding for a protein similar to a cell wall hydrolase, CwlJ, from B. subtilis that is one of two proteins necessary to hydrolyze the spore cortex during germination. A cell wall lytic enzyme has been purified and characterized from B. megaterium KM by Foster and Johnstone (13), but to our knowledge the gene has not been sequenced. CwlJ has recently been determined to localize to the spore coat of B. subtilis (3, 8). In addition to the germination and cell division genes, several ORFs showed similarity to B. subtilis genomic gene products of unknown function (YcnI, YcgR, and YpiB). Four small ORFs, ORF46 to -50, were also found within the rRNA operon. Small genes within the rRNA operon have been shown to be expressed in E. coli, but their functions are unknown (6).
Possible
virulence plasmid gene exchanges.
Three ORF products had significant
similarities to proteins from the B. anthracis virulence
plasmid pXO1. ORF34 was 60% similar to B. anthracis
pXO1-07, encoding a reverse transcriptase, while ORF40 was similar to
pXO1-103, encoding an integrase-recombinase (see below). The ORF41
product was similar to pXO1-101, which has been reported to have
considerable nucleotide homology with the virulence factor
bceT from B. cereus
(37). ORF41 has a motif
similar to the HTH-XRE group of transcriptional regulators that include
the lambda repressor.
Possible mobile
elements.
A putative
insertion sequence (IS) element was found in which 39-bp IRs
(74% identical) flank ORF18. ORF18 codes for a protein that is
highly similar to transposase ISEf1 of vancomycin-resistant
Enterobacter faecalis V583
(38) and to transposases
from IS1201 of Lactobacillus helveticus and
IS905 of Lactococcus lactis. Both IS1201 and
IS905 belong to the IS256 family of transposons
(31), which have 24- to
41-bp IRs. IS1201 is a 1,387-bp IS that carries a single ORF
coding for a putative transposase
(53). While the pBM400
IS, designated ISBm400, has a similar transposase, its 39-bp IRs
(5'AACTGAGAAAAATTAATACATTTAGTTTTTCCATTTCTT 3'
and
5'AAGAGAATGAAAAACTAGTTGAATTAATTATTTTAAATT 3')
(labeled IS in Fig.
3) have no significant
homology to any known IS element. It is also large (2,653 bp) but
contains no obvious ORF other than that for the
transposase.
ORF40 exhibited 48% similarity to the integrase-recombinase gene (pX01-103) of B. anthracis virulence plasmid pXO1. Plasmid pXO1 carries two putative integrase-recombinase ORFs similar to Methanobacterium thermoautotrophicum Int/Rec and B. subtilis RipX, the homolog of E. coli XerD required for chromosomal dimer resolution (36). ORF40 of pBM400 is more similar to the pXO1 RipX integrase-recombinase and is flanked by large IRs of 127 bp (Fig. 3) with 79% identity (see below). The possible pBM400 integron also contained two other predicted ORFs (38 and 39) with unknown functions. Its intervening sequence was about 1.5kb. On pXO1, the integrase-recombinase gene is flanked by IRs and is part of a region with several IS elements, integrases, and transposases within the 44.8-kb pathogenicity island. The presence on pBM400 of integrase-recombinase and transposase genes flanked by IRs strongly suggests the possibility of DNA mobility.
GC Analysis.
To test the possibility of horizontal
transfer, the entire pBM400 sequence was analyzed by use of Artemis
software (see Materials and Methods). Horizontal transfer can sometimes
be identified by one or more anomalies in GC content, GC deviation, and
dinucleotide usage (21).
As can be seen in Fig.
5, the rRNA operon showed typical high GC content (52%). Most of
the rest of the plasmid was near 36.5% GC, except for a few
regions that were very AT-rich, including the possible integron, the
copper export gene, and the styrene monooxygenase. The copper and
integrase regions also showed Karlin signature differences. There was a
smaller GC-rich region near or within the IS region, which also
exhibits a Karlin signature peak. (The transposase gene is 40.4%
GC.) The FtsK gene is 44% GC, but it is in a region with no
obvious anomalies. Interestingly, the conserved plasmid ORF2C gene has
the highest Karlin signature difference peak and a small region of GC
deviation.
![]() View larger version (55K): [in a new window] |
FIG. 5. Whole
plasmid GC analysis of pBM400. The entire plasmid was analyzed for
possible alien genes or regions (pAs) by use of Artemis software for
three different parameters: GC content, GC deviation, and Karlin
signature difference. The line in the GC content graph is at
36.5%. Possible regions of sequence anomalies in one or more
graphs are labeled. rrn, rRNA operon; Cu, copper export gene region;
Sty, styrene monooxygenase region; Int, integron; 2C, ORF2C (ORF6); IS,
ISBm400.
|
Potential sigma A promoters were found based on sequence similarity to the sigma A consensus of B. subtilis: -10, TATAAT, and -35, TTGACA (34). There were five possible promoter sequences found in close proximity to each other upstream of the small ORF21, the product of which has no similarity to those in the databases. Five adjacent putative promoters were also found upstream of ORF41, coding for the putative integrase-recombinase. Overall, there were 76 sequences that showed significant similarity to the Bacillus sigma A consensus sequence. However, actual promoter predictions from sequences have not proved reliable.
|
|
|---|
This research was supported in part by NIH grant IR 155GM59599-01/G12401R.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»