Mga is sufficient to activate transcription in vitro of sof-sfbX and other Mga-regulated virulence genes in the group A Streptococcus.

Abstract

The group A streptococcus (GAS), or Streptococcus pyogenes, is a strict human pathogen of medical significance, causing infections ranging from pharyngitis (strep throat) to necrotizing fasciitis (flesh-eating disease). Several virulence genes that encode factors important for colonization, internalization, and immune evasion are under the control of the multiple gene regulator of the GAS, or Mga. Mga functions as a DNA-binding protein that interacts with sites both proximal (Pemm and PscpA) and distal (PsclA) to the start of transcription for the genes that it regulates. The genes encoding serum opacity factor, sof, and a novel fibronectin-binding protein, sfbX, are cotranscribed and represent two uncharacterized Mga-regulated virulence genes in the GAS. Analysis of the promoter region of sof-sfbX identified a putative Mga-binding site 278 bp upstream of the regulated start of transcription as determined by primer extension. Electrophoretic mobility shift assays demonstrated that Mga is able to bind specifically to the single distal site in a fashion similar to the previously characterized PsclA. In order to better understand the events that take place at this and other Mga-regulated promoters, an in vitro transcription assay was established. Using this assay, we showed that Mga is sufficient to activate transcription in vitro for Mga-regulated promoters containing both proximal (Pemm) and distal (PsclA and Psof-sfbX) binding sites. These results indicate that additional factors are not required for Mga-specific activation at diverse promoters in vitro, although they do not rule out the potential influence of other components on the Mga virulence regulon in vivo.

Full text

Page 1

JOURNAL OF BACTERIOLOGY, Mar. 2006, p. 2038–2047
0021-9193/06/$08.00?0 doi:10.1128/JB.188.6.2038–2047.2006
Copyright © 2006, American Society for Microbiology. All Rights Reserved.
Vol. 188, No. 6
Mga Is Sufficient To Activate Transcription In Vitro of sof-sfbX and
Other Mga-Regulated Virulence Genes in the Group A Streptococcus
Audry C. Almengor, Matthew S. Walters, and Kevin S. McIver*
Department of Microbiology, University of Texas Southwestern Medical Center, Dallas, Texas 75390-9048
Received 23 November 2005/Accepted 4 January 2006
The group A streptococcus (GAS), or Streptococcus pyogenes, is a strict human pathogen of medical signif-
icance, causing infections ranging from pharyngitis (strep throat) to necrotizing fasciitis (flesh-eating disease).
Several virulence genes that encode factors important for colonization, internalization, and immune evasion
are under the control of the multiple gene regulator of the GAS, or Mga. Mga functions as a DNA-binding
protein that interacts with sites both proximal (Pemm and PscpA) and distal (PsclA) to the start of transcrip-
tion for the genes that it regulates. The genes encoding serum opacity factor, sof, and a novel fibronectin-
binding protein, sfbX, are cotranscribed and represent two uncharacterized Mga-regulated virulence genes in
the GAS. Analysis of the promoter region of sof-sfbX identified a putative Mga-binding site 278 bp upstream of
the regulated start of transcription as determined by primer extension. Electrophoretic mobility shift assays
demonstrated that Mga is able to bind specifically to the single distal site in a fashion similar to the previously
characterized PsclA. In order to better understand the events that take place at this and other Mga-regulated
promoters, an in vitro transcription assay was established. Using this assay, we showed that Mga is sufficient
to activate transcription in vitro for Mga-regulated promoters containing both proximal (Pemm) and distal
(PsclA and Psof-sfbX) binding sites. These results indicate that additional factors are not required for Mga-
specific activation at diverse promoters in vitro, although they do not rule out the potential influence of other
components on the Mga virulence regulon in vivo.
Differential control of gene expression in bacteria is an im-
portant adaptive response to various environments and is often
mediated by transcription factors that either activate or repress
initiation of transcription. In the case of activators, the factors
are thought to increase transcription initiation by recruiting
RNA polymerase (RNAP) to the promoter via direct protein-
protein interactions (10). Such interactions between RNAP
and the activator have been shown to occur with the carboxyl
terminus of the ? subunit for class I transcription factors or
with the ? subunit for class II transcription factors. Class I
activators typically bind to DNA just upstream of the ?35
region of the promoters that they regulate in order to facilitate
their interaction with the ? subunits. Class II activator binding
sites, on the other hand, tend to overlap the ?35 region while
still promoting transcriptional activation (10). Less common,
however, are transcription factors capable of activating tran-
scription from a single or multiple distant sites, especially when
the Escherichia coli housekeeping ?70is employed by the
RNAP (6). In fact, the majority of transcription factors that
activate from far upstream sites are required to initiate ?54-
dependent promoters, such as the prototypical regulator NtrC.
Despite their binding distal to the ?35 region, they are still
considered a class II transcription factor (10).
Transcriptional regulators play a vital role in the pathogen-
esis of the group A streptococcus (GAS), or Streptococcus
pyogenes. The GAS is a strict human pathogen that is capable
of causing disease in varied locations throughout the body,
including the skin (impetigo), the respiratory tract (pharyngi-
tis), and deeper tissues (necrotizing fasciitis). Unlike many
other prokaryotes, the ability of the GAS to coordinately reg-
ulate expression of genes does not depend on alternative sigma
factors, since the GAS has only one that is thought to control
transcription of a limited number of genes (21). Rather, the
organism is able to respond to environmental cues by relaying
information to a variety of transcriptional regulators.
One such regulator is the multiple gene regulator of the
GAS, or Mga. Mga is responsible for activating the expression
of several surface-associated or secreted factors that enable the
bacterium to adhere to epithelial surfaces, invade cells, and
evade the immune system. These factors include the serum
opacity factor (OF), M protein, and the streptococcal collagen-
like protein (SclA) (4, 15, 20, 23, 24). Mutations in mga and
Mga-regulated genes result in strains that are significantly at-
tenuated for virulence in a variety of animal models (7, 12, 13,
15, 25). Furthermore, a recent longitudinal study of pharyngitis
in macaques found that the Mga regulon was highly expressed
in vivo during the acute phase of infection (30). Mga is known
to bind to the promoters of the genes that it activates (1, 16,
19), but its mechanism of action in transcriptional activation
remains to be determined.
Since the Mga-binding sites in the emm (M protein) and
scpA (C5a peptidase) promoters are located proximal to the
start of transcription (16), it is thought that Mga acts like a
class I transcription factor to activate gene expression by bind-
ing to the ? subunit of the RNAP. However, it was recently
discovered that Mga is capable of activating transcription from
a distal binding site in the sclA promoter (1). Thus, Mga-
regulated promoters were grouped on the basis of the number
and location of Mga-binding sites: a single proximal site (cat-
egory A; emm and scpA), a single distal site (category B; sclA),
* Corresponding author. Mailing address: Department of Microbi-
ology, University of Texas Southwestern Medical Center, Dallas, TX
75390-9048. Phone: (214) 648-1255. Fax: (214) 648-5907. E-mail:
Kevin.Mciver@UTSouthwestern.edu.
2038

Page 2

and multiple distal sites (category C; mga) (19). The differ-
ences in these promoters suggest that Mga may be capable of
activating transcription by multiple mechanisms. In this study,
a second category B Mga-regulated promoter is identified (sof-
sfbX), and Mga is demonstrated to be sufficient to activate both
category A and B promoters in vitro.
MATERIALS AND METHODS
Bacterial strains and media. Bacterial strains used in this study are listed in
Table 1. The GAS was cultured in Todd-Hewitt medium supplemented with
0.2% yeast extract (THY), and growth was followed by optical density using a
Klett-Summerson photoelectric colorimeter with the A filter. E. coli was grown
in Luria-Bertani broth (LB). Antibiotics were used at the following concentra-
tions: ampicillin at 50 to 100 ?g/ml for E. coli, rifampin at 150 ?g/ml for E. coli,
and spectinomycin at 100 ?g/ml for the GAS.
DNA manipulations. Plasmid DNA was isolated from E. coli using either the
Wizard Miniprep system (Promega) or Maxi/Midi prep purification systems
(QIAGEN). DNA fragments were isolated from agarose gels using the QIAquick
gel extraction kit (QIAGEN). Chromosomal DNA from the GAS was isolated
using the FastDNA kit and a FastPrep cell disruptor (Bio101, Inc.). PCR for
cloning and promoter probes was performed using Platinum Pfx high-fidelity
DNA polymerase (Invitrogen), and reactions were purified using the QIAquick
PCR purification system (QIAGEN). PCR for diagnostic assays was performed
using Taq DNA polymerase (New England Biolabs). DNA sequencing was per-
formed either using the Excel II cycle sequencing kit (Epicenter, Inc.) or through
the automated sequencing core facility in the McDermott Center at University of
Texas Southwestern Medical Center.
Expression and purification of Mga-His protein from E. coli. Mga-His protein
was purified from E. coli as previously described (1). Briefly, cultures of E. coli
BL21(DE3) containing pKSM170 (PT7-mga-his) were grown to an optical density
at 600 nm (OD600) of 0.6 at 30°C in LB plus ampicillin (50 ?g/ml), and expression
of the protein was induced for 1.5 h by addition of 1 mM IPTG (isopropyl-?-D-
galactopyranoside). Cells were lysed by two passages through a prechilled French
pressure cell, and Mga-His was purified over a Ni-nitrilotriacetic acid (Ni-NTA)
resin column (QIAGEN) under native conditions. Protein concentrations were
determined using the Bio-Rad protein assay kit. Integrity of the purified protein
was assessed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-
PAGE) and Western blot analysis as described below. Mga-His purified for the
purposes of in vitro transcription was dialyzed twice against 1 liter of transcrip-
tion buffer minus bovine serum albumin (33 mM Tris acetate, pH 8, 10 mM
magnesium acetate, and 0.5 mM dithiothreitol) for 4 h to overnight prior to
storage at ?80°C.
Western blot analysis. Anti-Mga-His immunoblots were performed as de-
scribed previously (17). Briefly, purified Mga-His proteins were separated on
SDS-10% PAGE. Proteins were transferred to nitrocellulose membranes and
reacted with the anti-His monoclonal antisera (Novagen) at a 1:2,000 dilution for
2 h at room temperature. Blots were reacted with horseradish peroxidase-con-
jugated anti-mouse immunoglobulin G secondary antibody (Chemicon Int.),
developed using the Western Lighting chemiluminescence system (Perkin-
Elmer), and visualized on autoradiography film (Kodak).
Electrophoretic mobility shift assays (EMSAs). Promoter probes were gener-
ated by PCR amplification using either serotype M4 AP4 (OF?), serotype M1
SF370 (OF?), or serotype M6 strain JRS4 (OF?) chromosomal DNA and the
relevant primer pairs (Psof, Pemm, PscpA, Pmga, PsclA, and nonspecific) listed in
Table 2. PCR fragments were end labeled with [?32P]ATP using T4 polynucle-
otide kinase (New England Biolabs). Labeled fragments were excised from a 5%
polyacrylamide gel, extracted by crush and soak elution, and purified using the
QIAquick PCR purification system (QIAGEN).
EMSA was performed as described previously (16). Briefly, a constant amount
of labeled promoter probe DNA (ca. 1 to 5 ng) and increasing amounts of
purified Mga-His (1 to 2 ?g) were used in each reaction mixture. Competition
assays were performed by addition of 750 ng unlabeled promoter probes to
binding reaction mixtures. Reaction mixtures were run on a 5% polyacrylamide
gel at 4°C. The gel was dried under vacuum at 80°C for 1 h and exposed overnight
to a phosphorimaging screen. The phosphor screen was scanned using a
Storm820 (Amersham Biosciences), and resulting data were analyzed with the
ImageQuant analysis software (version 5.2).
Primer extension analysis. Total RNA was extracted from samples in late
exponential phase as described previously (17) using the FastRNA kit and a
FastPrep cell disruptor (Bio101, Inc.). A primer extension was performed on 5
?g of total RNA from GAS strains AL168str-r and nr4 as described previously
(19) using the primer Psof-R1 (Table 2). The primer extension product was run
on a 6% denaturing polyacrylamide gel (Amresco), and the gel was processed as
described above for EMSAs. Sequence was generated using labeled Psof-R1
primer on a PCR product amplified from AL168str-r genomic DNA (Table 2;
Psof-L5 and Psof-R1) as described above.
Construction of in vitro transcription plasmids. A 421-bp fragment of PrpsL
was PCR amplified from the M1 strain SF370 by using the primers GAS-rpsL1
and GAS-rpsL7 (Table 2) and ligated into the EcoRV-digested pBluescript II
KS(?) (Stratagene) to form pKSM415 (Table 1). A 420-bp fragment of Pemm
TABLE 1. Bacterial strains and plasmids used in this study
Strain or plasmid Description
Reference
or source
Strains
GAS
AL168str-r
nr4
JRS4
JRS4-PolHis (12Z05A)
JRS519
SF370
AP4
E. coli
DH5?
BL21(DE3)
M22, OF?, streptomycin-resistant derivative of AL168
M22, Tn916 mutant of AL168str-r in mga
M6, streptomycin-resistant derivative of D471
M6, derivative of JRS4 with chromosomal His-tagged ?? subunit of RNA polymerase
M6, derivative of JRS4 (mga-10)
M1, sequenced strain
M4, OF?
28
28
26
21
18
27
14
hsdR17 recA1 gyrA endA1 relA1
F?ompT hsdSB(rB
9
Novagen
?mB
?) gal dcm (DE3)
Plasmids
pBluescript II KS(?)
pCD2
pKSM164
pKSM170
pKSM415
pKSM416
pKSM417
pKSM418
pKSM419
pKSM420
ColE1 ori AmprlacZ?
PT7-?A(Bacillus subtilis)
Pmga-mga-his
PT7-mga-his
PrpsL, 3? HindIII site
Pemm, 3? EcoRI site
PsclA (sequence downstream of Mga-binding site), 3? XhoI site
PsclA (?Mga-binding site), 3? XhoI site
PsclA (full length), 3? XhoI site
Psof, 3? BamHI site
Stratagene
5
1
1
This study
This study
1
1
1
This study
VOL. 188, 2006Mga IS SUFFICIENT TO ACTIVATE GENES IN S. PYOGENES2039

Page 3

containing the Mga-binding site was PCR amplified from SF370 by using the
primers Pemm1-L1 and Pemm1-R1 (Table 2) and ligated into the EcoRV-
digested pBluescript II KS(?) to form pKSM416 (Table 1). A 692-bp fragment
of Psof containing the Mga-binding site was PCR amplified from the M4 strain
AP4 by using the primers Psof-L5 and Psof-R1 (Table 2) and ligated into the
EcoRV-digested pBluescript II KS(?) to form pKSM420 (Table 1).
Expression and purification of Mga-His protein from the GAS. Mga-His
protein was purified from the GAS as previously described (1). Briefly, cultures
of JRS519 carrying pKSM164 (Pmga-mga-his) were grown to late logarithmic
phase at 37°C in THY plus spectinomycin. Cells (20 50-ml aliquots) were lysed
using five 9-s pulses at 4°C on a FastPrep cell disruptor (Bio101, Inc.), and
Mga-His was purified over a Ni-NTA resin column (QIAGEN) under native
conditions. Protein concentrations were determined using the Bio-Rad protein
assay kit, and the integrity of the purified protein was assessed by SDS-PAGE
and Western blot analysis as described above. Mga-His purified for the purposes
of in vitro transcription was dialyzed twice against 1 liter of transcription buffer
minus bovine serum albumin (33 mM Tris acetate, pH 8, 10 mM magnesium
acetate, and 0.5 mM dithiothreitol) for 4 h to overnight prior to storage at
?80°C.
Expression and purification of core RNA polymerase from the GAS. Core
RNA polymerase was purified from the GAS as previously described (21) with
modifications. Briefly, cultures of JRS4-PolHis were grown to an OD600of 0.8 at
37°C in THY plus spectinomycin. Cells (20 50-ml aliquots) were lysed using five
9-s pulses at 4°C on a FastPrep cell disruptor (Bio101, Inc.), and RNA polymer-
ase was purified over a Ni-NTA resin column (QIAGEN) under native condi-
tions. Fractions containing the polymerase were then passed over a HiTrap Q
Sepharose FF anion exchange column (Amersham Biosciences) and eluted with
a linear salt gradient. Protein concentrations were determined using the Bio-Rad
protein assay kit, and the integrity of the purified protein complex was assessed
by SDS-PAGE.
Expression and purification of ?Aof Bacillus subtilis from E. coli. ?Awas
purified from E. coli as previously described (3). Briefly, E. coli BL21(DE3)
carrying pCD2 (5) was grown to an OD600of 0.8 at 37°C in LB plus ampicillin
(100 ?g/ml), and expression of the protein was induced for 0.5 h by addition of
1 mM IPTG. Rifampin was added at 150 ?g/ml, and the cells were grown for
another 3.5 h. Cells were lysed by two passages through a prechilled French
pressure cell, and sodium deoxycholate was added to solubilize the cell mem-
branes. Inclusion bodies containing ?Awere pelleted, washed, and solubilized
with 0.4% Sarkosyl, and the protein was then refolded by dialysis. The protein
was loaded onto a HiTrap Q Sepharose FF anion exchange column (Amersham
Biosciences) and eluted by a linear salt gradient. Protein concentrations were
determined using the Bio-Rad protein assay kit, and the integrity of the purified
protein was assessed by SDS-PAGE.
In vitro transcription analysis. In vitro transcription reactions were performed
as previously described (31). Briefly, core RNA polymerase, ?A, and Mga were
incubated together on ice for 10 min. Transcription buffer, RNaseOUT (Invitro-
gen), linearized plasmid template, and distilled H2O were added to a final
volume of 40 ?l, and each mixture was incubated at 37°C for 10 min. The reaction
was started by the addition of ribonucleotides (including [?32P]CTP) and incu-
bated at 37°C for 2 min. Heparin (10 ?g) was added to inhibit reinitiation of
transcription, and each mixture was incubated at 37°C for 5 min. The reaction
products were then chased with 1 ?l of 100 mM cold CTP and incubated at 37°C
for 5 min. Products were ethanol precipitated, resuspended in formamide gel
loading dye, and run on a 6% denaturing polyacrylamide gel (Amresco), and gels
were processed as described above.
RESULTS
The sof-sfbX promoter contains a single distal Mga-binding
site. Although expression of the cotranscribed sof, encoding
serum opacity factor, and sfbX, encoding a fibronectin-binding
protein, are Mga dependent (20, 22), a direct role for Mga in
this process has yet to be established. Using sequence from the
chromosomal locus of sof-sfbX in a serotype M12 strain of the
GAS (11), a region homologous to the Mga-binding site con-
sensus (16) was found approximately 278 bp upstream of a
predicted start of transcription that was identified on the basis
of its similarity to conserved regions of the E. coli ?70promoter
(Fig. 1A). A comparison of the sof-sfbX promoter regions from
this M12 strain, a recently sequenced OF?M28 strain (8), and
the sequence of an OF?M22 strain available in the laboratory
determined that the putative Mga-binding site and its distance
to the predicted start of transcription were highly conserved
among several OF?strains. The location of the Mga-binding
site is reminiscent of that found in the promoter of sclA (1),
suggesting that Mga may function to activate transcription of
these two promoters by a similar mechanism.
TABLE 2. PCR primers used in this study
TargetPrimerSequence (5?–3?)Reference
NonspecificrpsL-1
rpsL-2
Pemm-L1
Pemm-R1
Pemm1-L1
Pemm1-R1
OYL-25
OYR-25
GAS-rpsL1
GAS-rpsL7
PsclA-L2
PsclA-R
C5-L1
C5-R1
Psof-L3
Psof-L4
Psof-L5
Psof-L6
Psof-R1
Psof-R2
Psof-R3
Psof-R4
Psof-R5
Psof-R6
GAATGTAGATGCCTACAATTAACCA
GTGCGCCACGAACGATATG
GCATGGATCCCATCGCAAAGAGCTTA
GCGGCTCGAGTAGTGTCTATTCGTGTTATT
AAGCGATTATTGACAAGCTC
AGTCAAAGCTACCGCTACTG
TACCATAAAATACCTTTC
GGTTGTACCATAACAGTC
TGTCTAAAATCACATCTTCGA
TTACGTACCAACTGGTTAATT
TGGTAATCGTAATTGTCTGC
ATGGTGCTTTGATGTCAACA
AAGAATGAGATTAAGGAGGTCACA
GCGCAATGGCAAGTTTGTC
TTTGGTCTCAGACGGCGCCA
AATCAAGGTCACAAACTTTC
AATCGTCTGATAAGCTCTTG
CACTCAGTTGCCTCCAAGCT
ACCTCCTGTCCTAAAACTGT
GAAAAAAGGGGTTGCCGCGA
GTCAATGTTTCGTTAACCTT
GGAAAGTTTGTGACCTTGAT
CGTCTGAGACCAAAGCGTTA
GCTGCGCAAGGAGAAGCTCA
17
17
17
17
1
1
17
17
This study
This study
1
1
16
16
This study
This study
This study
This study
This study
This study
This study
This study
This study
This study
Pemm
Pemm1
Pmga
PrpsL
PsclA
PscpA
Psof
2040ALMENGOR ET AL. J. BACTERIOL.

Page 4

Identification of the Mga-regulated transcriptional start
site of sof-sfbX. In order to confirm the location of the Mga-
binding site in the sof-sfbX promoter, its transcriptional start
site was mapped by a primer extension analysis using total
RNA from the OF?M22 strain AL168str-r and its Mga?
derivative nr4. Using the antisense primer Psof-R1 (Table 2)
that is located at the 5? end of sof-sfbX in the extension reac-
tion mixture, a doublet product was clearly observed in the
Mga?strain AL168str-r that was noticeably absent from the
Mga?strain nr4 (Fig. 1B). A repeat of the primer extension
with a second antisense primer Psof-R5 (data not shown) also
showed a doublet at this location and confirmed it as the true
Mga-regulated transcriptional start site. Notably, the ?10 and
?35 sequences (Fig. 1A) of this site are homologous to estab-
lished E. coli consensus sequences (six/six nucleotides and four/
six nucleotides, respectively).
Specific binding of Mga-His to Psof in vitro. A series of
electrophoretic mobility shift assays were performed to deter-
mine whether Mga binds in vitro to the putative binding site in
the sof-sfbX promoter. Increasing concentrations of Mga-His
were incubated with overlapping promoter probes that were
PCR amplified from the OF?serotype M4 strain AP4. Ini-
tially, an M4 strain was chosen for EMSA analysis to allow use
of purified M4 Mga-His, which was available in the laboratory.
However, because of technical difficulties in purifying M4
Mga-His to sufficient levels for EMSA analysis, the established
M6 Mga-His (1, 29) was used in future experiments. The ability
of M6 Mga-His to bind to Psof was found to be similar to that
of M4 Mga-His (data not shown), validating its use.
A summary of the probes used and the ability of Mga to bind
those probes are shown in Fig. 2A. The promoter fragments
that shifted upon addition of protein helped to delineate an
FIG. 1. Identification of an Mga-regulated transcriptional start site for sof-sfbX. (A) Sequence of sof-sfbX promoter region from serotype M22
indicating the transcriptional start site (asterisk and black arrow), ?10 and ?35 regions (solid bars), the putative Mga-binding site (darkened box),
and the sof start of translation (gray arrow). The sequence is numbered using the Psof-sfbX transcriptional start site as ?1. (B) Primer extension
analysis was performed on total RNA isolated from serotype M22 AL168str-r (Mga?) and nr4 (Mga?) using the radiolabeled antisense primer
Psof-R1 (Table 2) as described in Materials and Methods. The start of transcription for sof-sfbX (asterisk) and its corresponding ?10 hexamer are
shown.
VOL. 188, 2006Mga IS SUFFICIENT TO ACTIVATE GENES IN S. PYOGENES 2041

Page 5

FIG. 2. Identification of a specific Mga-binding site in the sof-sfbX promoter region. The ability of M6 Mga-His to bind to DNA within the
sof-sfbX promoter was assessed by an EMSA. (A) Region surrounding sof-sfbX (hatched arrows) in the GAS genome, including a protein export
protein (prsA) and a hypothetical cytosolic protein (Spy2036 in the SF370 M1 genome) (black arrows). Overlapping promoter probes in relation
to the transcriptional start site of sof-sfbX (circle with arrow) are shown by the black lines. The ability of Mga-His to bind to the probes is indicated
by both a plus sign and shaded region. The putative Mga-binding site (gray box) and a possible second low-affinity binding site (dashed box) are
denoted on the chromosomal diagram. (B) EMSAs on Psof promoter probes (Table 2). Constant amounts (1 to 2 ng) of labeled promoter probes
were incubated with increasing amounts (1.0 to 2.0 ?g) of Mga-His for 15 min at 16°C prior to separation on a 5% polyacrylamide gel.
2042 ALMENGOR ET AL.J. BACTERIOL.