135
Anti-sigma
factors
John D Helmann
Anti-o factors
modulate
controlled
by alternative
regulated
by either
the expression
of numerous
regulons
0 factors.
Anti-a factors
are themselves
secretion
from
through
the hook-basal
body),
(i.e. phosphorylation
regulated
interaction
effecters
with
proteins
molecule
regulators
function
rs factors).
Recent
description
of the opposed
highlights
o/anti-o
detailed
in Bacillus
determination
dissection
for an anti-o
of two
complex
of extracytoplasmic
the genetic
surfaces;
the
destabilization
to 028 binding;
antagonist;
and the
the first
and the
partner-switching
modules
subtilis.
Addresses
Section of Microbiology,
Wing Hall, Cornell
14853-8101,
USA; e-mail: jdh9@cornelLedu
Current
or small
include
binding
unexpected
role of FlgM in holoenzyme
finding that folding
of FlgM is coupled
structure
export
by an anti-anti-o
modules),
or
extracytoplasmic
(i.e. transrnembrane
the cell (i.e. FlgM
sequestration
partner-switching
Opinion
in Microbiology
1999,
University,
2:135-l
Ithaca,
NY
41
http://biomednet.com/elecref/1369527400200135
Q Elsevier
Science
Ltd ISSN
Abbreviations
AA
SpollAA
AB
SpollAB
ECF
extracytoplasmic
1369-5274
function
general stress response). In this review, I summarize
progressin the characterization of these and related systems,
with an emphasison resultsreported in 1997and 1998.
Regulators interacting with Escherichia
co/i 070
The first demonstration of an inhibitor protein binding
to (3 emerged from studies of RNA polymerase modification during phage T4 infection (reviewed in [4]). This
anti-o factor, AsiA, associatestightly with E. coli 070and
inhibits transcription from both host promoters and early,
o70-dependent phage promoters. Because T4 late transcription depends on a phage-encoded alternative r~
(o~P”~), it initially appeared that AsiA might function in
0 switching. The story, however, is not this simple,
because AsiA is also required for the activation of phage
middle genes, which depend on 07~ together with the
phage-encoded h4otA protein [.5,6].
Recent biochemical analyseshave clarified this mechanism.
The small 10.6 kDa AsiA protein binds in a 1:l complex
with the carboxy-terminal 63 amino acids(conserved region
4) of 070and thereby blocks recognition of the -35 promoter region [7’,8”]. AsiA modified holoenzyme recognizes
middle phage promoters containing bound MotA and lacking a -35 recognition region [5,6,9,10’]. It is not yet known
whether AsiA alsobinds to MotA, and whether this interaction is important for middle gene transcription.
Introduction
Promoter recognition in bacteria requires that RNA polymerase core enzyme @Q’cQ) associateswith a sigma (6)
subunit to form a holoenzyme [1,2]. Most bacteria contain
multiple CJfactors,including both a primary (3factor, controlling essential housekeeping functions, and alternative (T
factors, activated by specific signals or stressconditions.
Most rs factors (the 07’Jfamily) are related in sequenceand
presumably in structure. The exceptions are members of
the 054family, often controlling aspectsof nitrogen metabolism, which are structurally and functionally distinct.
Genome sequencing has revealed that the numbers of (r
subunits vary greatly between different bacteria:
Mycoplasmn genittzliumcontains a single CYfactor, whereas
most other bacterial genomes encode at least three. The
proliferation of d factor paralogsis particularly noteworthy in
the Gram-positi\:e lineage including 17 in Badus
subtilis
and 13in Mycobacm-hm tuberculosis. Several of these are associated with known or suspectedanti-o factors.
Alternative (3 factors are regulated at the transcriptional,
translational, and post-translational levels. One common
mechanismis the reversible interaction of r~ and a protein
inhibitor, designatedan anti-o (reviewed in [3’]). Anti-o factors figure prominently in regulons controlled by 028(i.e.
flagellar biosynthesis), extracytoplasmic function (ECF) CT
factors, and B. szlbtil’as
OF (i.e. early sporulation) and & (i.e.
Together with previous studies, these results indicate
that phage T4 has evolved the ability to recruit and modify the host RNA polymerase for the transcription of both
early and middle phage genes, prior to the eventual
c@5-dependent activation of late genes. In the first
stage of infection, T4 ADP-ribosylates the RNA polymerase a subunits which greatly reduces transcription
from strong host promoters that require an upstream (UP)
element for optimal activity (cited in [7’]). This increases the pool of RNA polymerase available to transcribe
phage early genes. In the second step, AsiA selectively
inhibits transcription from other o70-dependent promoters by blocking recognition of the -35 element.
Remarkably, a variant of this mechanismmay alsofunction
in uninfected E. coLi.Biochemical fractionation of o7o-associated proteins from stationary phase cells identified a
single, specifically associatedpolypeptide designated Rsd
[11’1.Rsd binds to region 4 of 070and appearsto block association with RNA polymerase core enzyme. The effects of
Rsd on in a&o transcription are modest
and appear to be
promoter specific, with a maximum observed inhibition of
four fold. Since the levels of Rsd in v&,>oare only sufficient
to complex 20% of 0 ‘O, the role of this protein is not immediately obvious. Ongoing generic analysesof rsd mutants
will be needed to establishthe role of Rsd in vjwo.
136
Figure
Cell regulation
identified that leads to conditional (ionic strength sensitive) FlgM secretion [18], which further supports the
notion that secretion of flagellar proteins occurs through
the hollow channel inside the hook and flagellar filament.
1
Significantly, high resolution NMR experiments have
establishedthat FlgM is largely unfolded in solution [ 19”]
with, at best, transient helical elements [ZO’]. Upon interaction with &a, the carboxy-terminal region of FIgM
becomesordered [19”]. It is postulated that this unfolded
state may facilitate the passageof FlgM through the narrow channel of the hook-basal body structure.
The mechanismby which FlgM inhibits the activity of the
02s RNA polymeraseinvolves both the sequestrationof 02s
and the destabilization of existing 02~ holoenzyme [Zl”].
Using surface plasmonresonance,the equilibrium dissociation constant of 02s and FlgM has been estimated at
-2 x lo-10M [Zlo’]. For unknown reasons,this value is much
lower than that inferred from NMR analyses[19”]. Genetic
experiments suggestthat multiple sitesof 028,involving portions of both regions2 and 4, bind to FlgM [Zl”].
Current
Opmion in Microbiology
Regulation
of late flagellar
genes
by export
of the FlgM anti-o.
FlgM
(square)
inhibits
the activity
of’o*s
either
by binding
to free cr*s or by
[21”*].
When
bound
to
inducing
dissociation
of the o 2s holoenzyme
o*s, FlgM adopts
a partially
folded
structure
(diamond).
Once the hook
and basal body structure
is complete,
FlgM can be exported
from the
cell by the flagellar
export
machinery
114-l
6). FlgM passes
through
the hollow
inner core of the basal-body,
hook, and flagellar
filament
structures
[la).
Once o*s transcription
commences,
flagellin
becomes
a competing
substrate
for this export
channel
and FlgM levels
in the
cell again begin
to rise.
FlgM and the regulation
gene expression
of flagellar
One of the most dramatic examples of anti-o factor regulation emerged from analyses of flagellar biogenesis in
SahoneLa
typRimuri2lm. The synthesis of late flagellar
genes (encoding flagellin and chemotaxis functions)
depends on an alternative (J factor of the 02s subfamily
(reviewed in [lZ]). During the early stages of flagellar
biogenesis, oz8 is held inactive by tight association with
FlgM [13]. Once the hook and basal body is assembled,
FlgM is secreted by the flagellar export system thereby
freeing active 028 [14-161 (Figure 1).
The mechanisms acting to couple secretion of FlgM to
completion of the hook-basal body structure are complex.
Genetic studies implicate two proteins, FhlB and RflH, as
molecular ‘gatekeepers’ that act to sense, by an as yet
unknown mechanism, the completion of the hook [17’].
Even after the hook is complete and FlgM export commences, the FliD, FliS, and FliT proteins still partially
inhibit export, and thereby prevent the complete derepressionof &-dependent flagellar late genes. Finally,JgE
(encoding hook protein) missensemutations have been
02~homologsare widely, but not universally, present in flagellated bacteria [12], and many of these systems contain
FlgM. Although the details differ substantially, f&M is
under complex control often including a &-dependent
promoter that provides an auto-repressing circuitry. In S.
~phimzdrium,f1gM is also regulated at the translational level
by Flk, which acts asa checkpoint for flagellar ring assembly [Z&23]. Thus, the mechanisms of morphogenetic
coupling that coordinate flagellar assembly include the
control of FlgM synthesis (by Flk) in addition to the previously mentioned hook assemblycheckpoint that controls
FlgM export. In B. subtilis, f/gM is downstream of a competence operon [24]. As a result, JgM is expressed under
partial control of the ComK transcription factor and this, in
part, accounts for the mutually antagonistic expression of
the motility and competence regulons.
.Regulation
of anti-a factors
partner-switching
modules
by
The B. subtilis crF and & regulons are both modulated by
partner-switching modules involving the mutually exclusive binding of an anti-o to either of two partners: the
corresponding CYor an anti-anti-o [ZS]. Although CJFfunctions early during sporulation [26] and on regulates a
complex stress response [27], both these cs factors and
their regulators are paralogs.
Regulation
of d activity during
sporulation
Prior to sporulation-specific septation, @ is held in an
inactive complex with its corresponding anti-o (SpoIIAB;
AB for short) via contacts to three conserved CJfactor
regions: 2.1, 3.1 and 4.1 [ZS]. Active crF is released when
AB instead complexes with the antagonist (anti-anti-o)
protein SpoIIAA (AA). In work reviewed elsewhere [26], it
was established that the binding of AA to AB is stabilized
Anti-sigma
by ADP, whereas the ABeoF complex is stabilized by ATP.
In addition, AB acts as a protein kinase to phosphorylate
AA, Phosphorylation inactivates AA thereby allowing AB
to inhibit OF. On.ce the sporulation septum forms, AA is
dephosphorylatecl by the membrane-bound SpoIIE phosphatase. This allows AA to bind AB and is a key event in
the release of active CJFin the forespore.
Figure
factors
Helmann
137
2
Recent studies provide a more detailed picture of this partner-switching
module and support an induced release
mechanism for ~1; activation (Figure 2). First, kinetic studies have shown that phosphorylation of AA by AB is very
slow (0.005 set-1 [29”,30’]). Once phosphotransfer occurs,
AA-P dissociates at a relatively rapid rate (0.017 set-l) but
the remaining AB*ADP complex, designated AB”, is only
slowly recycled (0.0002 set-1; a one hour half-life!) [30’].
This kinetic bottleneck may correspond to the accumulation of stable AA*AB+ADP
complexes [31”].
Unexpectedly,
studies of AB mutants altered near the
ATP-binding domain indicate that kinase activity is needed during the process of AA-mediated
release of CJF
(Figure 2). In this induced release model [31”,32], AA is
phosphorylated directly by the AB~ATPwF
ternary complex, leading to an unstable AB*ADP*oF
complex that
dissociates to liberate OF and AB*ADl? The AB*ADP is
ultimately trapped in a long-lived complex with AA.
Interestingly, genetic experiments have revealed that certain missense mutations in the amino-terminal region of
AB specifically impair binding to AA, whereas alanine substitutions at the same positions affect binding to both AA
and OF [33’]. This suggests that the binding surface used
by AB to bind AA overlaps with that used to bind OF.
Further work will be needed to sort out the remaining
details in this mechanism. Specifically, a direct comparison
of the slow rate of AA phosphorylation by AB*ATP with that
of the AB*ATP*oF ternary complex is needed. In addition,
AB is a dimer [34], but AA and CJFare not. Thus, one can
imagine that two AA molecules are needed to fully discharge CJFfrom an AB,*ZATP%?oF complex (Figure 2). The
resulting AB,*2ADP complex can either undergo ADP/ATP
exchange reactions and rebind oF, or bind another two AA
monomers to form the ‘dead-end’ AB,*ZAA*ZADP
complex. Inspection of this pathway suggests that inactivation of
the dimeric anti-o, AB,, requires net consumption of four
unphosphorylated AA monomers. Therefore, activation of
GF may be highly cooperative with respect to the concentration of unphosphorylated AA generated by SpoIIE.
Recent inroads have also been made into the structural
analysis of these regulatory components. NMR spectroscopy
has allowed the three-dimensional structure of AA to be visualized [35”]. This small protein contains a four-stranded p
sheet flanked by four a helices. This structural information,
together with genetic data on AA and AB mutant proteins,
provides a prelirninary glimpse of the likely interaction surfaces (e.g. [33-l). Further structural analyses of AA, and other
Regulation
of early forespore
gene expression
by the SpollAB
anti-o. In
the predivisional
cell, oF is held in an inactive complex with the dimeric
SpollAB
(AB; rectangle) anti-a and ATP (261. The anti-anti-o,
SpollAA
(AA), is presumably
phosphorylated
and therefore
inactive. As
sporulation
proceeds,
the SpollE phosphatase
triggers a rise in
unphosphotylated
AA levels and the resulting AA binds to the
AB*ATP*oF complex to induce the release of active oF [32]. This
induced release reaction requires the kinase activity of AB, suggesting
that AA phosphorylation
and aF release are coupled events [31**1. AB
is subsequently
trapped as a long-lived species containing
ADP
[29”,321.
This is presumably
an altered conformation
of AB (AB*)
[30*] and may also include AA (the AA*ADP*AB
complex)
[31”].
components of this and related systems, will allow the biochemical basis of partner switching to be elucidated.
Regulation
of the &-controlled
stress
responses
Remarkably, the regulation of gB, controlling the expression of more than 60 proteins in response to several
stress conditions (reviewed in [27]), is even more complex than that described for CY~.Activation of the CSB
regulon is governed by two, sequentially linked partnerswitching modules [36,37]. The regulatory components
are encoded in a large, complex operon containing sigB
and seven regulator of sigma-B (rsb) genes (Figure 3).
The output of the upstream module, which integrates
environmental stress signals [38], is RsbT, which stimulates oB activity indirectly by activating the RsbU
phosphatase of the downstream module. Interestingly,
with the significant exception of RsbX (see below), there
is correspondence between the genetic and functional
order of these genes.
138
Cell regulation
Figure 3
rsbR
rsbS
rsbT
rsbU
r.sbV
Downstream
Upstream
rsbW
rsbX
sig8
module
module
Current
Recent genetic and biochemical analyses have confirmed and extended the essential features of the dual
partner-switching model [36,37]. Genetic analyses of
&X mutants, and their suppressors, confirm that this
gene encodes a negative regulator of ~$2 that acts
upstream of RsbU and RsbT but that is not required for
the response to at least some stresses[39’,40]. As originally proposed 1371,RsbX appearsto act asa homeostatic
feedback signal that functions to downregulate gB.
Thus, signals that liberate active bB from its anti-o
(RsbW) also lead to increased transcription of the genes
for the downstream module from an internal, autoregulatory promoter. This up-regulation of crB levels may
initially magnify the response, but the concomitant upregulation of RsbX will act to reduce the signal
generated by the upstream module. Specifically, RsbX
dephosphorylates RsbS that thereby sequesters RsbT
preventing activation of the RsbU phosphatase.
Intriguingly, mutational studies of RsbT reveal that
kinase activity is not essential for stimulation of RsbU,
but is required for the transmission of stress signals to
the downstream module [41’].
Recent analyseshave alsoshedlight on the heretofore mysterious RsbR protein [42’]. RsbR appears to act as a
modulator of RsbS activity in the upstream module and
representsa new branch in this signaltransduction cascade.
Anti-c
factors
that regulate
ECF CTfactors
ECF CJfactors [43] area largeand growing subfamily of 0 factors that typically regulate functions related, in the broadest
sense,to the cell envelope (see [44] for a recent review).
Typically, the genes encoding ECF (J factors are positively
Opinion
in MicrobiDl~y
Regulation of the & stress response by two
coupled partner-switching
modules. The Bach
subtilis sigB gene is co-transcribed
with seven
regulatory genes from a &dependent
promoter
(PA). Genes that positively affect oB activity (and
s&S itself) are shaded grey, whereas negative
regulators are white. The downstream
portion of
the operon, which can also be expressed from a
&dependent
~oregu~to~
promoter (P&
encodes the downstream
partner-switching
module [36,371. The RsbW protein (W)
functions as an anti-o and binds to either 06 or
to RsbV &). W also functions as a protein
kinase and phosphorylates
V. This reaction is
postulated to sense the energy status of the cell:
when ATP levels drop, W will become
sequestered
by V and the oB regulon will be
induced [38,3*].
The upstream module
(proteins R, S, and T) controls the activity of the
U phosphatase,
which can regenerate active V
protein [36,371. T is an allosteric activator of U
and also has protein kinase activity targeted
against its antagonist S [41 *I. Active S, in turn,
can be regenerated
by X, which serves to
downregulate
signals from the upstream
module 1401.
autoregulated and are transcriptionally coupled to the
expressionof a cognate anti-cr factor. In E. co& ECF (Tfactors
include &=I, an activator of ferric citrate transport (reviewed
in [45*]), and (TE,a mediator of a periplasmicstressresponse.
ECF (Sfactorsare alsowell representedin other bacteriawith
seven in A. subtilisand 10 in M. tuberculosis.
Most ECF CTfactors are regulated by anti-o factors. In the
absenceof an external signal, the CTis held in an inactive,
stoichiometric complex with an anti-o, often located in
the cytoplasmic membrane. By virtue of its transmembrane disposition, the anti-o is poised to activate a
transcriptional responsesignaled by the presence of molecules external to the cytoplasm&z membrane. Systems
regulated in this manner include various stressresponses,
uptake systems for ferri-siderophore
complexes,
carotenoid biosynthesis in response to light, and synthesisof the exopolysaccharide alginate 1441.
Several recent studies have investigated the topological
arrangement of anti-o factors in the cytoplasmic membrane as well astheir interactions with regulatory ligands.
In E. coli, ferric citrate transport requires crFecr, which is
regulated by FecR. Although initially thought to function
as an anti-o, some results suggest that FecR may be
required to activate crFecl[45’]. FecR appears to have a
single membrane spanning segment with a cytoplasmic
amino-terminal domain and a periplasmic carboxy-terminai domain 1461.Activation of FecR requires that the
periplasmic domain interacts with the amino-terminal
domain of the outer-membrane FecA transporter [47’]. In
B. subt~Z~s,
C+ is aiso regulated by an anti+ factor 148,491.
Despite the report that this (5 activates a @ecr-dependent
Anti-sigma
gene in E. co& [48], in 8. subtills ox modulates cell wall
structure and does not control iron transport [SO]. The B.
5~~~~~~s&+’ regulon is also regulated by an anti-cr that senses perturbations
‘of cell wall structure (JD Helmann,
unpublished data).
Ouhammouch
Arri-ECF
o flicro~ respond co a wide variety of s&u&
transmitted either by other proteins or by small molecules.
In the case of E. LX& GE, release from the transmembrane
anti-o (RseA) requires signals perceived in the periplasm
that may be modulated bq’ RseB and RseC [51*,X*}. In a
CICDS&,Yreiateb ywiem T~~&kng
aY&natf: s~~&on
in
~se~~o~o#as
aerqkosa,
a & homolog (022) is regulated by
an anti-0 MucA. MucA contains a single transmembrane
domain and interacts with the periplasmic Mu& protein
[53]. Not aitl ECF ‘0 factors are controlled by signalsexternal to the cell: the 8. co&olor gR regulon is induced by
oxidants perceived by the redox activity of a cytoplasmic
anti-c factor (cited in [54’]).
25b:235-Z~ir4B.
SR, Orsini
Hinton
DM,
March-Amegadzie
Severinova
E, Severinov
RNA polymerase
279:9-l 0.
8.
l
*
R, Gerber
K, Darst
by bacteriophage
Colland
F, Orsini
G, Brady
JS, Sharma
H, Kolb
A: The bacteriophage
within
Pahari
D: Interaction
of bacteriophage
14 AsiA protein
coli sigma70 and its variant. FEBS Leti 1997,
S, Chatterji
10.
Adelman
.
between
subunit
K, Orsini
1I I.
l
A, Grariani
L, Brady
that the Asi&070
that AsiA alters
complex
oromoter
Helmann
is identified
in stationary
JD: Alternative
W#s’ks
J Biol
of special
interest
**of outstanding
interest
1.
2.
Helmann
JD: Bacterial
and Regulation.
Edited
Raven Press;
1994:1-l
Lonetto
M, Gribskov
conservetion
sigma factors.
by Conaway
7.
forms with 1:l stoichiomeselectivity
at the initial bind-
m
ev.@m.
phase
E. co/i
M, Gross
andi evolutionary
CA: The u70 family: sequence
relationships.
J Bacterial
1992,
Rev
Aizawa
Microbial
SI: Flagellar
1996,
Hughes
assembly
in Salmonella
typhimurium.
KT, Gillen
KL, Semon
MJ, Karlinksy
JE: Sensing
Kutsukake
machinerv
~~hjrnu~~rn
K: Hook-length
control of the export-switching
involves a double-locked
gate in Se&none&
flagelfar mor&ogenesis.
f Bacterial
1997,
Saito
T, Ueno
1998,27:1129-
Brody EN, Kassavetis
GA, Ouhammouch
M, Sanders
GM, Tinker RL,
Geiduschek
EP: Old phage, new insights: two recently recognized
mechanisms
of transcriptional
reguiation
in bacteriophage
T4
development
FEMS ~icrobiaf
Letll995,
128:1-6.
structural
of a
17.
.
T, Kubori
T, Yamaguchi
filament elongation
can be impaired
protein FlgE of Sa~m~e//a
~phi~urium:
hook as a passage for the anti-sigma
,1F!
l
4.
by export
280.
Kutsukake
K: Excretion
of the anti-sigma
factor through a flagellar
substructure
couples flagellar gene expression
with flagellar
assembly
in Salmonella
fyphimurium.
MO/ Gen Genet 1994,
24’5-%m-%7 2.
1%.
of the
MO/
19:1-5.
16.
~icrob;of
of many
as a a7o-bind-
14.
174:3%43-3849.
3.
Hughes
KT, Mathee
K: The anti-sigma
factors. Annu
.
199%, 52:231-26%.
This review provides
a current
and comprehensive
account
wul:rslar;rraPJrifflr:~ii;,~yn~tkrjiuR.
cells
sigma factors and the regulation
of
IbrJ, !bim-9kd
‘&iv, , ‘%2%7%2%%2.
179:126%-l
273.
A genetic
study revealing
that mutation
of both
required
to allow FlgM export through
the basal
hook structure.
In Transcription:
Mechanisms
RC, Conway
J. New York:
1997,
XutsuWke X, lyoda S, Dhti~sh’~ X, YmoT : Gene&
and mojecular
analyses of the interaction
between the flagellum-shirt
sigma
and antirs@~
fa&a<s in S&man&a
~~~~~u~~u~.
EM%0 I tQ04,
13:456%-4576.
of review.,
l
Chem
13.
reading
period
EN: The interaction
4 and the sigma‘ll)
)ishaga
M, \&ihama
A: R stationary
phase protein in
Escheritfik
coKwifh
bindfng
acfivify fo fhe ma[or sigma
subunit of RNA polymerase.
Proc Nafl Acad Sci USA 199%,
95:4953-495%.
The Rsd protein
ing protein.
12.
G, Kolb
WIT l&IA oroWn 01 ba&eriophageI
of Escberkhia
coti RNA polymerase.
272:27435-27443.
The authors
&tablish
try and demonstrates
yammer-switching
fhe annual
T4
switch for sigma 70-dependent
1998,27:8~
9-829.
in&mediates
in bacterial flageffar assembly
negative regulator. Science 1993,262:1277-l
published
M:
made at a
of the T4
70 binding
MO! 8ioll996,
SA: lnh~b~jon of Escherichie
c&i
T4 A.&A. I Moi Bioif 996,
EN, But
AsiA protein: a molecular
promoters
MO/ Micmbio/
with Escherichk
411:60-62.
15.
hisrest,.
as:
EN:
This paper
demonstrates
that AsiA modifies
holoenzyme
without
dissociation of 0~0. The modified
enzyme
is blocked
from initiating
transcription
from the RNA 7 promoter
(containing
-35 and -10 regions)
but not from
galPI,
Elegant
protein
footprinting
studies,
with labeled
0~0, map sites of
AsiA protection
to region
4 and demonstrate
that A&A binding
modifies
the conformation
of region
2.
modules as found in the gF and 6 regulons, and interactkn wirh sma’t’mokcB\e 01 proxh
‘tjgaf&s
as Sound
j,
systemsreguiated by ECF d factors.
Papers of pariicular
have been highlighted
139
In vitro studies
demonstrate
that AsiA modiied
RNA polymerase
can bind and
tib~e
trmjht,
frm an exw-ded
-10 PpmE?
s?& does r&cd rzq.ire
rewgnition of the -35 element
kaPl1 but not from two promoters
Cr7 Al and A21
&i*,aqafie
hi+ a -35&r&a
- M~~~~~*e~
&w cess&
Iti @$?
Genomic analyseshighlight the widespread occurrence of
bcth alcernaciueCTEactocsand anti-cr &ccors as cegulators
of gene expression. The systemsreviewed range from the
relatively simple, the secretion of the anti-@ FlgM as a
mechanism to sense the completion of the flageliar
hcodnlr>as&'DD~,Y SX~XXW& ‘so -$ne‘>ncreti+Y
d>d~D~Iate
signal transduction cascade ulrimately regulating the
RsbS an&c~ factor. The mecZlanismsbj? whic’n anti-b
factors exerr their effects are still an active area of study,
but seem to involve both sequestration and holoenzyme
dissociation. To date, the pathways serving to control
arri-cr acrivicv can be broadly grouped into three care-
References and recommended
G, Brody
Characterization
of pre-transcription
complexes
bacteriophage
14 middle promoter:
involvement
MotA activator
and the T4 AsiA protein, a sigma
protein, in the formation
of the open complex.
J
9.
i%~m the cell (i.e. FlgM{>
K, Harvey
Helmann
Bacteriophage
T4 MotA and AsiA proteins suffice to direct
Escherichia
co/i RNA polymerase
to initiate transcription
at T4
middle promoters.
Proc i&f/ Acad Sci USA i 995,92:1451-l
455.
Conclusions
gcries: expm
M, Adelman
factors
*
&2q!w+L?rilG~P!,
the f/I18 and rf/ff genes
is
body in the absence
of the
SI: Flagellar
by mutations
in the hook
a possible role of the
factor FIgM. MO/ ~;crob;ol
S, line T, Aizawa
1139.
C~~~;r~~,rYar}i~;~~.
The C-terminal
half of the anti-sigma
structured
when bound to its tamet,
Mq$rexU;T,&&q&$3.k!-:
factor, FIgM, becomes
sigma 28. Nat Sfruct B;o/
1997,4:285-291.
NMR studies
reveal that FigM is mostly
unfolded
in solution
but
partially
folded
structure
upon binding
to ~12s. The binding
interaction
adopts
a
affects
140
Cell regulation
residues
of FlgM
20.
l
in the carboxy-terminal
region
mutants
impaired
in binding.
Daughdrill
GW,
21.
*
l
with
LJ, Dahlqui~
FW: The C-terminal
flghl contains a dynamic equilibrium
helical conformations.
Biochemistry
Hanely
anti-sigma
factor
structure favoring
37:1078-l 082.
In extension
alpha-heii~i
of 028.
of FIgM, consistent
of the results
reported
in [I 9”1,
structure
in the ~~~-term~nal
the location
required
for oF binding,
on AA and oF binding.
half of the
solution
34.
Duncan
L, Losick
R: SpollAB
is an anti-sigma
factor that binds
to and inhibits transcription
by regulatory
protein sigma F
from Bacillus
subtilk
Proc Nat/ Acad Sci USA 1993,
90123252329.
35.
*
Kovacs H, Comfort
D, Lord M, Campbelt
ID, Yudkin MD: Solution
structure of SpollAA, a pho~ho~atable
component
of the
system that regulates transcription
factor sigmaF of Bacillus
subtilis.
Proc Nat/ Acad Sci USA 1998, 95:5067-5071,
1996,
FlgM is found to have
region
even in the
transient
absence
l
KT: The flagellar
Chadsey
MS, Karlinsey
JE, Hughes
factor
FlgM actively dissociates
sigma28 RNA polymerase
12:3123-3136.
Salmonella
hoioenzyme.
anti-sigma
typhimurium
Genes
22.
Karlinsey
JE, Tsui HCT, Winkier
translation
J Bacterial
23.
Karlinsey
JE, Pease
gene of Saimonelta
AJ, Winkler
~~imurium
JL, Hughes
26.
Stragier
subtilis.
27.
Hecker
The sigma
factors
of i3atillus
subfi/is.
.
Microbial
of sporulation
in Bacillus
AL, Losick R: Three sites of contact between
transcription
factor sigmaF and its antisigma
SpoliAB. Genes Dev 1996,10:2348-2358.
T, Lord
and SpollAA
Najafi
Contribution
of partner
to regulation
of sigmeF activity
M, Yudkin
cycling
reaction
MD:
DA, Yudkin
the Bacillus
factor
DA, Duncan
L, Paskowitz
DM, Losick
0.
the antisigma factor SpollAB is required for activation
inhibition of tran~~ption
factor of during striation
subfilis.
J MO/ Bioi 1998,284:569-578.
DM, Duncan
42.
.
.
common
sites of contact between
SpotlAB
and its partners
SpoIlA&
transcription
factor UF in BaciUus
subfilis.
R: Evidence
43.
N, Scott
obtained
that
are apparently
J, Voelker
U, Haldenwang
sigB
inhibit
also
regulator
that RsbX
is
module
in the
Isolation and
mutations
that
RsbX. J Bacterial
CW: Serine kinase activity of a Bacillus
is required to transduce
envi~men~l
but not to activate its target PPPC phosphatase.
Kang CM, Vijay K, Price
subtitis
switch protein
Akbar
S, Kang
CM,
Gaidenko
TA, Price
CW:
Modulator
protein RsbR regulates
environmental
signalling
stress pathway
of Beciifus
subtitis.
Mot Microbioi
24~567-578.
in the general
1997,
Lonetto
MA, Brown
KL, Rudd KE, Buttner
MJ: Analysis of the
Streptomyces
coelicolor
sigE gene reveals the existence
of a
subfamily of eubacterial
ofactors
extracytoplasmic
functions. Proc
9117573-7577.
44.
Missiakas
D, Raina
role and regulation.
45.
Braun
l
starting
A review
46.
Welz
47.
involved
Natl Acad
of the FecR-Feel
D, Braun
in the regulation
Sci USA
system
regulating
ferric
inbox
of
1994,
S: The extracytoplasmic
function sigma
MO/ Microbial
1998,28:1059-l
066.
V: Surface signaling: novel tmn~~ption
from the ceil surface. Arch Microbioi
functional
J Bacferiof
.
J MO/ Biol1998,
with [31 *I, AB mutants
were
many of the same residues
Smirnova
the idea
(upstream)
Genetic
experiments
demonstrate
that the RsbR protein
affects
the activity of the upstream
partner-switching
module
but is not required
for sig
naling.
RsbR
interacts
with the RsbS
antagonist
and mutation
of a
possible
phosphorylation
site CT2051 suggests
that RsbS is phosphorylated by the RsbT kinase.
activity of
as well as
in f&&t/us
for
the antisigma
factor
and the developmental
L, Losick
984.
MO/ Microbial
1998,
30:189-196.
The authors
engineer
a point mutation
into the ATP-binding
region
of the
RsbT protein
kinase and demonstrate
that this protein
is still competent
to
allosterically
activate
the RsbU phosphatase
that connects
the upstream
partner-switching
module
to the downstream
one.
of
L, Alper S, Losick R: SpollAA governs
the release of the
cell-type specific transcription
factor sigma F from its anti-sigma
factor Spoil&B. J MO/ Viol 1996, 260:147-l 64.
DA, Paskowitz
A, Haldenwang
WG:
subtilis
in
J Bacterial
Voelker
U, Luo T, Smirnova
N, Haldenwang
W: Stress
activation
of Bacillus
subtilis
sigma B can occur in the absence of the
sigma B negative regulator
RsbX. J Bacterial
1997,
stress signals
switching
in sporulating
Duncan
Garsin
M, Dufour
WG:
operon
41.
A genetic
screen
was conducted
to identify
AB mutants
that were able to
inhibit &activity,
but unable to release
oF properly
during sporulation.
An AB
mutant
(El04K)
is described
that is defective
in the AA phosphorylation.
Unlike
the co~esponding
AA mutant
that cannot
be phospho~lated
(and
therefore
ieads to high oF activity),
this mutant
leads to low oF activity.
The
results
suggest
that phosphorylation
of AA by the AB*ATP*oF
complex
is
necessary
for CIF release
(induced
release,
see [321).
33.
B, Hecker
40.
R: The kinase
Garsin
A, Maul
mechanisms
activate sigma B of Baci//us
to environmental
and metabolic
stresses.
with
.
MD:
31.
284:557-568.
In work done in parallel
AA binding.
Remarkably,
1996,10:2265-2275.
U, Voelker
chara~erization
of l38ciifus
subtilis
supprass the loss of the negative
1998,
180:3671-3680.
Properties
of the phospho~iation
catalyzed
by SpollAB that help to regulate sporuiation
subtilis.
J Bacterial
1997, 179:5628-5631,
SM, Harris
Dev
The authors
present
genetic
data consistent
involved
in down-regulation
of the stress-induction
& regulatory
system.
136.
Decatur
subfilis
Magnin
39.
.
Bacillus
In an extension
of [29”1,
the authors
demonstrate
that AA increases
the
binding
of ATP to the AB kinase and that the rate-limiting
step in turnover
of
the AB kinsse is release
of ADP. A tight~binding
conformation
of AB, designated AB’, is proposed.
32.
Voelker
1995,17?:3771-3780.
BaciiJus
subfitis.
J Bacterial
1997, 179~3922-3927.
Using surface
plasmon
resonance,
the authors
measured
the binding
affinities of the regulators
affecting
c+ activity.
The results
establish
a very tight
binding
interaction
for the AA*AB*ADP
complex
and kinetic
analyses
iilustrate that this complex
is essentially
trapped
on the time scale of bacterial
development.
30.
38.
Separate
response
U: Non-specific,
general and multiple stress
of gro~h-restri~ed
Baciilus subtiiis
cells by the
of the sigmaB regulon. MO/ Microbioi
1998,
29:1129-l
CM, Brody MS, Price CW: Opposing
pairs of serine
protein kinases end phosphatases
transmit signals of
envimnm~tal
stress to activate a bacterial transcription
factor.
Genes
M, V(ilker
resistance
expression
1996,178:3846-3853.
Yang X, Kang
179:1980-l
P, Losick
R: Molecular
genetics
Annu Rev Genet
1996,30:297-241,
of mutations
CM, Brody MS, Akbar S, Yang X, Price CW: Homologous
pairs of regulatory
proteins control activity of &c&s
subfilis
transcription
factor sigma(B)
in response to envimnm~~l
stress.
37.
KT: The ffk
I? A molecular switch controlling
competence
and
motility: competence
regulatory
factors Corns, Me&, and ComK
control sigmaD-dependent
gene expression
in Badus
subtilis.
J Bacterial
1998, 180:4243-4251.
Haldenwang
WG:
Rev 1995,59:1-30.
29.
..
ME, Bailey
effects
Kang
J Bacfeiioi
KT: Flk couples @y&f
in Salmonella
typhimurium.
Liu J, Zuber
25.
28.
36.
couples flagellar P- and L-ring
morphogenesis.
J Bacterial
1997,
assembly to flegellar
1792389-2400.
24.
surthe
of
ME, Hughes
to flagellar ring assembly
1998, 180:5384-5397.
by the allele-specific
Provides
a first glimpse
of the three-dimensional
structure
of a component
of
the pawner-switchjng
regulation
system. This structure will be useful
in
understanding
the effects
of phospho~ation
on the AA*AB
interaction
and
in modeling
similar effects
in homologous
systems.
Dev 1998,
A combination
of promoter
filter-binding,
native gel eledrophoresis,
and
face plasmon
resonance
support
a model
in which
FlgM can increase
rate of dissociation
of the 02s holoenzyme
in addition
to sequestration
free &s.
as judged
factors:
mechanism
1997,167:325-331.
citrate transport
in f. co/i.
citrate transport
of Escherichia
co/k
of the FecR transmembrane
regulatory
protein.
V: Ferric
regions
1998,180:2387-2394.
A, Plantor
S, Angerer
A, Braun V: Transcription
induction of the ferric citrate transport
genes via the N-terminus
of the FecA outer membrane
protein, the Ton system and the
electrochemical
potential of the cytoplasmic
membrane.
Mof
Kim I, Stiefel
Microbial
1997, 23:333-344.
Identifies
a unique
amino-terminal
protein
that is essential
for signal
rate transport.
extension
transduction
of the FecA outer membrane
but dispensable
for ferric cit-
Anti-sigma
48.
Brutsche
Gen
49.
Genet
Huang
V: SigX of Bacillus subtilis replaces the ECF
Fact of Escherichia
co/i and is inhibited
by RsiX MO/
S, Braun
sigma factor
1997,
256:416-425.
X, Decatur
A, Sorokin
A, Helmann
JD: The Bacillus
protein is an extracytoplasmic
function sigma factor
to the survival of high temperature
stress. / Bacterial
179:2915-2921.
50.
Huana
B&L&IS
X. Helmann
subti/is
J MO/ &o/1998,
51.
.
De Las Penas
JD: Identification
ox factor
using
279:165-l
73.
A, Connolly
L, Gross
sobtilis
contributing
of tamet oromoters
a consen&directed
1997,
for the
search.
wrth 152.1 esfablrshes
that RseA is a membrane-bound
& activity
in E. co/i. RseB, which has only a modest
is localized
to the periplasm
and interacts
with RseA.
anti-sigma
effect on oE
Missiakas
D, Mayer
MP, Lemaire
M, Georgopoulos
Helmann
C, Raina
141
S:
Modulation
of the Escherichia
co/i sigma-E (RPoE) heat-shock
transcription-factor
activity by the RseA, RseB and RseC proteins.
MO/ Microbial
1997,24:355-371.
Together
with [51 l ] establishes
that RseA is an anti-sigma
affecting
ity in E. co/i. The binding
interactions
between
Rse proteins
were
both biochemically
and using the yeast two-hybrid
system.
oE activevaluated
K, McPherson
CJ, Ohman
DE: Posttranslational
control
the a/gr (aQU)-encoded
@* for expression
of the alginate
regulon in Pseudomonas
aeruginosa
and localization
of its
antagonist
proteins MucA and MucB (AION). J Bacterial
1997,
179:371 l-3720.
53.
Mathee
54.
.
Paget MS, Kang JG, Roe JH, Buttner
MJ: sigmaR, an RNA
polymerase
sigma factor that modulates
expression
of the
thioredoxin
system in response to oxidative
stress in
Streptomyces
weliwlor
A3(2). EMBO
J 1998,
17:5776-5702.
CA: The sigma-E-mediated
response to extracytoplasmic
stress in Escherichia
co/i is
transduced
by RseA and RseB, two negative regulators
of
sigma-E. MO/ Microbial
1997, 24:373-385.
-
Together
affecting
activity,
19
52.
.
factors
An ECF o factor responding
edly regulated
by an anti-a
to oxidative
stress is described
which
factor regulated
by thiol oxidation/reduction.
of
is report-