Protein-protein communication within the transcription apparatus.

JOURNAL

OF

BACTERIOLOGY,

May

1993,

p.

2483-2489

Vol.

175,

No.

9

0021-9193/93/092483-07$02.00/0

Copyright

©

1993,

American

Society

for

Microbiology

MINIREVIEW

Protein-Protein

Communication

within

the

Transcription

Apparatus

AKIRA

ISHIHAMA

Department

of

Molecular

Genetics,

National

Institute

of

Genetics,

Mishima,

Shizuoka

411,

Japan

INTRODUCTION

Despite

the

rapid

expansion

of

catalogues

of

cis-acting

transcription

signals

on

DNA

and

of

trans-acting

transcrip-

tion

regulatory

proteins

found

in

both

prokaryotes

and

eukaryotes,

relatively

little

is

known

about

precisely

how

these

transcription

signals

and

factors

ultimately

influence

transcription.

Recent

progress

in

studies

of

the

molecular

anatomy

of

Escherichia

coli

RNA

polymerase,

however,

has

led

to

a

breakthrough

in

our

understanding

of

the

molecular

mechanisms

underlying

transcription

regulation

by

tran-

scription

factors.

The

RNA

polymerase

holoenzyme

of

E.

coli

is

composed

of

core

enzyme

with

the

subunit

structure

a2133'

and

one

of

the

several

species

of

v

subunit

which

are

involved

in

the

specific

recognition

of

promoters.

The

core

enzyme

is

the

basic

machinery

of

RNA

synthesis:

the

catalytic

site

of

RNA

polytherization

is

located

on

the

1

subunit,

while

RNA

polym

rase

binds

to

DNA

nonspecifically

via

the

13'

subunit.

Subunit

a

links

these

two

large

subunits

into

the

core

enzyme

complex

(for

reviews,

see

references

15,

17,

and

43).

The

core

enzyme

is

functionally

differentiated

into

the

various

forms

of

holoenzyme

by

binding

one

of

the

different

molecular

species

of

cr

subunit

(for

a

review,

see

reference

10).

Simple

promoters

are

recognized

by

one

or

more

of

these

holoenzymes,

but

some

promoters

require

additional

transcription

factors

for

transcription

initiation

(1,

16,

17).

An

increasing

amount

of

evidence

indicates

that

interplay

between

RNA

polymerase

and

transcription

factors

involves

direct

protein-protein

contacts

(19).

Thus,

the

molecular

architecture

of

the

transcription

apparatus

for

specific

and

accurate

initiation

differs

in

detail

from

promoter

to

pro-

moter.

Each

transcription

apparatus

is

responsible

for

tran-

scription

of

only

a

set

of

genes.

Since

the

number

of

core

enzyme

molecules

is

fixed

at

a

constant

level

characteristic

of

the

rate

of

cell

growth,

i.e.,

on

average

about

2,000

molecules

per

genome

equivalent

of

DNA,

the

degree

to

which

each

of

the

approximately

4,000

genes

on

the

E.

coli

chromosome

is

transcribed

is

primarily

determined

by

the

relative

numbers

of

each

kind

of

transcription

apparatus

with

a

particular

promoter

selectivity

(16,

18).

In

this

article,

I

will

summarize

the

protein-protein

communication

between

RNA

polymerase

and

transcription

factors

from

E.

coli.

REPLACEMENT

OF

SIGMA

SUBUNIT:

FIRST-STEP

DIFFERENTIATION

OF

RNA

POLYMERASE

Functional

differentiation

of

RNA

polymerase

by

replace-

ment

of

the

or

subunit

was

first

found

in

Bacillus

subtilis,

and

the

sequential

pathway

of

v

replacement

during

sporulation

has

been

established

(for

a

recent

review,

see

reference

8).

On

the

other

hand,

six

different

species

of

cr

subunit

in

E.

coli

have

been

identified

(10,

26,

40)

(Fig.

1).

Most

genes

ex-

pressed

in

exponentially

growing

cells

are

transcribed

by

the

holoenzyme

Eu70

(E

represents

core

enzyme

and

c70

repre-

sents

the

cr

subunit

with

a

molecular

weight

of

70,000

or,

alternatively,

qD,

the

rpoD

gene

product).

Under

stress

conditions,

specific

sets

of

genes

are

transiently

induced

for

adaptation

and

survival.

Minor

or

subunits

were

originally

identified

as

the

regulatory

components

for

expression

of

stress-induced

genes.

For

example,

the

genes

for

heat

shock

proteins

are

transcribed

(or

transcribed

more

vigorously)

by

the

holoenzyme

Eu32

(or

u11,

the

rpoH

gene

product)

and

Ea4

(a'E)

(10).

The

rpoH

gene

is

expressed

upon

exposure

of

steady-state

cells

to

heat

shock,

and

uE

is

synthesized

at

extreme

high

temperatures

(the

rpoE

gene

has

not

yet

been

isolated).

The

rpoN

gene

encodes

u54

(uN),

which

is

required

for

transcription

of

the

genes

which

are

controlled

by

the

availability

of

nitrogen

sources

(10).

This

ur

subunit

does

not

form

a

stable

holoenzyme

complex,

but

it

can

bind

to

some

cognate

promoters

in

the

absence

of

core

enzyme.

Thus,

cr54

is

functionally

similar

to

eukaryotic

TATA-binding

protein

(TBP;

one

of

the

basic

transcription

factors

with

TATA

box-binding

activity)

(4).

Structurally,

arI4

is

different

from

other

v

subunits,

lacking

most

of

the

conserved

sigma

domains.

The

rpoF

gene

codes

for

another

v

subunit

(ou)

which

is

needed

for

transcription

of

a

set

of

genes

coding

for

the

formation

of

flagella.

During

the

transition

from

exponential

growth

to

station-

ary

phase,

a

set

of

genes

is

induced,

at

least

some

of

which

are

essential

for

starvation

survival

in

the

stationary

phase

(36).

The

rpoS

gene

was

first

identified

as

katF,

a

regulatory

gene

for

katE

which

encodes

HPII

catalase

(24).

KatF

was

later

found

to

be

a

positive

regulator

for

a

number

of

other

stationary-phase-specific

genes.

Sequence

analysis

indicated

that

the

rpoS

gene

product

is

a

member

of

the

r70

family

(29).

Recently,

Tanaka

et

al.

(40)

detected

the

sigma-like

activity

of

purified

ao8

(crs,

the

rpoS

gene

product)

in

vitro.

The

promoters

of

genes

that

are

expressed

only

in

stationary

growth

phase

can

be

recognized

only

by

Ea38.

Thus,

re-

placement

of

u70

(oD)

with

a-8

(cs)

is

essential

for

the

expression

of

at

least

some

stationary-phase-specific

genes.

However,

the

promoters

of

some

genes

which

are

expressed

in

both

exponentially

growing

and

stationary-phase

cells

were

found

to

be

recognized

by

both

Eu70

and

Ecr8.

This

result

indicates

a

new

type

of

transcription

regulation,

in

which

the

transcription

level

from

a

single

promoter

is

controlled

by

the

molecular

species

of

the

a

subunit.

On

the

basis

of

these

observations,

we

proposed

that

a38

is

a

second

principal

cr

factor,

selectively

used

in

stationary-phase

cells.

The

intracellular

level

of

o38

increases

concomitantly

with

the

cessation

of

cell

growth.

Moreover,

some

E.

coli

strains

lacking

the

functional

rpoS

gene

cannot

survive

in

stationary

phase

(40).

2483

2484

MINIREVIEW

Core

enzyme

a70

<rpoD>

a32

<rpoH>

a

24

<rpoE>

Eur7O

-

Regular

genes

Eo,32

-

--

Heat-shock

genes

Eu24

--,

Extreme

heat-shock

genes

a54

<rpoN>

w

EaM

-4v

Nitrogen-regulated

genes

a28

<poF>

-

Eol8

---

Flagella-Chemotaxis

genes

a38

<rpoS>~,

EU38

-o-

Oxidative

stress

response

genes

Stationary

phase-specific

genes

Sigma

subunit

Holoenzyme

Transcribed

genes

FIG.

1.

Functional

differentiation

of

RNA

polymerase

core

enzyme.

The

core

enzyme

with

the

subunit

composition

a2%3'

carries

the

basic

function

of

RNA

polymerization

but

is

inactive

in

transcription

initiation

from

promoters.

By

binding

one

of

the

multiple

species

of

a

subunit,

core

enzyme

is

converted

into

holoenzyme.

Each

holoenzyme

is

responsible

for

transcription

of

a

set

of

genes.

INTERPLAY

WITH

TRANSCRIPTION

FACTORS:

SECOND-STEP

DIFFERENTIATION

OF

RNA

POLYMERASE

The

holoenzyme

at

least

in

the

case

of

Eu70

and

Eu54

is

further

differentiated

into

various

forms

of

transcription

apparatus

through

interaction

with

transcription

factors

(16,

19).

Some

of

these

regulatory

factors

bind

to

RNA

poly-

merase

in

the

absence

of

DNA,

but

most

bind

to

the

enzyme

only

when

both

components

are

aligned

on

DNA.

Interac-

tion

between

RNA

polymerase

and

a

regulatory

protein

that

binds

to

neighboring

DNA

sites

at

the

promoter

results

in

cooperative

binding

of

both

components.

Binding

of

a

pro-

tein

molecule

to

a

stronger

affinity

DNA

site

helps

the

binding

of

a

second

molecule

to

a

weak

affinity

site,

thereby

increasing

the

effective

concentration

of

the

second

mole-

cule

for

binding

to

the

weaker

site.

Protein-protein

contacts

help

to

position

the

RNA

polymerase

in

register

with

the

promoter

or

increase

the

rate

of

isomerization

of

the

RNA

polymerase-promoter

complex

from

the

closed

to

open

state.

The

specificities

of

protein-protein

contacts

between

RNA

polymerase

and

transcription

factors

are

determined

through

interaction

of

short

peptide

segments

on

the

exposed

surface

of

each

component

(19).

The

molecular

assemblies

thus

formed

might

be

thermodynamically

unstable

but

could

be

stabilized

either

by

DNA

binding

of

both

components

or

by

secondary,

less-specific

protein-protein

contacts.

Several

lines

of

evidence

indicate

that

direct

protein-protein

contacts

are

involved

in

the

specificity

determination

of

transcription

activation

of

catabolite-sensitive

genes

by

cyclic

AMP

(cAMP)

receptor

protein

(CRP)

(or

catabolite

gene

activa-

tion

protein)

(for

a

recent

review,

see

reference

32).

These

lines

of

evidence

include

cooperative

binding

of

the

two

components,

contact-induced

fluorescence

polarization,

and

isolation

of

activation-negative

CRP

mutants

which

can

bind

to

target

DNA

sites

but

cannot

activate

transcription.

In

addition,

we

isolated

a

number

of

RNA

polymerase

mutants

with

defective

responses

to

activation

by

cAMP-CRP

(14,

44).

One

productive

approach

for

identification

of

the

contact

sites

on

RNA

polymerase

for

various

transcription

factors

is

to

perform

a

systematic

search

for

activation-negative

RNA

polymerase

mutants

and

locate

the

mutations

on

the

RNA

polymerase

genes.

Up

to

the

present

time,

two

contact

site

regions

have

been

identified

on

E.

coli

RNA

polymerase

(Fig.

2),

one

in

the

C-terminal

region

of

the

a

subunit

(contact

site

I)

and

another

in

the

C-terminal

region

of

the

c70

subunit

(contact

site

II).

PROTEIN-PROTEIN

CONTACTS:

MAPPING

OF

CONTACT

SITE

I

ON

RNA

POLYMERASE

The

N-terminal

region

of

the

RNA

polymerase

at

subunit

is

involved

in

subunit-subunit

contacts

for

core

enzyme

assem-

bly

(12),

while

mutations

in

its

C-terminal

region

affect

the

response

of

the

enzyme

to

various

transcription

factors.

Thus,

we

proposed

that

one

of

the

contact

sites

(site

I)

is

located

in

this

region

(12).

In

our

collaboration

with

a

number

of

laboratories

working

with

specific

transcription

factors

from

E.

coli,

the

factors

are

being

classified

into

those

that

can

or

cannot

activate

mutant

RNA

polymerases

containing

C-terminally

truncated

a

subunits.

Figure

3

shows

a

tentative

list

of

transcription

factors

that

require

site

I

for

transcription

activation.

The

locations

of

activator-

binding

sites

on

DNA

vary

from

protein

to

protein

and

from

promoter

to

promoter

(5).

Most

activator-binding

sites

are

located

close

to

the

promoter

and

upstream

of

the

promoter.

However,

the

binding

site

can

be

moved

kilobases

away

by

genetic

manipulation,

at

least

in

the

case

of

NtrC

(NR1)

without

loss

of

activation

proficiency

(31).

Below

are

the

class

I

transcription

factors

(Fig.

3),

which

appear

to

make

contact

with

contact

site

I

on

the

a

subunit.

These

activator

proteins

bind

target

DNA

sites

located

upstream

of

the

relevant

promoter.

Ada.

Exposure

of

cells

to

methylating

agents

triggers

an

adaptive

response

by

generating

methyl-phosphotriester

ste-

reoisomers

on

DNA.

This

signal

enhances

transcription

of

the

genes

encoding

several

DNA

repair

enzymes,

including

the

regulatory

protein

Ada

itself.

The

Ada

protein

directly

reverses

the

major

mutagenic

lesion

in

DNA,

06-methylgua-

nine,

by

transferring

the

methyl

group

from

guanine

to

a

cysteine

residue

near

the

N-terminal

end

of

Ada.

The

methylated

Ada

then

functions

as

an

activator

for

transcrip-

tion

of

the

genes

of

the

Ada

regulon

including

ada.

Sakumi

et

al.

(34)

found

that

Ada

requires

the

C-terminal

contact

site

I

of

RNA

polymerase

a

for

transcription

activation

of

ada.

Ada

binds

just

upstream

of

the

ada

promoter

(the

binding

center

is

-46).

J.

BACTERIOL.

MINIREVIEW

2485

a

subunit

(329

aa)

a

subunit

(613

aa)

Contact

site

I

lp

RNA

polymnerase

holoenzyme

(Eu7O)

Contact

site

I

~~~~~~~~~~~~~~~~~~~~~Iq

FIG.

2.

Transcription

factor

contact

sites

on

RNA

polymerase.

The

C-terminal

contact

site

I

on

a

is

necessary

for

transcription

activation

by

class

I

transcription

factors

and

may

be

involved

in

direct

protein-protein

contacts.

Two

class

II

factors,

PhoB

and

CRP

(in

the

case

of

gal

transcription),

make

contacts

with

RNA

polymerase

in

the

C-terminal

region

(contact

site

II)

of

the

a70

subunit.

aa,

amino

acids.

CRP.

The

location

of

the

CRP-binding

site

varies

between

the

many

catabolite-sensitive

operons

(about

10%

of

all

E.

coli

genes

are

considered

to

be

under

the

control

of

the

cAMP-CRP

complex).

Changing

the

location

of

CRP

binding

[J

Class-I

Transcription

Factors

-100

-50

Promoter

,.35

_10

Trl

Trpl

MW

-

trpAB

Ada

_

_

~~~~ada

OxyR

OxYR

katG

OxyR

OxyR

~~

~~~

ahpc

Ox~~~

R

_

oxyX

OmpR

OmrpR

OmpR

ompC

CRP

uxuAB

CRP

affects

its

activation

proficiency.

The

angular

orientation,

rather

than

the

absolute

distance

of

its

binding

site

with

respect

to

the

promoter,

is

important

(39).

Transcription

in

vitro

from

the

CRP-dependent

lacP1

promoter

(the

major

E

Class-Il

Transcription

Factors

-100

y

-50

-35

-10

PhoB

PhoB

_

_l~s

pst

CRP

ga

j*j

Non-classLo

Factors

(Unclassified)

[i

Non-class

I

Factors

(Unclassitied)

-100

V

Promoter

-50

-35

-10

MerR

_--

Xcii

i

merT

XPRE

CRP

Xcl

Xcl

_n

m

lac

XPRM

NtrC

_XPL

ESN::::::'::-::'

ginA

FIG.

3.

Location

of

transcription

factor-binding

sites.

(A)

Class

I

factors.

Many

of

the

class

I

transcription

factors

which

require

the

C-terminal

contact

site

I

on

RNA

polymerase

a

subunit

bind

to

target

DNA

signals

located

upstream

of

the

promoter.

(B)

Class

II

factors.

Both

PhoB

and

CRP

(in

the

case

of

gal

activation)

belonging

to

class

II

factors

bind

the

DNA

overlapping

promoter

-35.

(C)

Other

non-class

I

factors.

The

protein-protein

contact

sites

of

other

non-class

I

factors

have

not

been

identified

yet.

I

r..

am

i

I

I

I

m

1..

0

uilgffdm..

I-

--

VOL.

175,

1993

t

gal

IHF

mmlml

_S

2486

MINIREVIEW

Assembly

defect

TS

I&-A

Y

1

100

200---

AraC

Ogr(P2)

CysB

6(P4)

OxyR

MOR

300

329

Fnr

*

OmpR,

0

A

'%

L

lI

*

*

"

CRP

U.

* -

250

260

270

280

290

300

310

320

329

No.

of

Amino

Acids

from

N-terminus

FIG.

4.

Mapping

of

the

transcription

factor

contact

site

I.

RNA

polymerase

mutants

with

defective

responses

to

transcription

activation

by

various

class

I

factors

carry

mutation(s)

in

the

C-terminal

region

of

the

a

subunit

as

indicated:

CRP

(5a,

44);

OxyR

(41);

AraC,

CysB,

and

MeIR

(42);

Ogr(P2)

or

b(P4)

(37);

Fnr

(25);

OmpR

(37).

downstream

promoter)

was

observed

with

wild-type

RNA

polymerase,

but

not

with

mutant

enzymes

containing

C-ter-

minally

truncated

a

subunits

(14).

Kolb

et

al.

(20)

clearly

demonstrated

cooperative

binding

of

CRP

and

wild-type

RNA

polymerase

to

lacP1,

but

no

such

binding

with

the

mutant

RNA

polymerases.

This

result

indicates

that

at

least

for

CRP,

contact

site

I

is

indeed

involved

in

direct

protein-protein

contact.

Furthermore,

they

showed

that

DNase-sensitive

sites

appeared

between

the

CRP-

and

RNA

polymerase-binding

sites

only

when

the

mutant

RNA

poly-

merases

were

used

(the

center

of

CRP

binding

on

lacP1

is

-61.5).

This

result

suggests

that

at

least

one

a

subunit

of

RNA

polymerase

in

the

open

complex

is

located

at

the

upstream

end

(RNA

polymerase

covers

about

60

bp

of

the

promoter

DNA

between

-45

and

+15).

The

absence

of

contacts

between

the

truncated

RNA

polymerases

and

CRP

could

explain

the

deficiency

in

transcription

activa-

tion.

IHF.

Integration

host

factor

(IHF)

is

one

of

the

E.

coli

DNA-binding

proteins

with

DNA-bending

activity

and

is

involved

in

a

variety

of

cellular

processes,

such

as

transpo-

sition,

site-specific

recombination,

and

control

of

gene

tran-

scription.

IHF

stimulates

transcription

from

the

phage

APL

promoter

(IHF

binds

two

tandem

sites

centered

at

-86

and

-180).

The

mutant

RNA

polymerase

containing

truncated

a

showed

no

IHF

stimulation

ofPL.

This

result

defines

IHF

as

a

class

I

transcription

activator

(7).

OmpR.

OmpR

is

an

activator

protein

involved

in

osmo-

regulation

of

expression

of

the

genes

encoding

the

alterna-

tive

porin

proteins,

OmpC

and

OmpF.

When

OmpR

is

phosphorylated

by

EnvZ,

the

activator

binds

to

three

target

sites

clustered

upstream

of

the

ompC

promoter

(the

binding

center

of

the

closest

site

is

-46)

and

activates

transcription.

OmpR

is,

however,

unable

to

activate

mutant

RNA

poly-

merases

containing

truncated

at

subunits

(13),

indicating

that

OmpR

makes

contact

with

the

deleted

region

of

a

OxyR.

Upon

exposure

to

a

sublethal

dose

of

H202,

a

number

of

genes

for

detoxification

of

oxidants

are

induced

by

an

activator,

OxyR

(38).

The

OxyR

protein

is

a

bifunc-

tional

protein,

acting

as

both

a

sensor

for

oxidative

stress

and

transcription

activator

for

the

stress-induced

genes,

including

katG

coding

for

catalase

I

and

ahp

coding

for

alkylhydroperoxide

reductase.

Tao

et

al.

(41)

demonstrated

that

purified

OxyR

activates

in

vitro

transcription

from

the

katG

and

ahp

promoters

by

wild-type

RNA

polymerase

but

not

by

mutant

polymerases

containing

C-terminally

trun-

cated

a.

Moreover,

they

showed

that

the

OxyR

protein

causes

cooperative

binding

(an

indication

of

protein-protein

contact)

of

wild-type

RNA

polymerase

but

not

mutant

enzymes.

Detailed

mapping

of

the

contact

site

for

CRP

in

the

case

of

lac

transcription

activation

was

done

with

a

library

of

point

mutations

within

the

contact

site

I

region

of

the

rpoA

gene

encoding

the

RNA

polymerase

a

subunit

(44).

To

my

sur-

prise,

all

the

activation-negative

rpoA

mutations

mapped

within

a

narrow

region

encoding

amino

acid

residues

265

to

270

(Fig.

4).

Using

a

different

screening,

Ebright

(5a)

isolated

similar

mutants

carrying

rpoA

mutations

in

the

same

region.

The

R265C

mutant

we

isolated

was

found

also

to

be

defec-

tive

in

response

to

activation

by

OxyR

(41).

The

rpoA341

mutation

leading

to

defects

in

response

to

activation

by

CysB,

MelR,

and

AraC

was

mapped

at

codon

271

(42).

Thus,

a

number

of

activator

proteins

seem

to

interact

with

this

region

of

about

10

amino

acids

in

length

between

amino

acids

260

and

270.

The

contact

domain

on

CRP

for

RNA

poly-

merase

has

also

been

mapped

within

a

narrow

region,

again

consisting

of

about

5

to

10

amino

acid

residues

(2, 6).

Taken

together

these

results

suggest

that

the

RNA

polymerase-

CRP

contact

is

similar

to

the

epitope-paratope

interaction

between

antigens

and

antibodies.

Recently,

Slauch

et

al.

(37)

isolated

rpoA

mutations

con-

ferring

defective

responses

to

transcription

activation

by

OmpR,

some

of

which

affect

two

adjacent

amino

acid

residues,

322

and

323.

Likewise,

Lombardo

et

al.

(25)

mapped

rpoA

mutations

causing

defective

responses

to

activation

by

Fnr,

a

regulatory

protein

for

genes

expressed

under

anaerobic

conditions,

mainly

within

a

region

including

amino

acids

311

and

317.

Thus,

there

seems

to

exist

another

contact

site

cluster

near

the

C

terminus

of

the

RNA

poly-

merase

a

subunit

(for

recent

reviews,

see

references

19

and

33).

RNA

Polymerase

at

Subunit

Ts

J.

BAcTWERIOL.

MINIREVIEW

2487

PROTEIN-PROTEIN

CONTACTS:

LOCATION

OF

CONTACT

SITE

II

ON

RNA

POLYMERASE

Some

activator

proteins,

most

of

which

bind

overlapping

the

promoter,

do

not

require

contact

site

I

on

the

C-terminal

region

of

a

for

transcription

activation

(13),

indicating

that

these

proteins

make

contact

with

an

as

yet

unidentified

site

on

RNA

polymerase.

Recently,

Makino

et

al.

(27)

showed

that

the

contact

site

for

PhoB

is

located

in

region

4

of

the

a70

subunit

at

the

upstream

end

of

subregion

4.2.

Kumar

et

al.

(21)

suggested

that

CRP

also

makes

contact

with

u70

in

the

case

of

gal

activation.

Furthermore,

Lee

et

al.

(23)

proposed

the

contact

site

of

NtrC

on

Or4.

Thus,

we

propose

to

designate

the

contact

site

on

the

a

subunit,

including

not

only

u70

but

also

other

minor

or

factors

as

contact

site

II,

and

to

classify

activators

that

require

contact

site

II

into

class

II

factors.

CRP.

cAMP-CRP

is

unable

to

activate

lacP1

transcription

by

mutant

RNA

polymerases

containing

C-terminally

trun-

cated

a

(13).

However,

transcription

of

gal

by

the

same

mutant

RNA

polymerases

was

activated

by

CRP

in

the

presence

of

cAMP

(the

CRP

binding

center

is

-41.5),

indicating

that

protein-protein

contacts

can

still

occur

with

the

truncated

polymerases.

Thus,

contact

site

II

for

gal

transcription

was

considered

to

be

located

on

another

region

of

a

or

on

another

subunit

of

RNA

polymerase.

Kolb

et

al.

(20)

showed

synergistic

binding

of

CRP

and

the

mutant

RNA

polymerases

at

galPl,

but

not

at

lacP1.

The

correlation

between

transcription

activation

and

synergistic

binding

also

supports

a

direct

protein-protein

contact

between

CRP

and

RNA

polymerase.

During

studies

of

the

molecular

anatomy

of

the

&70

sub-

unit,

Kumar

et

al.

(21,

22)

discovered

that

deletion

of

the

C-terminal

region

(the

promoter

-35

recognition

domain)

of

this

subunit

does

not

prevent

transcription

initiation

at

"extended-10"

promoters,

which

lack

the

-35

consensus

sequence

but

have

a

strong

-10

signal

with

TGNTATAAT

sequence.

However,

C-terminal

deletion

of

a70

impairs

the

ability

of

the

holoenzyme

to

respond

to

CRP

activation

of

gal

transcription.

These

observations

indicate

that

activation

site

II

(or

contact

site

II)

for

CRP

is

located

within

the

deleted

region,

close

to

the

PhoB

contact

site

(below).

PhoB.

PhoB

is

the

activator

protein

for

a

number

of

genes

organized

in

the

phosphate

regulon,

which

are

included

upon

depletion

of

external

phosphate.

All

thepho

regulon

operons

carry

the

Pho

box,

the

binding

site

of

PhoB

protein,

over-

lapping

the

promoter

-35

region.

PhoB

is

activated

upon

phosphorylation

by

the

phosphotransferase

PhoR,

and

acti-

vated

PhoB

enhances

transcription

in

vitro

of

pst,

the

gene

for

phosphate

transport,

catalyzed

by

both

wild-type

and

mutant

RNA

polymerases

(13).

This

result

indicates

that

the

C-terminal

a-subunit

contact

I

is

not

needed

for

transcription

activation

by

PhoB.

Makino

et

al.

(27)

demonstrated

that

mutations

near

the

upstream

end

of

the

conserved

region

4.2

(the

recognition

domain

for

the

promoter

-35

signal)

of

Or70

render

RNA

polymerase

inactive

in

response

to

PhoB.

This

result

strongly

indicates

that

the

contact

site

for

PhoB

is

located

within

this

region.

Results

obtained

using

C-termi-

nally

truncated

a70

subunits

have

confirmed

this

conclusion

(21).

As

noted

above,

many

of

the

class

I

activators

bind

upstream

of

the

promoter

-35

signal.

Two

class

II

factors,

PhoB

and

CRP

(in

the

case

of

gal

activation),

bind

overlap-

ping

the

-35

region

or

within

the

promoter

sequence

(Fig.

3).

The

contact

site

selection

by

an

activator

protein,

how-

ever,

depends

not

only

on

the

location

of

its

DNA

binding

site

but

also

on

the

global

structure

of

the

protein.

For

instance,

among

activators

that

produce

footprints

overlap-

ping

the

promoter

-35

region,

some

make

contact

with

site

I

(Ada,

TrpI,

and

OxyR)

but

others

do

not

(XCI,

XCII,

PhoB,

and

CRP

in

the

case

of

gal

activation).

Furthermore,

DNA

looping

mediated

by

NtrC

allows

direct

protein-protein

contact

between

this

activator

and

RNA

polymerase.

OTHER

NON-CLASS

I

FACTORS

AND

DNA

CURVATURE

Besides

the

class

II

factors

noted

above,

at

least

four

factors,

XCI,

XCII,

MerR,

and

NtrC,

have

been

found

not

to

require

the

C-terminal

region

of

a

subunit

for

action,

but

their

contact

sites

have

not

been

identified

yet.

XCI

and

XCII.

Transcription

of

the

phage

X

promoter,

PRM'

for

expression

of

the

repressor

cI

gene

is

activated

by

CI

protein

bound

to

the

operator

OR2

(the

binding

site

is

centered

at

-42)

(11).

Some

cI

mutations

cause

a

defect

in

activation

of

pRm

but

do

not

affect

DNA

binding,

indicating

that

the

mutations

affect

a

surface

of

the

CI

protein

that

interacts

with

RNA

polymerase.

Gussin

et

al.

(9)

demon-

strated

that

transcription

in

vitro

from

the

APRM

promoter

was

normal

even

with

mutant

RNA

polymerases

lacking

the

C-terminal

contact

site

I

region

of

ao,

indicating

that

CI

makes

contact

with

putative

site

II

or

other

portions

of

RNA

polymerase.

Likewise,

transcription

initiation

from

XpRE

is

activated

by

binding

of

the

CII

protein

to

two

TTGC

sequences

that

flank

the

promoter

-35

region

(the

binding

center

is

-32.5

bp).

The

mutant

RNA

polymerases

were

activated

by

CII,

although

to

a

lesser

extent

than

the

wild-type

polymerase

(9).

Again,

the

CII

protein

appears

to

make

contact

with

a

site

other

than

contact

site

I

on

ax.

MerR.

Increases

in

the

extracellular

concentration

of

heavy

metals

such

as

mercuric

ion

induce

toxic

stress.

The

signal

is

received

by

the

regulatory

activator

protein

MerR,

which

ultimately

activates

transcription

of

the

mer

operon

for

mercury

resistance.

The

product

of

mer

lowers

the

effective

concentration

of

mercury

by

chelating

it.

The

Hg2+-MerR

complex

binds

between

the

-35

and

-10

signals

of

the

mer

promoter,

thus

overlapping

the

RNA

polymerase-

binding

site

(30).

The

activator

protein

and

RNA

polymerase

occupy

two

different

sides

of

the

same

region

of

DNA.

Deletion

of

the

contact

site

I

on

a

does

not

interfere

with

transcription

activation

by

the

MerR

protein

(la).

NtrC.

When

E.

coli

is

starved

for

nitrogen,

the

synthesis

of

several

enzymes

for

the

utilization

of

alternative

nitrogen

sources

is

induced

under

the

control

of

the

NtrC

protein,

which

activates

transcription

in

conjunction

with

Er54

ho-

loenzyme.

Transcription

of

glnA

encoding

glutamine

syn-

thetase,

for

example,

requires

NtrC

bound

to

two

sites

more

than

100

bp

upstream

from

the

transcription

start

site,

but

protein-protein

contact

is

achieved

by

means

of

DNA

loop

formation

(31).

Lee

et

al.

(23)

have

found

that

the

C-terminal

region

of

the

RNA

polymerase

a

subunit

is

not

needed

for

this

NtrC-dependent

transcription

of

lnA

by

E&54

(or

Eo-N).

They

propose

that

activators

of

E&4

make

direct

physical

contact

with

contact

site

II

on

o54

subunit

rather

than

contact

site

I

on

a.

DNA-binding

transcriptional

activator

proteins

such

as

MerR

can

influence

transcription

indirectly

through

induc-

tion

of

a

promoter

conformation

competent

for

transcription

initiation.

In

fact,

intrinsic

DNA

bending

or,

alternatively,

that

induced

by

protein

binding

is

known

to

be

required

in

VOL.

175,

1993

2488

MINIREVIEW

certain

cases

of

transcription

(28,

35).

For

instance,

the

requirement

of

CRP

for

gal

transcription

can be

eliminated

by

the

insertion

of

curved

DNA

(3).

Such

artificial

elements

promote

transcription,

provided

that

the

intrinsic

bend

is

phased

identically

to

the

bend

induced

by

bound

CRP.

However,

it

should

be

emphasized

that

studies

with

CRP

mutants

show

that

DNA

bending

is

not

sufficient

for

tran-

scription

activation,

indicating

that

both

DNA

curvature

and

protein-protein

contacts

are

involved.

PERSPECTIVES

The

pathway

of

signal

transduction

in

prokaryotes

in-

volves

sensors,

regulators

(transcription

factors),

and

RNA

polymerase.

An

external

or

internal

signal

(stimulus)

re-

ceived

by

a

sensor

is

transmitted

to

the

regulatory

protein

(transcription

factor).

The

activated

transcription

factor

may

bind

directly

to

RNA

polymerase

in

the

absence

of

DNA,

forming

a

functional

transcription

apparatus

with

promoter

selectivity,

but

in

most

cases

the

factor

binds

to

specific

DNA

sites

(prokaryotic

enhancers),

interacts

with

RNA

polymerase

on

DNA,

and

leads

to

transcription

activation

by

facilitating

either

RNA

polymerase

binding

to

the

promoter

or

initiation

of

RNA

synthesis.

Protein-protein

communica-

tion

between

RNA

polymerase

and

transcription

factors

thus

modulates

the

selectivity

of

promoter

recognition

by

RNA

polymerase

and

ultimately

determines

the

relative

order

of

gene

expression

in

a

given

environment.

The

identification

of

contact

sites

between

E.

coli

transcription

factors

and

RNA

polymerase

may

provide

important

insights

into

the

chal-

lenging

problem

of

elucidating

protein-protein

communica-

tions

within

the

highly

complex

transcription

apparatus

of

eukaryotic

organisms.

ACKNOWLEDGMENTS

I

thank

Richard

S.

Hayward

and

Richard

Losick

for

providing

insight

on

this

minireview

and

for

critically

reading

the

manuscript.

Research

in

Mishima

was

supported

by

grants-in-aid

from

the

Ministry

of

Education,

Science

of

Culture

of

Japan,

the

Joint

Research

Program

of

The

Graduate

University

for

Advanced

Stud-

ies,

and

the

NIG

Cooperative

Research

Programs.

REFERENCES

1.

Adhya,

S.,

and

S.

Garges.

1990.

Positive

control.

J.

Biol.

Chem.

265:10797-10800.

la.Ansari,

A.,

et

al.

Unpublished

data.

2.

Bell,

A.,

K.

Gaston,

R.

Williams,

K.

Chapman,

A.

Kolb,

H.

Buc,

S.

Minchin,

J.

Williams,

and

S.

Busby.

1990.

Mutations

that

alter

the

ability

of

the

Escherichia

coli

cyclic

AMP

receptor

protein

to

activate

transcription.

Nucleic

Acids

Res.

18:7243-7250.

3.

Bracco,

L.,

D.

Kotlarz,

A.

Kolb,

S.

Dieckman,

and

H.

Buc.

1989.

Synthetic

curved

DNA

sequences

can

act

as

transcriptional

activators

in

Eschenchia

coli.

EMBO

J.

8:4289-4296.

4.

Buck,

M.,

and

W.

Cannon.

1992.

Specific

binding

of

the

tran-

scription

factor

sigma-54

to

promoter

DNA.

Nature

(London)

358:422-424.

5.

Collado-Vides,

J.,

B.

Magasanik,

and

J.

D.

Gralla.

1991.

Control

site

location

and

transcription

regulation

in

Escherichia

coli.

Microbiol.

Rev.

55:371-394.

5a.Ebright,

R.

Personal

communication.

6.

Eschenlauer,

A.,

and

W.

Reznikoff.

1991.

Escherichia

coli

ca-

tabolite

gene

activator

protein

mutants

defective

in

positive

control

of

lac

operon

transcription.

J.

Bacteriol.

173:5024-5029.

7.

Giladi,

H.,

K.

Igarashi,

A.

Ishihama,

and

A.

B.

Oppenheim.

1992.

Stimulation

of

the

phage

X

pL

promoter

by

integration

host

factor

requires

the

carboxy

terminus

of

the

a-subunit

of

RNA

polymerase.

J.

Mol.

Biol.

227:985-990.

8.

Gross,

C.,

M.

Lonetto,

and

R.

LosickL

1992.

Bacterial

sigma

factors.

In

S.

L.

McKnight

and

K.

R.

Yamamoto

(ed.),

Tran-

scription

regulation,

Cold

Spring

Harbor

Laboratory

Press,

Cold

Spring

Harbor,

N.Y.

9.

Gussin,

G.

N.,

C.

Olson,

K.

Igarashi,

and

A.

Ishihama.

1992.

Activation

defects

caused

by

mutations

in

Eschenichia

coli

rpoA

are

promoter

specific.

J.

Bacteriol.

174:5156-5160.

10.

Helmann,

J.

D.,

and

M.

J.

Chamberlin.

1988.

Structure

and

function

of

bacterial

sigma

factors.

Annu.

Rev.

Biochem.

57:

839-872.

11.

Hochschild,

A.,

N.

Irwin,

and

M.

Ptashne.

1982.

Repressor

structure

and

the

mechanism

of

positive

control.

Cell

32:319-

325.

12.

Igarashi,

K.,

N.

Fujita,

and

A.

Ishihama.

1992.

Identification

of

a

subunit

assembly

domain

in

the

alpha

subunit

of

Escherichia

coli

RNA

polymerase.

J.

Mol.

Biol.

218:1-6.

13.

Igarashi,

K.,

A.

Hanamura,

K.

Makino,

A.

Aiba,

A.

Aiba,

T.

Mizuno,

A.

Nakata,

and

A.

Ishihama.

1991.

Functional

map

of

the

a

subunit

of

Escherichia

coli

RNA

polymerase:

two

modes

of

transcription

activation

by

positive

factors.

Proc.

Natl.

Acad.

Sci.

USA

88:8958-8962.

14.

Igarashi,

K.,

and

A.

Ishihama.

1991.

Bipartite

functional

map

of

the

E.

coli

RNA

polymerase

a

subunit:

involvement

of

the

C-terminal

region

in

transcription

activation

by

cAMP-CRP.

Cell

65:1015-1022.

15.

Ishihama,

A.

1981.

Subunit

assembly

of

Eschenichia

coli

RNA

polymerase.

Adv.

Biophys.

14:1-35.

16.

Ishihama,

A.

1988.

Promoter

selectivity

of

prokaryotic

RNA

polymerases.

Trends

Genet.

4:282-286.

17.

Ishihama,

A.

1990.

Molecular

assembly

and

functional

modula-

tion

of

Escherichia

coli

RNA

polymerase.

Adv.

Biophys.

26:

19-31.

18.

Ishihama,

A.

1991.

Global

control

of

gene

expression

in

bacte-

ria,

p.

121-140.

In

A.

Ishihama

and

H.

Yoshikawa

(ed.),

Control

of

cell

growth

and

division.

Springer-Verlag,

Berlin.

19.

Ishihama,

A.

1992.

Role

of

the

RNA

polymerase

a

subunit

in

transcription

activation.

Mol.

Microbiol.

6:3283-3288.

20.

Kolb,

A.,

K.

Igarashi,

A.

Ishihama,

M.

Lavigne,

M.

Buckle,

and

H.

Buc.

1993.

E.

coli

RNA

polymerase,

deleted

in

the

C-termi-

nal

part

of

its

a-subunit,

interacts

differently

with

the

cAMP-

CRP

complex

at

the

lacP1

and

at

the

galP1

promoter.

Nucleic

Acids

Res.

21:319-326.

21.

Kumar,

A.,

B.

Grimes,

N.

Fujita,

K.

Makino,

R.

A.

Malloch,

R.

S.

Hayward,

and

A.

Ishihama.

Role

of

the

sigma70

subunit

of

Escherichia

coli

RNA

polymerase

in

transcription

activation.

Submitted

for

publication.

22.

Kumar,

A.,

R.

A.

Malloch,

N.

Fujita,

D.

A.

Smillie,

A.

Ishihama,

and

R.

S.

Hayward.

The

minus

35-recognition

region

of

Es-

cherichia

coli

sigma

70

is

inessential

for

initiation

of

transcrip-

tion

from

an

"extended

minus

10"

promoter.

Submitted

for

publication.

23.

Lee,

H.-S.,

A.

Ishihama,

and

S.

Kustu.

1993.

The

C

terminus

of

the

a

subunit

of

RNA

polymerase

is

not

essential

for

transcrip-

tion

activation

of

a"4

holoenzyme.

J.

Bacteriol.

175:2479-2482.

24.

Loewen,

P.

C.,

and

B.

L.

Briggs.

1984.

Genetic

mapping

of

katF,

a

locus

that

with

katE

affects

the

synthesis

of

a

second

catalase

species

in

Escherichia

coli.

J.

Bacteriol.

160:668-675.

25.

Lombardo,

M.

S.,

D.

Bagga,

and

C.

G.

Miller.

1991.

Mutations

in

rpoA

affect

expression

of

anaerobically

regulated

genes

in

Salmonella

typhimurium.

J.

Bacteriol.

173:7511-7518.

26.

Lonetto,

M.,

M.

Bribskov,

and

C.

L.

Gross.

1992.

The

a70

family:

sequence

conservation

and

evolutionary

relationships.

J.

Bacteriol.

174:3843-3849.

27.

Makino,

K.,

M.

Amemura,

S.-K.

Kim,

A.

Nakata,

and

H.

Shinagawa.

1993.

Role

of

the

d70

subunit

of

RNA

polymerase

in

transcription

activation

by

activator

protein

PhoB

in

Es-

cherichia

coli.

Genes

Dev.

7:149-160.

28.

Matthews,

K.

S.

1992.

DNA

looping.

Microbiol.

Rev.

56:123-

136.

29.

Mulvey,

M.

R.,

and

P.

C.

Loewen.

1989.

Nucleotide

sequence

of

katF

of

Eschenchia

coli

suggests

KatF

protein

is

a

novel

a

transcription

factor.

Nucleic

Acids

Res.

17:9979-9991.

30.

O'Halloran,

T.

V.,

B.

Frantz,

M.

K.

Shin,

D.

M.

Ralston,

and

J.

G.

Wright.

1989.

The

MerR

heavy

metal

receptor

mediates

J.

BACTERIOL.

MINIREVIEW

2489

positive

activation

in

a

topologically

novel

transcription

com-

plex.

Cell

56:119-129.

31.

Reitzer,

L.

J.,

and

B.

Magasanik.

1986.

Transcription

of

ginA

in

Escherichia

coli

is

stimulated

by

activator

bound

to

sites

far

from

the

promoter.

Cell

45:785-792.

32.

Reznikoff,

W.

S.

1992.

Catabolite

gene

activator

protein

activa-

tion

of

lac

transcription.

J.

Bacteriol.

174:655-658.

33.

Russo,

F.

D.,

and

T.

J.

Silhavy.

1992.

Alpha:

the

Cinderella

subunit of

RNA

polymerase.

J.

Biol.

Chem.

267:14515-14518.

34.

Sakumi,

K.,

K.

Igarashi,

M.

Sekiguchi,

and

A.

Ishihama.

1993.

The

Ada

protein

is

a

class

I

transcription

factor

of

Escherichia

coli.

J.

Bacteriol.

175:2455-2457.

35.

Schleif,

R.

1988.

DNA

looping.

Science

240:127-128.

36.

Siegele,

D.

A.,

and

R.

Kolter.

1992.

Life

after

log.

J.

Bacteriol.

174:345-348.

37.

Slauch,

J.

M.,

F.

D.

Russo,

and

T.

J.

Silhavy.

1991.

Suppressor

mutations

in

rpoA

suggest

that

OmpR

controls

transcription

by

direct

interaction

with

the

a

subunit

of

RNA

polymerase.

J.

Bacteriol.

173:7501-7510.

38.

Storz,

G.,

L.

A.

Tartaglia,

and

B.

N.

Ames.

1990.

OxyR,

a

transcriptional

regulator

of

oxidative

stress-inducible

genes:

direct

activation

by

oxidation.

Science

248:189-194.

39.

Straney,

D.

C.,

S.

M.

Straney,

and

D.

M.

Crothers.

1989.

Synergy

between

Eschenchia

coli

CAP

protein

and

RNA

poly-

merase

in

the

lac

promoter

open

complex.

J.

Mol.

Biol.

206:41-

57.

40.

Tanaka,

K.,

Y.

Takayanagi,

N.

Fujita,

A.

Ishihama,

and

H.

Takahashi.

Heterogeneity

of

the

principal

sigma

factor

of

RNA

polymerase

in

stationary

phase

Escherichia

coli.

Proc.

Natl.

Acad.

Sci.

USA,

in

press.

41.

Tao,

K.,

N.

Fujita,

and

A.

Ishihama.

Involvement

of

the

RNA

polymerase

a-subunit

C-terminal

region

in

cooperative

interac-

tion

and

transcription

activation

with

OxyR.

Mol.

Microbiol.,

in

press.

42.

Thomas,

M.

S.,

and

R.

E.

Glass.

1991.

Escherichia

coli

rpoA

mutation

which

impairs

transcription

of

positively

regulated

systems.

Mol.

Microbiol.

5:2719-2725.

43.

Yura,

T.,

and

A.

Ishihama.

1979.

Genetics

of

bacterial

RNA

polymerases.

Annu.

Rev.

Genet.

13:59-97.

44.

Zou,

C.,

N.

Fujita,

K.

Igarashi,

and

A.

Ishihama.

1992.

Mapping

of

the

cAMP

receptor

protein

contact

site

on

the

a

subunit

of

Escherichia

coli

RNA

polymerase.

Mol.

Microbiol.

6:2599-

2605.

VOL.

175,

1993