emory 1992 e.coli
TRANSCRIPT
-
8/16/2019 Emory 1992 E.coli
1/15
10.1101/gad.6.1.135Access the most recent version at doi: 1992 6: 135-148Genes Dev.
S A Emory, P Bouvet and J G Belasco Escherichia coli.A 5'-terminal stem-loop structure can stabilize mRNA in
References
http://genesdev.cshlp.org/content/6/1/135#related-urlsArticle cited in:
http://genesdev.cshlp.org/content/6/1/135.refs.htmlThis article cites 41 articles, 15 of which can be accessed free at:
serviceEmail alerting
click herethe top right corner of the article orReceive free email alerts when new articles cite this article - sign up in the box at
http://genesdev.cshlp.org/subscriptions go to:Genes & Development To subscribe to
Copyright © Cold Spring Harbor Laboratory Press
Cold Spring Harbor Laboratory Presson October 10, 2011 - Published by genesdev.cshlp.orgDownloaded from
http://genesdev.cshlp.org/lookup/doi/10.1101/gad.6.1.135http://genesdev.cshlp.org/lookup/doi/10.1101/gad.6.1.135http://genesdev.cshlp.org/content/6/1/135#related-urlshttp://genesdev.cshlp.org/content/6/1/135#related-urlshttp://genesdev.cshlp.org/content/6/1/135.refs.htmlhttp://genesdev.cshlp.org/content/6/1/135.refs.htmlhttp://genesdev.cshlp.org/cgi/alerts/ctalert?alertType=citedby&addAlert=cited_by&saveAlert=no&cited_by_criteria_resid=genesdev;6/1/135&return_type=article&return_url=http://genesdev.cshlp.org/content/6/1/135.full.pdfhttp://genesdev.cshlp.org/cgi/alerts/ctalert?alertType=citedby&addAlert=cited_by&saveAlert=no&cited_by_criteria_resid=genesdev;6/1/135&return_type=article&return_url=http://genesdev.cshlp.org/content/6/1/135.full.pdfhttp://genesdev.cshlp.org/subscriptionshttp://genesdev.cshlp.org/subscriptionshttp://genesdev.cshlp.org/subscriptionshttp://genesdev.cshlp.org/subscriptionshttp://genesdev.cshlp.org/subscriptionshttp://www.cshlpress.com/http://www.cshlpress.com/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/subscriptionshttp://genesdev.cshlp.org/cgi/alerts/ctalert?alertType=citedby&addAlert=cited_by&saveAlert=no&cited_by_criteria_resid=genesdev;6/1/135&return_type=article&return_url=http://genesdev.cshlp.org/content/6/1/135.full.pdfhttp://genesdev.cshlp.org/content/6/1/135#related-urlshttp://genesdev.cshlp.org/content/6/1/135.refs.htmlhttp://genesdev.cshlp.org/lookup/doi/10.1101/gad.6.1.135
-
8/16/2019 Emory 1992 E.coli
2/15
A 5'-terminal stem-loop structure can
stabilize mRNA in Escherichia coli
S h e r i A . E m o r y / P h i l i p p e B o u v e t / ' ^ a n d J o e l G . B e l a s c o ^ ' ^
^Department of Microbiology and Molecular Genetics, Harvard Medical School, Boston, Massachusetts 02115 USA;
^Laboratoire de Biologic et Genetique du Developpement, Universite de Rennes I, 35042 Rennes Cedex, France
The 5'-untranslated region of the long-lived Escherichia coli ompA transcript function s as an mR NA stabilizer
capable of prolonging the lifetime in
E. coli
of a num ber of heterolo gous m essag es to whic h it is fused. T o
elucidate the structural basis of differential mRNA stability in bacteria, the domains of the ompA
5'-untranslated region that allow it to protect mRNA from degradation have been identified by mutational
analysis . The presence of a stem-loop no more than 2-4 nucleot ides from the extreme 5 ' terminus of this
RNA segment is crucia l to it s stabil iz ing inf luence, whereas the sequence of the stem-loop is relat ively
unimportant. The potential to form a hairpin very close to the 5' end is a feature common to a number of
stable prokaryotic messages. Moreover, the lifetime of a normally labile message {bla mRNA) can be
prolonged in £. coli by adding a simple hairpin structure at its 5' terminus. Accelerated degradation of ompA
mRNA in the absence of a 5'-terminal stem-loop appears to start downstream of the 5' end. We propose that
E. coli m essag es beginning wit h a single-stranded RNA segm ent of significant length are preferentially
targeted by a degradative tibonuclease that interacts with the mRNA 5' terminus before cleaving internally at
one or more distal sites.
[Key Words: mR NA s tab i l i ty ; 5 ' - u n t r an s la ted r eg io n ; Escherichia coli; g en e r eg u la t io n ; ompA; R N A s t ru c tu r e ]
Received October 18, 1991; revised version accepted November 27, 1991.
Degradation of mRNA is a cel lu lar process that is h ighly
impor tant for contro ll ing gene express ion . The capacity
to degrade mRNA is essential to the ab il i ty of cel ls to
al ter their patterns of pro tein synthesis rap id ly in re-
sponse to their changing needs. In addit ion , the s teady-
s ta te ce l lu la r co n cen t r a t io n o f a co n t in u o u s ly sy n th e-
s ized me ssage is d irect ly propor t i onal to i ts half - l ife . T he
lifet imes of indiv idual messages can vary widely with in
a s ingle cel l . For ins tance, mRNA half - l ives in Escherich-
ia coh
range f rom seconds to near ly an hour , with an
average l ifet ime of 2-4 min (Pedersen et a l . 1978; Nils-
son et a l . 1984; Donovan and Kushner 1986; Emory and
B elasco 19 9 0) . Desp i te th e imp o r tan ce of mR N A in s ta -
b i l i ty to g en e r eg u la t io n , th e mech an isms an d s t ru c tu r a l
d e te rmin an ts o f p ro k ary o t ic mR NA d ecay a r e n o t we l l
unders tood (Kennell 1986; Belasco and Higgins 1988) .
Elements at the 5 ' and 3 ' ends of bacter ial and phage
messag es h av e b een imp l ica ted in mR NA s tab i l iza t io n .
M o s t p ro k a r y o t i c m R N A s e n d i n a 3 ' - t e r m i n a l s t e m -
loop s tructure, which serves as a pro tective barr ier
against degradation by 3 ' exor ibonucleases (Belasco et a l .
1985; Mott et a l . 1985; Newbury et a l . 1987; Chen et a l .
1988).
Ho wev er , f o r b ac te r ia l mR NAs th a t en d wi th a 3 '
hairp in , there is hard ly any evidence to indicate that s ta-
bility differences in vivo result from differential rates of
ex o n u c leo ly t ic p en e t r a t io n o f th ese 3 ' s tem- lo o p s t ru c-
^Corresponding author.
tu res (Belasco et a l . 1986; Chen et a l . 1988; Chen and
Belasco 1990) . Ins tead , i t appears that the d isparate l i fe-
t imes o f mo s t p ro k ary o t ic mR NAs a r e co n t ro l led b y d e-
cay d e te rmin an ts lo ca ted u p s t r eam o f th e 3 ' en d . Esp e-
c ia l ly n o te wo r t h y a r e a smal l n u m b er o f 5 ' - t e rm in a l
mR NA seg men ts th a t h av e b een sh o wn to s tab i l ize h e t -
ero logous messages to which they are fused . These 5 '
mR NA s tab i l ize r s co mp r ise th e 5 ' - u n t r an s la ted r eg io n
(UTR) of cer tain s tab le bacter ial and phage messages ,
such as the
ompA
tran scr ip t of
E. coli,
t h e
ermC
and
ermA trans cr ip ts of Bacillus subtilis an d Staphylococcus
aureus,
and the gene 32 mes sage of phage T4, as we ll as
a 5 ' segment of the phage \ p ^ t r an scr ip t (Yamamo to an d
Im am oto 1975; Gors ki et a l . 1985; Belasco et a l . 1986;
Bechhofer and Dubnau 1987; Sandler and Weisblum
1988).
Ap p aren t ly , th ese 5 ' e lemen ts a r e ab le to p ro tec t
mR NA f ro m a t tack b y an imp o r tan t ce l lu la r r ib o n u -
c lease th a t in i t i a tes mR NA d eg rad a t io n in v iv o ; th e
id en t i ty o f th i s k ey d eg rad a t iv e en zy me r emain s u n cer -
ta in . Mo s t o f th e k n o wn 5 ' mR NA s tab i l ize r s f u n c t io n
only under special condit ions . For example, the gene 32
5 ' UTR can in c r ease messag e l i f e t imes o n ly in T4 - in -
fected cel ls (Gorski et a l . 1985) , and s tab il izat ion of
m R N A b y t h e ermC a n d ermA 5 ' UT R s req u i r es rib o -
so me s ta l l in g in d u ced b y an an t ib io t ic th a t in h ib i t s
translat ion (Bechhofer and Dubnau 1987; Sandler and
We isb lu m 1 9 88 ). S imi la r ly , messag e s ta b i l iza t io n b y th e
5 ' por t ion of the X P t r an scr ip t i s mo s t p ro n o u n ced o n ly
GENES & DEVELOPMENT 6 :135-148 © 1992 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/92 $3.00
135
Cold Spring Harbor Laboratory Presson October 10, 2011 - Published by genesdev.cshlp.orgDownloaded from
http://www.cshlpress.com/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/http://www.cshlpress.com/http://genesdev.cshlp.org/
-
8/16/2019 Emory 1992 E.coli
3/15
Emory et al.
af te r p ro lo n g ed X in f ec t io n (Y amam o to an d Ima mo to
1975).
In contras t , the ompA 5 ' UT R can s tab i l ize mR NA in
£.
coli
under normal condit ions of rap id cel l g rowth
(Belasco et al. 1986; Em ory an d Belas co 1990). Th is RN A
segment is der ived f rom the
ompA
gene transcr i p t ,
wh ich en co d es a majo r E. coli o u t e r m e m b r a n e p r o t e i n
(OmpA). In cel ls growin g rapid ly at 30°C, the
ompA
m e s -
sage decays with a half-life of ~ 17 min, making it one of
t h e m o s t s t a b l e m R N A s i n E. coli (von Gabain et al.
1983). Th e l i f e t ime o f ompA m R N A is g ro wth ra te r eg-
u lated and fal ls by as much as a factor of 4 in s lowly
growing cells (Nilsson et a l . 1984) . Growth rate regula-
tion of ompA m R NA stabil i ty , l ike the longevity of the
ompA
trans cr ip t unde r condit io ns of rap id growth , is de-
te rmin ed b y th e 1 3 3 -n u c leo t id e ompA 5 ' UTR , wh ich
can confer both of these proper t ies onto o ther messages
to which i t is jo ined (Emory and Belasco 1990) .
As the l i fet imes of a var iety of lab ile messages can be
prolonged by fusion to the ompA 5' UT R (Belasco et al.
1986;
M. Ha nse n and J . Belasco , unpub l. ) , i t appears th at
m a n y
E. coli
mR NAs a r e d eg rad ed v ia a co mmo n p a th -
way ag a in s t wh ich th i s 5 ' R NA seg men t p ro v id es p ro -
tec t io n . M essag e s tab i l iza t io n b y th e
ompA
5 ' UT R is
not an indirect consequence of close r ibosome spacing ,
wh ich th eo re t ica l ly mig h t s te r ica l ly h in d er r ib o n u c lease
attack (Lundberg et a l . 1988; Emory and Belasco 1990;
M. Han sen and J. Belasco , unpu bl. ) . Ins tead , th is R NA
seg men t seems to ac t d i r ec t ly to p ro tec t mR NA f ro m
degradation by a r ibonuclease that responds to one or
mo re e lemen ts n ear th e mR NA 5 ' en d . B o th th e seco n d -
ary s tructure of the
ompA
5 ' UT R and i ts ab il i ty to func-
t io n as a g ro wth - r a te - r eg u la ted m R N A s tab i lize r a r e
highly conserved among enter ic bacter ia (Chen et a l .
1991). However , the identi ty of the specif ic ompA 5 '
UTR s t ru c tu r a l f ea tu r es th a t a r e imp o r tan t f o r mR NA
stabil izat ion has not been explored previously .
To e lu c id a te th e mech an ism o f mR NA s tab i l iza t io n
b y th e
E. coli ompA
5 ' UTR , we h av e d is sec ted th i s R NA
seg men t to id en t i fy th e s t ru c tu r a l co mp o n en ts th a t can
p ro tec t mR NA f ro m d eg rad a t io n in £ .
coli.
Ou r s tu d ies
reveal that the presence of a s tem-loop at or very near
th e ex t r eme 5 ' t e rmin u s o f th e ompA 5 ' UT R is essen tial
for i ts function as a potent mRNA stabil izer . Fur ther-
more, the half - l i fe of a normally lab ile E. coli messag e
can be pro longed by adding a s imple s tem-loop at i ts 5 '
en d . Th e 5 ' s tem- lo o p p ro tec t s ompA mR NA f rom d eg-
radation that appears to begin downstream of the 5 ' end .
By def in ing impo r ta nt aspec ts of the sub strat e specif ici ty
of the
E. coli
mR NA d eg rad a t io n mach in ery , th ese f in d -
ings provide key insights in to both the s tructural basis of
d i f f e r en t ia l messag e s tab i l i ty in b ac te r ia an d th e mech -
an ism o f mR NA d ecay .
Results
Two domains of the o m p A 5 ' UTR are important for
mRNA stabilization
We h av e sh o wn p rev io u s ly th a t th e E. coli ompA 5 ' UT R
comprises two imperfect s tem—loop s tructures (hpl and
hp2) and two s ingle-s tranded RNA segments (ss l and
ss2) (Fig. 1; Chen et al. 1991). The ompA 5 ' UT R was
d is sec ted to id en t i fy th e s t ru c tu r a l d o main s r esp o n s ib le
for i ts act iv i ty as an mRNA stabil izer in rap id ly growing
E. coli c ells . Var ious 5 ' UT R segm ent s we re delete d f rom
a p lasmid-borne copy of a pseudo-wild- type ompA gene
{ompA
+ 4]
th a t en co d es a t r an scr ip t v i r tu a l ly id en t ica l
in sequence and s tab il i ty to the wild- type
ompA
messag e
(Emory and Belasco 1990) . Th ese dele ted se gm ents cor-
responded to one or more RNA structural domains of the
ompA 5 ' UTR u p s t r ea m of th e r ib o so me-b in d in g s i te
(Fig. 1). Plasm ids enco ding the resu l t ing m ut an t
ompA
messag es were in t ro d u ced in to E. coli s train C600S, and
to tal cel lu lar RNA was iso lated f rom rapid ly growing
cu l tu r es a t t ime in te rv a ls a f te r t r an scr ip t io n in h ib i t io n
wi th r i f amp ic in . Deg rad a t io n o f each mu tan t messag e
and of the endogenous wild- type E. coli ompA messag e
was mo n i to r ed s imu l tan eo u s ly b y S I an a ly s i s o f th ese
RN A sa mpl es (Table 1). Th e paral lel analy sis of the de-
cay of wild-type ompA mR NA p ro v id ed an in te rn a l s tan -
dard that faci l i ta ted in terpretat ion of s tab il i ty d if fer -
en ces amo n g th e v ar io u s mu tan t t r an scr ip t s .
Two d is t in c t s t r u c tu r a l d o main s ap p ear to co n t r ib u te
to mR NA s tab i l iza t io n b y th e ompA 5 ' UT R. On e is the
5 ' - te rmin a l s tem - lo o p s t ru c tu r e (h p l ; F ig . I ) , wh o se p r e -
cise deletion reduces the half-life of ompA m R N A b y a
factor of 3 to just 5.7 ± 0.4 min {ompAA64) Fig. 2, left).
20
G U
A A
A U
G C
C G
C G
G U
V A G
: A u
G U
G U
C G
U A
C G
G C
U A
1 ^
A
1 u
G C
G C
G C
G C
A
A
G
U
u
A
-
A C -6 0
C G
C G
^40
U
G
C
U
C
U
A
U
U
G
U
G
C
G UGAAGGAUUUAAC
u
G
G
A
G
A
U
A
U
U
c
A -100
U
G
120
1
GCGUAUUUUGGAUGAUAACGAGGCGCAAAAAAUG
•
hpl ss1 hp2
ss2
Figure 1. Secondary stru cture of the
ompA
5' UTR. Brackets
delineate the boundaries of the four secondary-structure do-
mains w ithin this RNA segment (Chen et al. 1991): hpl (nucle-
otides 1-63), ssl (64-74), hp2 (75-103), and ss2 (104-133). The
Shine-Dalgamo element and translation initiation codon are
underlined. Precise
5'-end
mapping by primer extension shows
that roughly equal numbers of E. coli orapA transcripts begin
with the indicated 5'-terminal nucleotide or with a cytosine
residue 1 nucleotide upstream (data not shown).
136 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Presson October 10, 2011 - Published by genesdev.cshlp.orgDownloaded from
http://www.cshlpress.com/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/http://www.cshlpress.com/http://genesdev.cshlp.org/
-
8/16/2019 Emory 1992 E.coli
4/15
mRNA stabilization by a 5' stem- loop
Th e other is a single-stranded R NA segme nt (ss2; Fig. 1),
the 3' portion of which contains the ribosome-binding
site. The stabiUty of
ompA
mRN A is halved by deleting
from ss2 an 11-nucleotide RNA segment (5'-ss2) located
a few nucleotides upstream of the
ompA
Shine-Dalgamo
sequence
{ompAA104-114)
Fig.
2,
middle). Larger dele-
tions that remove either of these two elements (hpl or
5'-ss2) also destabilize the
ompA
message
{ompAA104,
ompAAllS)
Table 1). In contrast, deletio n of the inte rnal
hairpin (hp2) within the 5' UTR and of the single-
stranded segment (ssl) between hp l and hp2
[ompAA65-
104, onipAA74-103)
has little or no effect on the degra-
dation rate of the
ompA
transcript (Fig. 2, middle; Table
1).
All of these deletions leave the
ompA
ribosome-bind-
ing site intact, and none impair translation of
ompA
mRNA (Emory and Belasco 1990; Emory 1991).
Stabilization by the 5'-terminal stem—loop
is independent of its sequence
The 5' UTRs of
ompA
mRNA from
Seiratia maicescens
an d
Enterobacter aerogenes
function as effective mRNA
stabilizers in
E. coli
despite extensive sequence differ-
ences in both hpl and hp2 (Chen et al. 1991). To assess
the importance of the sequence of hpl to
ompA
mRNA
stability, the degradation of variant
E. coli ompA
mes-
sages with altered 5'-terminal hairpins was examined
(Fig. 3; Table 2). Tru nca tion of hp l by deleting 44 nucle-
otides from its top reduces th e ^df-life of the
ompA
mes-
sage by
-
8/16/2019 Emory 1992 E.coli
5/15
Emory et al.
:
i
I
i
1
1 1
o o>
CO 00
1
t
1
1
f
1
o
I
I
f
1
1
11
I
{
•
1
•
•
Q. Q.
E E
o o
V
I
I
l i
It
m
m
m
n
• f
g 05 ^
CO 00 o
o o) en
138 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Presson October 10, 2011 - Published by genesdev.cshlp.orgDownloaded from
http://www.cshlpress.com/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/http://www.cshlpress.com/http://genesdev.cshlp.org/
-
8/16/2019 Emory 1992 E.coli
6/15
Table 1. Half-lives of mutant ompA transcripts lacking 5'
UTR segments
ompA
allele
Mutant
5' UTR half-life
structure'' (min)
Wild-type
half-life
(min)''
mRNA stabilization by a 5' stem-loop
ompA
+
4"
(pseudo-wild- type)
ompAl ^64^
(Ahpl)
ompAM04-114'-'^
(A5'-ss2)
ompAM4-103'^
(Ahp2)
ompA^65-104^
(Assl, hp2)
ompAM04^'^
(Ahpl , ssl , hp2)
ompAMlS^'^
(Ahpl , ssl , hp2, 5 '-
-ss2)
XL
\
\l
If
f
—
.._-
19 ±
1'^
SJ ±0.7
9.5 ± 0.8
14 ± 1
1 7 ± 1
4.4 ± 0.2
3.6 ± 0.6'^
N D
16 ± 2
18 ± 1
14 ± 1
19 ± 1
1 7 ± 3
1 9 ± 4
' 'The expected secondary structure of each mutant 5' UTR is
represented diagrammatically (for a detailed secondary struc-
ture of the wild-type 5' UTR, see Fig. 1). All were transcribed
from the hla promoter, whose transcription initiation site was
mapped precisely by primer extension (data not shown). (Solid
line) Retained 5' UTR segment; (dotted or absent line) deleted 5'
UTR segment; (solid rectangle) Shine-Dalgarno element.
''The half-life of the endogenous wild-type
ompA
message was
measured simultaneously in the same strain as each mutant
transcript. (ND) Not determined.
'=The ompA + 4, ompAM4-103, and ompAM04-114 5' UTRs
differ at the 5' end from that of wild-type
ompA
mRNA in that
the 5'-terminal G has been replaced by the sequence GAUCA.
The last 3 nucleotides of this substituted pentanucleotide
(UCA) are expected to augment hp 1 by 2 bp by pairing with the
last nucleotide of hpl (U) and the first 2 nucleotides of ssl (GA)
(see Table 2).
''Emory and Belasco (1990).
'^In addition to deletion of the number of 5'-terminal nucle-
otides indicated by the ompA allele name, 4 nucleotides
(GAUC) have been added at th e 5' end of ompA/: i64 mRNA, and
6 nucleotides (GAUCAG) have been added at the 5' end of
ompAM04
and
ompAMlS
mRNA.
'Segment
5'-ss2
comprises
ompA
nucleotides 104—114.
®In ompAA65-104 mRNA, the G-U pair normally at the bottom
of hpl has been changed to an A-U pair, and hpl has been
augmented at its base by 4 additional nucleotide pairs (GAUC-
hpl-GAUC). Furthermore, the deleted RNA segment (ssl-hp2)
has been replaced by 2 nucleotides (AG).
Owin g to th e p r esen ce o f th e w i ld - ty p e ompA t r an -
scr ip t in s train C600S, i t was unclear in i t ia l ly whether
th e r ap id d i sap p earan ce o f th e ex ten d ed ompA messag es
r ep resen ted mR NA d eg rad a t io n o r mere r emo v a l o f th e
single-s tranded por t ion of the 5 ' ex tension to generate a
s tab le p ro cess in g p ro d u c t r esemb l in g w i ld - ty p e ompA
U A
C G
G C
U A
G C
G C
G C
G C
A C
C G
C G
A U
C G
G A U A A G G A U .
ompA+4
G C
G C
G C
A C
C G
C G
A U
C G
G A U A A G G A U .
ompAA9-52
G CA AGG AU. . .
G A U A A G G A U .
ompAA64 ompAA29
Figure 3. Variant 5' stem-loop structures . For each ompA
message with an altered 5' hairpin, the sequence and expected
secondary structure of a 5'-terminal RNA segment tha t extends
into ssl are shown. In every case, the remainder of the mRNA
sequence is identical to that of the wild-type ompA transcript.
Adenosine, cytosine, and uridine residues susceptible to signif-
icant methylation by DMS in £.
coli
are indicated for the 5'-
terminal segment of
ompAA29
mRNA. (#) Heavy methylation;
(•) moderate methylation; (A) weak methylation.
mR NA . If s tab le , su ch a p ro cess in g p ro d u c t wo u l d b e
ex p ec ted to accu mu la te in th e ce l l to a s tead y - s ta te co n -
cen t r a t io n h ig h er th a n th a t o f i t s mo re lab i le p r ecu r so r s ,
the fu l l- length ompA t r an scr ip t s w i th 5 ' ex ten s i o n s . To
d is t in g u ish th ese p o ss ib i l i t i e s , R NA was i so la ted f ro m
an isogenic host s train (SE600) that bore a chromosomal
delet ion of the
ompA
g en e an d ca r r ied p lasm id
pOMPA-h 16a or pOMPA-l- 16b . SI and pr imer-extension
an a ly s i s of th ese s tead y - s ta te R NA sam p les sh o we d l i t t l e
or no detectable ompA + 16a
OT
ompA + 16b m R N A p r o -
cess in g p ro d u c t th a t was s imi la r in len g th to th e w i ld -
ty p e ompA mes sage (Fig . 5 , cont ro l lane s; Fig. 6) . The re-
fore,
addit io n of a shor t , s ingle-s t randed R NA se gm ent t o
Table 2. Half-lives of mutant ompA transcripts with variant
5' hairpins
ompA
al lele
ompA + 4
ompAA9-52
ompAA64*
ompAA29
m u t a n t
m e ssa g e
1 9 ± 2 ' '
13 ± 1
12 ± 1
15 ± 2
Half-life (min
1
wi ld- type
message^
N D
1 8 ± 2
1 8 ± 3
1 7 ± 2
^The half-life of the endogenous wild-type ompA message in
each strain was measured simultaneously. (ND) Not deter-
mined.
''Emory and Belasco (1990).
GENES & DEVELOPMENT 139
Cold Spring Harbor Laboratory Presson October 10, 2011 - Published by genesdev.cshlp.orgDownloaded from
http://www.cshlpress.com/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/http://www.cshlpress.com/http://genesdev.cshlp.org/
-
8/16/2019 Emory 1992 E.coli
7/15
Emory
et
al.
Table 3.
Mutant
ompA
transcripts with or without
destahihzing 5' extensions
ompA al lele
ompA
ompA
+
4
ompA + 12a
ompA + 16a
ompA + 16b
ompA + 12a*
ompA + 16b*
ompAM3a
ompAA73b
ompAA104»
5 '
U T R
st ructure^
(
L
H
ii
2
1
*
1
M u t a n t
half-life
(min)
19
±2
3. 2
± 1.1
5.8
± 0.3
6. 2
+ 0.8
13
± 1
13
+ 1
6. 0
± 0.3
13
+ 1
24
± 3
Wild- type
half-life
(min) ' '
17
± 2
N D
15
± 3
16
± 2
12
+ 1
17
± 1
15
+ 1
14
± 1
15
+ 1
19
+ 3
^The secondary structure
of
each muta nt 5' UTR is represented
diagrammatically. Only 5'-terminal single-stranded extensions
>2 nucleotides in length are show n. (Solid rectangle) Shine-
Dalgamo element;
[1] ompA
hpl; (2)
ompA
hp2; (*) synthetic
hp*.
''The half-life of the endogenous wild-type ompA message in
each strain was measured simultaneously. (ND) Not determined.
•^Emory and Belasco (1990).
t h e
5' end of
ompA
m R N A d e s t ab i l iz e s
th e
en t i r e t r an -
script.
Five unpaired nucleotides at the 5' end can destabilize
o m p A
mRNA
A n
ompA
d e l e t i o n m u t a n t
[ompAAJSa]
lack in g h p l and
s s l an d b eg in n in g in s tead w i th h p 2 - s s2 p r eced ed b y ju s t
5 n u c leo t id es (GAUC A) d ecay s w i th a sho rt half-life
(6.0
±
0 .3 min ) co m p ared wi th w i ld - ty p e ompA m R N A
(14 ± 1 min) (Table 3; Fig. 7). As a 5 ' h a i rp in of any k ind
a n d
ss 2
ap p ear
to be the
only features
of the
ompA
5 '
U T R t h a t are n ecessa ry for m R N A l o n ge v i ty , the shor t
l i f e t ime
of
ompAA73a mR NA su g g es ts th a t
a
5 ' - te rmi-
n a l s in g le - s t r an d ed seg men t o n ly 5 n u c l e o t i d e s in len g th
m a y
be
e n o u g h
to
acce le r a te ompA m R N A d eg rad a t io n .
Th is h y p o th es i s was co n f i rmed by m e a s u r i n g th e decay
ra te of a m e s s a g e
{ompAAJSb]
t h a t is id en t ica l to
ompAA73a ex cep t th a t it h as o n ly 1 nuc leo tide (A) up-
s t r e a m of hp 2- ss 2 (Table 3). As expec ted , t he half - l i fe of
ompAA73b m R N A (1 3 ± 1 min ) p ro v ed to be a b o u t as
long as t h a t of the wi ld - ty p e
ompA
tran scr ip t (15 ± 1
min) (Fig.
7).
Th e se f in d in g s in d ica te th a t
as few as 5
u n p a i r ed b ases at the 5 ' end are suff icien t to target
ompA
m R N A
for
rap id d eg rad a t io n
in
E. coli.
Addition
of a
5'-terminal stem-loop
can
prolong
the lifetime
of a
normally short-lived mRNA
A s a f ina l d emo n s t r a t io n th a t a 5 ' s t e m - l o o p and the
s in g le - s t r an d ed R NA seg men t en co mp ass in g th e ompA
r ib o so me-b in d in g s i te are suff icien t for th e full efficacy
of th e
amp A
mR NA s tab i l ize r , a m i n i m a l 5 ' U T R c o m -
pr is ing only a sy n th e t ic te rm in a l h a i rp in (hp *) and the
ompA
s s2 seg me n t was ev a lu a ted . As sh o w n
in
Figure
7,
a n ompA t r an scr ip t b ear in g th i s mi n im al
5' UT R
{ompAA104*) Tab le 3 )
is as
s tab le
as
th e w i ld - ty p e mes -
sage .
It is also f ive t im es m ore s ta b le tha n a s imi la r m es -
00 1 -
10
N ^ ^ ~ ^
•
o'\
ô
A o m p A + 1 2 a »
• o m p A + 1 6 a
O o m p A + 1 2 a
•_
•
^
•"
5 CAGACUUUACAUC
10 20
Time (min)
loo t
10
N
•
A o m p A + 1 6 b »
• o m p A + 1 6 b
,
\
s
• N
\
^
5 ' g c a G A U C U A U A C U A U A A C C . . .
^_*
0 15 30 45
Time (min)
Figure
4.
Decay
of
mutant ompA transcripts with
5'
exten-
sions.
After transcription inhibition, total cellular RNA was
isolated periodically from
E.
coli strain C600S containing
pOMPA-l-12a, pOMPA-Hl6a, or pOMPA -l-12a»
[top]
or con-
taining pOMPA-l- 16b or pOMPA-l- 16b»
[bottom].
The mutant
transcripts and the endogenous wild-type ompA message were
detected by SI analysis, and mRNA concentration was plotted
semilogarithmically as a function of time. Beside the graphs are
drawn the 5' extensions of ompA
+ 12a* [top]
and ompA -\-16b*
[bottom]
mRNA. The 5' extensions of
ompA + 12a
and
ompA-\-16b mRNA
are
related
to
those
of
ompA-\-12a*
and
ompA + 16b* mRNA, respectively, but include only those nu-
cleotides shown in uppercase letters. The 5' extension of
ompA
+ 16a
mRNA is identical to that of ompA
+ 12a
except for
4 additional nucleotides (GAUC) at th e 5 ' end. Underlined nu-
cleotides
at
the 3' end
of
each extension are expected to base-
pair with the first 1-2 nucleotides of 5' UTR segment ssl . Tran-
scription initiation sites
for
these messages were mapped pre-
cisely by primer extension (data not shown). The measured half-
lives were 3.2 ± 1.1 min for ompA-\-12a mRNA (O) and 15 ± 3
m in for wild-type ompA m RNA
[top],
5.8 ± 0.3 mi n for
ompA + 16a
mRNA ( ) and 16 ± 2 min for wild-type
ompA
mRNA [top], 13
± 1 min for
ompA-\-12a* mRNA
(A) and
17
±
1 min for wild-type ompA mRNA [top], 6.2
±
0.8 min for
ompA
+
16b mRNA ( ) and 12 ± 1 m in for wild-type ompA
mRNA [bottom], and 13 ± 1 min for ompA
+
16b* mRNA (A)
and 15 ± 1 min for wild-type
ompA
mRNA
[bottom].
140 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Presson October 10, 2011 - Published by genesdev.cshlp.orgDownloaded from
http://file//-/-16b*http://www.cshlpress.com/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/http://www.cshlpress.com/http://genesdev.cshlp.org/http://file//-/-16b*
-
8/16/2019 Emory 1992 E.coli
8/15
mRNA stabilization by a 5' stem -loop
sage
{ompAA104)
with a 5' UTR consisting only of the
ss2 segment (Table 1; Fig. 7). Thi s finding confirms tha t
a simple 5'-terminal hairpin and the
ompA
ss2 segment
together are as effective as the com plete
ompA
5' UTR at
stabilizing mRNA. Moreover, when the same synthetic
stem-loop (hp*) was introduced at the 5' terminus of
bla
mRNA, which encodes p-lactamase, the resulting
bla202
messages were found to decay with a half-life of
6.8 ± 0.4 min (Fig. 7), about twice the lifetime of wild-
type
bla
mRNA (3.7 ± 0.3 min) (Fig. 7; von Gabain et al.
1983).
Thus, the ability of
bla
mRNA to resist degrada-
tion in
E. coli
can be enhanced simply by adding a st em -
loop at its 5' end.
Degradation in the absence of a 5'-terminal stem-loop
appears to begin downstream of the
ompA 5'
end
In principle, accelerated degradation of
ompA
mRNA
lacking a 5'-terminal hairpin migh t be initiated either by
a 5' exonuclease (defined here as a ribonuclease that re-
moves single nucleotides sequentially from the RNA 5'
terminus) or by a 5'-end-dependent endonuclease that
cuts internally but is sensitive to the presence or absence
of secondary structure at the 5' terminus of mRNA. Of
these two possibilities, degradation by a single-strand-
specific 5' exonuclease seems less likely because no 5'
exoribonuclease has ever been detected in £.
coli
and
because destabilizing single-stranded extensions added
upstream of
ompA
h pl are not preferentially removed
from the 5' end of
ompA + 16a
a nd
ompA + 16b
mRNA
to generate wild-type-like
ompA
mRNA as a readily de-
tectable processing product. As 5' exoribonucleases, by
definition, digest RNA from the 5' end, the plausibility
of 5'-exonucleolytic initiation of mRNA decay can be
tested further by determining whether degradation of the
5' mRNA segment is faster or slower than decay of the
rest of the message.
Degradation of three different
ompA
messages with-
out a 5'-terminal stem-loop
{ompAA64, ompA + 16a,
an d
ompA
+
16b
mRNA) was monitored by SI analysis
with a mixture of two DNA probes, one complementary
to the 5'-terminal segment of these transcripts (the 5'
UTR and codons 1-37) and the other complementary to
an internal
ompA
segment spanning codons 67-295. By
combining these two probes in each SI protection assay,
differences in the relative decay rates of the two RNA
segments could be measured w ith considerable accuracy.
As observed previously for wild-type
ompA
mRNA (von
Gabain et al. 1983), the internal segment of all three of
these m uta nt messages was found to decay more rapidly
than the corresponding 5'-terminal segment (Fig. 8). To
demonstrate that the differential lifetimes measured for
the 5'-terminal and internal
ompA
mRNA segments are
independent of the relative lengths of the two DNA
probes used simultaneously for SI protection, the mea-
surements were repeated with the same 5'-terminal
probe and a mu ch shorter internal probe com plementary
to
ompA
codons 249 -295 . Regardless of the length of
either the internal probe or the RNA segment with
which it hybridized, the same segmental difference in
stability was observed for each of the three mutant
ompA
messages (Fig. 8). In every one of these six inde-
pendent experiments, the relative concentration of the
internal mRNA segment declined steadily to a level be-
tween one-third and one-half that of the 5'-terminal seg-
ment within 30 min after transcription inhibition.
Thus,
it appears for three different labile
ompA
tran-
scripts lacking a 5'-terminal stem-loop that degradation
of an internal RNA segment precedes decay of the 5'
UTR. This conclusion is consistent with our finding, for
a number of mutant
ompA
messages, that there is no
correlation in rapidly growing cells between m RN A half-
life and the relative steady-state concentration of the
major products of endonucleolytic cleavage within the 5'
UTR (data not shown; Lundberg et al. 1990); only when
ompA
mRNA decay accelerates in slowly growing cells
does its rate appear to be controlled by 5' UTR cleavage
(Melefors and von Gabain 1988). Together, our data sug-
gest that under conditions of rapid bacterial growth, the
function of the 5'-terminal
ompA
hairpin is to protect
the message from degradation by a ribonuclease that ini-
tiates decay downstream of the 5' end.
Discussion
A key to understanding the molecular basis for differen-
tial mRNA stability in bacteria is to identify the struc-
tural features of stable messages that are responsible for
their unusual longevity in vivo. The studies reported
here show th at a 5'-termin al stem—loop is bot h crucial to
the function of the
ompA
5' UTR as a potent mRNA
stabilizer and sufficient to prolong the lifetime of a nor-
mally labile
E. coli
message. Remarkably, the 5'-termi-
nal stem-loop shields
ompA
mRNA from attack by a
cellular ribonuclease that appears to initiate degradation
far from the 5' end.
Our data also indicate that the
ompA
mRNA stabilizer
is bipartite, with a second, distinct domain of functional
importance. Both the 5' stem -loop and a single-stranded
RNA segment in the vicinity of the
ompA
ribosome-
binding site and its flanking sequences contribute to th e
remarkable longevity of the
ompA
transcript, and to-
gether they are sufficient for full activity of the
ompA
mRN A stabilizer. Alone, each of these elem ents can sta-
bilize mRNA to a lesser extent. Thus,
ompA
transcripts
that have both stabilizing elements (wild-type,
ompAA65-104]
are more stable than
ompA
messages
lacking either the 5' hairpin or
5'-ss2
[ompAA64,
ompAA104, ompAA104-114]
which, in turn, are more
stable than a message that lacks both of these RNA seg-
ments
[ompAAllS].
Sim ilarly, the half-life of
bla
mRNA , wh ich is increased twofold simply by adding a 5'
s tem-loop
[bla202],
increases about fivefold when its 5'
UTR is replaced with the entire
ompA
5' UTR (Belasco
et al. 1986; Emory and Belasco 1990).
The longevity of
ompAA104*
mRNA, in which hpl,
ss l ,
and hp2 are replaced by a synthetic stem -loo p, ru les
out any contributio n to
ompA
mRNA stability from pos-
GENES & DEVELOPMENT 141
Cold Spring Harbor Laboratory Presson October 10, 2011 - Published by genesdev.cshlp.orgDownloaded from
http://www.cshlpress.com/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/http://www.cshlpress.com/http://genesdev.cshlp.org/
-
8/16/2019 Emory 1992 E.coli
9/15
Emory et al .
V
• %.
»• IS i
1
*< t
s
1
f4 1
f
* J
i » •
f* »
f : • , ' : •
C D
+
<
Q .
E
O
mt<
8
1
1
^11
1
1
1 < i
V 0 V
• 0
• •
U D U D sC D
U D O U U
» •
*tt f^
0 •
•
-
8/16/2019 Emory 1992 E.coli
10/15
tnRNA stabilization by a 5' stem-loop
sib le s i tes of t ranslat ion in i t ia t ion upstream of ss2 .
These s i tes , which were identif ied on the basis of se-
quence analysis (Movva et a l . 1980) , probably are inac-
cess ib le to r ib o so me s d u e to o cc lu s io n b y in t r a mo le cu la r
base-pairing (Chen et al. 1991).
Stabilization of mRNA by a
5 '
stem-loop
Th e p resen ce o f a 5 ' - t e rmin a l s tem - lo o p can p ro lo n g th e
lifet ime of ompA m RN A by as m uc h as a factor of 5 . The
lo ca t io n of th i s s tem - lo o p a t o r v e ry n ear th e 5 ' t e rm in u s
is crucial to i ts s tab il izing ef fect , whereas the sequence
of th is hairp in and i ts posi t ion relat ive to the r ibosome-
binding s i te appear to be of l i t t le consequence to message
s tab i l i ty . Up to two u n p a i r ed n u c leo t id es u p s t r eam o f
the 5 ' hairpin (e.g. , ompA+4, ompAA29] are to lerat ed
wi th o u t an y r ed u c t io n in mR NA s tab i l i ty , b u t th e ad d i -
t io n o f 1 0 -1 5 u n p a i r ed n u c leo t id es o f r an d o m seq u en ce
to the
ompA
5 ' end
{ompA + 12a, ompA + 16a,
ompA + 16b] is as destab il izing as delet ion of the 5 ' hair -
p in . Moreover , the shor t l i fet ime of
ompAA73a
m R N A
v er su s ompAAlSb m R N A in d ica tes th a t 5 u n p a i r ed
bases at the 5 ' end are suff icien t to accelerate degrada-
tion of ompA mR NA , as th ese two messag es a re id en t i -
cal in sequence except for the number of nucleotides that
p r eced e h p 2 - s s2 . Th ere fo re , th e min imu m n u mb er o f 5 ' -
t e rmin a l u n p a i r ed n u c leo t id es th a t can d es tab i l ize ompA
m R N A i n E. coli is no more than f ive and may be as few
as three. The destab il izing ef fect of unpaired bases at the
ompA
5 ' t e rmi n u s d o es n o t ap p ear to b e seq u en ce sp e-
cific.
These f indings explain our previous observation that
the 5 ' UTRs of the
S. maicescens
an d
E. aerogenes ompA
t rans cr ip ts function as very ef fective mR N A stabil izers
in
E. coli
(Chen et a l . 1991) . Alt hou gh thes e
ompA
5 '
UTRs are quite s imilar in secondary s tructure to that of
E. coli and the sequence of ss l and ss2 is h ighly con-
served (80%), there is ex tensive sequence d ivergence in
hpl and hp2 (Chen et al. 1991). The efficacy of the Ser-
337 —
296 —
M
-
MM
m^
Figure 6. 5'-End m apping of ompA + 16a and ompA + 16b
mRNA at steady state. Total cellular RNA was isolated from E.
coli
strain C600S or from strain SE600 containing either
pOMPA-l-16a or pOMPA+16b. The wild-type and mutant
ompA transcripts were detected by SI analysis with 5'-end-la-
beled DNA probes complementary to the first 0.32-0.34 kb of
each message. Calibration is in nucleotid es. A set of m olecular
size standards (lane M) was generated as in Fig. 2.
ratia an d Enterobactei ompA 5' U T R s i n E. coli, d esp i te
n u m ero u s h p l an d h p 2 seq u en ce d i f fe r ences , i s n o w u n -
ders tandable in l igh t of our present f inding that v ir tually
any 5 ' - terminal hairp in fo l lowed by ss2 is suf f icien t to
p ro tec t ompA mR NA f rom rap id d eg rad a t io n .
As th e seq u en ce o f th e 5 ' s tem- lo o p i s r e la t iv e ly u n -
imp o r tan t to messag e s tab i l i ty an d i t s s tab i l iz in g in f lu -
ence is negated by the presence upstream of several un-
paired nucleotides , i t ev idently is the absence of a s ig-
n i f ican t s in g le - s t r an d ed R NA seg men t a t th e 5 ' t e rmin u s
ra th er th an th e p r esen ce o f ompA h p l t h a t m a k e s m e s -
sages bear ing the ompA 5 ' UT R res i s tan t to d eg rad a t io n .
We n o te th a t s ev era l o th er lo n g - l iv ed p ro k ary o t ic mes -
sages in their mature form (e.g . , E. coli papA m R N A , T 4
g en e 3 2 mR NA, an d Rhodobacter capsulatus pufBA
mR N A) h av e th e p o ten t ia l to fo rm a h a i rp in s t ru c tu r e
with in 1-4 nucleotides of the 5 ' end (Fig . 9) , (Youvan et
al.
1984; Belasco et al. 1985; Gorski et al. 1985; Baga et
al.
1988; Mc Phe ete rs et a l . 1988) . Tog ethe r wit h our
present f indings, th is observation suggests that a 5 '
Figure 5. Me thyla tion of
ompA
transcripts bearing 5' extensions.
[Top]
Total cellular RNA was isolated from £.
coli
strain SE600
containing pOMPA+16a
[left]
or pOMPA-t-16b
(right)
after treating each culture w ith DMS (DMS/in vivo). Alternatively, total
cellular RNA was purified from the sam e cultures with out prior DMS treatmen t and then a lkylated in vitro with either DM S (DMS/in
vitro) or CMCT. Sites of alkylation within the 5' UTR of ompA + 16a and ompA + 16 b mRNA were mapped by primer extension with
AMV reverse transcriptase, using complem entary oligodeoxynucleotides (AGCGA AACCAG CCAGTGC CACTGC or GATAACACG-
GTTAAATCCTTCAC) that annealed to mRNA sequences either 22-45 nucleotides downstream or 50-72 nucleotides upstream of
the translation initiation codon of
onipA + 16a
and
ompA + 16b.
Gel electrophoresis was performed in parallel with the products of
primer extension on unmethylated RNA templates in the presence (lanes
U, G, C, A]
or absence (control lane) of dideoxynucleoside
triphosphates. The sequencing lanes are labeled to indicate the sequence of the RNA, not the c omplem entary D NA. Blockage of prime r
extension by an alkylated RNA base results in a complementary DN A fragment t hat is 1 nucleotide shorter than that arising from
incorporation of a dideoxynucleotide opposite the same RNA base. Calibration is in nucleotides from the mRNA 5' end. (Bottom)
Summary of alkylation data for the 5' UTRs of ompA
+
lSa (left) and onipA
+
16b (right) mRN A. Adenosine, cytosine, and u ridine
residues susceptible to significant me thylation by DMS in
E. coli
are indicated. (•] Heavy methyla tion; (•) m oderate methy lation; (A)
weak me thylation . Also labeled are uridine and guanosine residues that are significantly alkylated by CMCT in vitro. (•) Heavy-to-
moderate alkylation; ( 0) weak alkylation. Com parisons of susceptibility to alkylation are meaningful only among nucleotides wit hin
the same message and at a similar distance from the annealed primer. Arrows identify the principal sites of termination by reverse
transcriptase on an unalkylated template; the structural significance of primer extension products ending at these sites on alkylated
RNA is often difficult to assess. Two of these termination sites on unalkylated RNA (at nucleotides 84 and 123) correspond to known
RNase K cleavage sites within the
ompA
5' UTR (Lundberg et al. 1990); whether or not the others also correspond to RNA 5' ends is
uncertain. N o differences w ere observed between the alkylation patterns of the ompA + 12a (data not shown) and ompA + 16a5' UTRs.
GENES & DEVELOPMENT 143
Cold Spring Harbor Laboratory Presson October 10, 2011 - Published by genesdev.cshlp.orgDownloaded from
http://www.cshlpress.com/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/http://www.cshlpress.com/http://genesdev.cshlp.org/
-
8/16/2019 Emory 1992 E.coli
11/15
Emory et al.
Figure 7. Degrad ation of a variant ompA
transcript bearing a minimal 5' stabilizer
preceded by 0, 1, or 5 nucleotides and of
bla mRNA with an added 5'-terminal
stem-loop. After transcription inhibition,
total cellular RNA was isolated periodi-
cally from E. coli strain C600S containing
pOMPAA 73a (top left), pOMPAA 73b (bot-
tom
left),
pOMPAA304*
(top
hght),
pBLA200
{bottom hght),
or pBLA202
{bot-
tom hght).
Transcripts of the
ompA,
ompAA73a, ompAA73b, ompAA104*,
bla200, an d bla202 genes were detected by
SI analysis, and mRNA concentration was
plotted semilogarithmically as a function
of time. Th e mea sured half-lives we re
6.0
±
0.3
min for ompAA73a
mRNA
(A)
and 14 ± 1 min for wild-type ompA
mRNA (• ) {top left), 13 ± 1 min for
ompAAJSb mRNA (A) and 15 ± 1 min for
wild-type ompA mRNA ( ) (bottom
left),
24 ± 3 min for
ompAA104*
mRNA (A)
and 19 ± 3 min for wild-type ompA
mRNA
(• )
{top hght),
3.7 ±
0.3
min for
bla200 mRNA (wild-type) (•) and 18 ± 2
min for wild-type ompA mRNA {bottom
hght), and 6.8 ± 0.4 min for
bla202
mRNA
(A) and 17 ± 2 m in for wild-type
ompA
mRNA {bottom hght). For comparison, de-
cay data for ompAA104 mRNA (•) are also
shown {top hght; Table 1).
100
10
^r
A ompA A73o
• ompA
100
g 10
A ompA A l 04-*
• ompA A104
• ompA
0 15 30
Time min)
15
30 45
Time (min)
10011
10
• b la202
• b la200
2 0 40
Time min)
10
Time min)
20
S t e m - l o o p m a y be of g en era l imp o r tan ce as a m e a n s b y
wh ich p ro k ary o t ic o rg an isms
can
se lec t iv e ly en h an c e
th e s tab i l i ty of m R N A .
Mechanism
of
mRNA degradation
in E.
coli
Th e ro u g h ly ad d i t iv e co n t r ib u t io n s to m R N A s tab i l i ty of
two d is t in c t s t r u c tu r a l d o main s of the ompA 5' UTR
suggest that the degradation of messag es th a t can be sta-
b il ized
by
fusion
to
t h i s
5' UTR is
in i t i a ted e i th e r
by a
s in g le r ib o n u c lease wh o se r a te of a t tack i s d e te rm in ed b y
a c o m b i n a t i o n of 5' UT R s t ru c tu r a l f ea tu r es or by a pair
of r ibonucleases with d if fer ing specif ici t ies , bo th of
w h i c h m u s t be b lo ck ed to a c h i e v e p r o n o u n c e d m R N A
lo n g ev i ty . Un d o u b ted ly th e r e are u n s tab le t r an scr ip t s
wh o se lab i l i ty r esu l t s f ro m a t tack by a dif ferent r ibonu-
c lease th a t is n o t sen s i t iv e to s t ru c tu r a l f ea tu r es n e ar the
m R N A 5 ' en d ; ad d i t io n of a 5 ' - te rmi nal ha irp in or fusion
to th e en t i r e ompA 5' U T R is not ex p ec ted to stab il ize
su ch messag es . Nev er th e les s ,
the
l i f e t imes
of
m a n y
shor t- l ived E. coli R NAs ( in c lu d in g bla, lacZ, an d phoA
mRNA) are contro lled by the r ibonuclease(s) impeded by
s u b s t i t u t i n g
the ompA 5' UT R or
ad d in g
a
s i m p l e
5'-
t e rmin a l s tem- lo o p (B e lasco et al. 1986; M. Han sen an d
J. Belasco , unpu bl. ) ; and , in p r in c ip le , all long- l ived mes-
sag es mu s t h av e s t ru c tu r es , s eq u en ces , or bou nd factors
th a t sh ie ld th em f ro m a t tack by th i s en zy m e.
Our data therefore suggest that E. coli c o n t a i n s a r ibo-
n u c lease th a t p r e f e r s to a t tac k messag es b eg in n in g w i th
> 2 - 4 u n p a i r e d n u c l e o t i d e s . T h i s e n z y m e is no t a 3 ' ex-
o n u c lease , as m u l t i p l e l i n e s of ev id en ce in d ica te th a t bla
mR NA d eg rad a t io n , wh ich is s lo wed by ad d i t io n of a
5 ' - te rmin a l h a i rp in , b eg in s u p s t r eam of th e 3 ' end (von
G a b a i n et al. 1983; Belasco et al. 1986). Instead, the sen-
s i t iv i ty
of
th i s r ib o n u c lease
to
base-pair ing
at
t h e m R N A
5 '
te rmin u s in d ica tes th a t
it is
e i th e r
a 5 '
ex o n u c lease
or
a 5 '- en d -d ep en d en t en d o n u c lease th a t cu ts in te rn a l ly b u t
in te r ac ts , at leas t in i t ia l ly , w i th th e 5' en d of m R N A .
Th ere a r e a n u m b e r of r easo n s to d o u b t th a t the r ib o n u -
c lease o b s t ru c ted by the ompA 5' UT R is a 5' ex o n u -
c lease th a t r em o v es s in g le n u c leo t id es seq u en t ia l ly f rom
th e R NA 5 ' end. First, as sh o w n p rev io u s ly for the wild-
ty p e
ompA
trans cr ip t (von Gab ain et al. 1983), rapid deg-
r ad a t io n of onipAA64, ompA + 16a, and ompA + 16b
m R N A a p p e a r s to b e gi n d o w n s t r e a m of th e 5' U T R . In
addit ion , degradation of ompA tran scr i p ts wi th ei ther of
two s in g le - s t r an d ed 5' ex ten s io n s d o es no t g en era te
wi ld - ty p e ompA m R N A as a readily dete ctab le decay in-
te rmed ia te ; s ig n i f ican t accu mu la t io n of s u c h an in ter -
m e d i a t e m i g h t
be
ex p ec ted
if
th ese messag es were
de-
graded by a h y p o t h e t i c a l 5 ' e x o n u c l e a se t h a t is imp ed ed
w h e n it e n c o u n t e r s an R NA s tem - lo o p s t ru c tu r e . Th i rd ,
n o 5 ' ex o r ib o n u c lease h as ev er b een d e tec ted in E. coli.
Fin a l ly , in ac t iv a t io n of the
ams/rne
g en e p ro d u c t in E.
coli s tab i l izes man y R NAs ( in c lu d in g ompA mR N A) an d
in h ib i t s th e i r c leav ag e in v iv o by R N a s e E or R N a s e K,
t w o £. coli en d o n u c leases th a t ap p ear to be in te r r e la ted
a n d m ay ev en be ident ical (Apir ion 197 8; O no and Ku-
wan o 1 9 7 9 ; Lu n d b erg et al. 1 9 9 0 ; Mu d d et al. 1990a,b;
B ab i tzk e and Ku sh n er 1 9 9 1 ; L in -C h ao and C o h e n 1 9 9 1 ;
144 GENES DEVELOPMENT
Cold Spring Harbor Laboratory Presson October 10, 2011 - Published by genesdev.cshlp.orgDownloaded from
http://www.cshlpress.com/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/http://www.cshlpress.com/http://genesdev.cshlp.org/
-
8/16/2019 Emory 1992 E.coli
12/15
mRNA stabilization by a 5' stem-loop
100
10 -
ompA+16o
• 5'
• in ternal (6 7- 29 5)
100 '
>A onr)pA+1 6a
10
1
V ^
k N.
• 5'
A in te rna l (249 -295 ) .
10 20
Tlnne (min)
30 10 20
Tim*
(min)
30
100
10
onnpA+16b
• 5'
• in ternal (67 -29 5) f
IOOI 'A
10
ompA+16b
• 5'
• in ternal (2 49 -2 95 )
10 20
TlnDe (min)
30
10 20
Time (min)
30
l O O o
10
o m p M 6 4
• 5'
• in ternal (67 -29 5)
100
10
ompAA64
• in ternol (24 9- 29 5)
10 20
Time (min)
30
10 20 30
Time (min)
Figure 8. Segmental differences in stability with in mu ta nt
ompA transcripts lacking a 5'-terminal stem-loop. After tran-
scription inhibition, total cellular RNA was isolated periodi-
cally from E. coli strain SE600 containing pOMPA-l-16a,
pOMPA-l-16b, or pOMPAA64. The degradation of 5'-terminal
and internal segments of each mutant ompA transcript was
monitored simultaneously by SI analysis with a mixture of two
5'-end-labeled DNA probes, one complementary to the 5' UTR
plus codons 1-37 and the other comp lementary either to codons
67-295 or to codons 249-295.
[Top]
Semilogarithmic plots of
the decay of 5'-terminal (#) and internal (A) segments of
ompA + 16a [top], om pA + 16b {middle], and ompAA64 {bottom]
mRNA, as determined by Si analysis with tw o pairs of segment-
specific probes. In every case, decay of the internal segment is
faster than decay of the 5'-terminal segment.
{Below]
Map of the
5'
and internal
ompA
mRNA segments detected by SI analysis
with each probe. (Thin lines)
ompA
UTRs; (open rectangle)
ompA protein-coding region; (arrowhead) ompA 3' end; (thick
lines) ompA mRNA segments complementary to each of the
three probes.
in g en d o n u c leo ly t ic c leav ag e a t o n e o r mo re s i te s d o wn -
stream of the 5 ' end . The precise location of these s i tes is
cu r r en t ly u n d er in v es t ig a t io n .
I f th e en zy me th a t i s imp ed ed b y th e ompA 5 ' UT R is
an endonuclease ( i .e . , a r ibonuclease that cu ts in ter -
nal ly) ,
th en th e r a te o f mR NA c leav ag e b y th i s en d o n u -
clease is determined not only by the presence of RNA
sequences that can serve as cleavage s i tes but also by the
p resen ce o r ab sen ce o f s t r u c tu r es n ear th e mR NA 5 ' en d
that can contro l access to these s i tes . Indeed , the ab il i ty
o f a 5 ' - t e rmin a l s tem- lo o p to sh ie ld an en t i r e t r an scr ip t
f ro m d eg rad a t io n su g g es ts th a t th e p r imary d e te rmin an t
of mRNA half - l i fe may of ten be the accessib i l i ty of po-
ten t ia l c leav ag e s i tes r a th er th an th e i r n u mb er o r th e i r
in tr i ns ic susce ptib i l i ty to cleavage. Th is pro t ecti ve ef fect
o f a 5 ' s tem- lo o p su g g es ts a d eg rad a t io n mech an ism fo r
E. coli messages that can be s tab il ized by fusion to the
ompA
5 ' UTR , wh ere b y a r a te -d e te rm in in g , 5 ' -en d -d e-
p en d en t en d o n u c lease f i r s t b in d s in a s t r u c tu r e -d ep en -
d en t s tep to th e m R N A 5 ' en d o r to a p ro te in b o u n d th ere
o p
mRNA
249-295
67-295
intemal
u
A
U
U
G
G
U
C
U
A
G
JO
G
U
A
A
U
C
G
G
G
U
C
AUUUUA
G
G
G
U
G
G
G
C
A
C
C
U
C
D
GO
C
c
c
c
u
c
G
0
G
G
0
G
G
A
AAOGCG
0 A
G 0
C G
„ G
M ac k ie 1 9 9 1 ; M ele fo r s an d v o n Gab a in 1 9 9 1 ; N i l s so n
and Uh lin 1991; Tara sevic iene et a l . 1991) ; therefore,
ei ther or both of these endonucleases are candidates for
th e k ey en zy me( s ) wh o se ac t io n i s imp ed ed b y th e ompA
5 ' UTR. I t seems l ikely , therefore, that the ompA 5 ' UT R
p ro tec ts th e ompA mess ag e f ro m in i t ia l , ra te -d e te rmin -
papA
pufBA
gene 32
Figure 9. Stem-loop structures near the 5 ' end of stable
prokaryotic mRNAs. The sequence and likely secondary struc-
ture is shown for a 5'-terminal untranslated segment of E. coli
papA mRNA, R. capsulatus pufBA mRNA, and T4 gene 32
mRNA (McPheeters et al. 1988).
GENES
&,
DEVELOPMENT
145
Cold Spring Harbor Laboratory Presson October 10, 2011 - Published by genesdev.cshlp.orgDownloaded from
http://www.cshlpress.com/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/http://www.cshlpress.com/http://genesdev.cshlp.org/
-
8/16/2019 Emory 1992 E.coli
13/15
Emory et al.
an d th en seek s o u t c leav ag e s i tes d o w n s t r ea m. Th e R NA
fragments thereby generated could be rap id ly degraded to
mo n o n u c leo t i d es b y fu r th er en d o n u c lease c leav ag e an d
ex o n u c lease d ig es t io n .
Th e in f lu en ce o f a 5 ' - t e rmin a l s t r u c tu r a l e lem en t o n
th e s tab i l i ty o f mR NA b ear s a s t r ik in g r esemb lan ce to
the select ive degradation of pro teins on the basis of their
amin o - te rmin a l amin o ac id r es id u e (Var sh av sk y e t a l .
1988). Moreo ver , the appa rent ef fect of an up stre am
s t ru c tu ra l e lemen t o n mR NA c leav ag e a t d o wn s t r eam
s i tes i s r em in isc en t o f ce l lu la r r eg u la to ry m ech an is ms
th a t co n t ro l t r an scr ip t io n an d t r an s la t io n r a tes v ia se -
q u en ce e lemen ts lo ca ted u p s t r eam o f in i t i a t io n s i tes f o r
R N A o r p ro te in sy n th es i s .
Mateiia ls and methods
Bacterial strains and plasmids
E. coli K-12 strains C600S and SE600 (Nilsson et al. 1987) are
streptomycin-resistant, supE44~ variants of strain C600. SE600
is identical to C600S except for a deletion of the chromosomal
ompA
gene (Emory and Belasco 1990).
Plasmid constructions were confirmed by restriction map-
ping and DNA sequencing. When necessary, protruding DNA
ends were made blunt prior to ligation by treatment with T4
DNA polymerase, the Klenow fragment of DNA polymerase I,
or mung bean nuclease. Oligonucleotide-directed mutagenesis
was performed as described previously (Kunkel 1985; Naka-
maye and Eckstein 1986) using purified oligodeoxynucleotides
synthesized on an Applied Biosystems 381A instrument.
Plasmid pOMPA-l-4 is a pBR322 derivative that encodes a
pseudo-wild-type ompA message transcribed from the bla pro-
moter (Emory and Belasco 1990); this plasmid has a
Bell
site at
th e
ompA + 4
transcription initiation site. Plasmid pOMPA-l-
4-1-3 was constructed by oligonucleotide-directed insertion of 3
bp (ATC) into pOMPA-l-4 to create a second Bell site at the
promoter-distal end of the segment encoding hpl (GTGAA-*
GTGATCAA). Deletion of the 0.07-kb Bell fragment of
pOMPA-f4-l-3 generated pOMPAA64. Plasmids pOMPAA29
and pOMPAA73a were c onstructed by deleting from pOMPA -I- 4 a
0.03-kb Bell (filled-in)-£coRV fragment or a 0.08-kb Bell (filled-
in)-Hphl (T4 DNA polymerase) fragment, respectively, thereby
reconstituting the Bell site. Plasmid pOMPAA73b was con-
structed by cleaving pOMPAA73a wi th Bell, treating the linear-
ized DNA with mung bean nuclease, and religating the result-
ing blunt ends. The 5' terminus of
ompAAlSb
mRNA mapped
to 1 nucleotide upstream of ompA hp2, and DNA sequencing of
pOMPAA73b revealed the unexpected loss of one additional
base pair upstream of the four that were planned. Plasmid
pOMPAA74-103 was constructed by oligonucleotide-directed
deletion of a 30-bp pOMPA -I- 4 fragment that encodes hp2 and
the preceding nucleotide; this mutation created a Sna^l at the
site of deletion. Plasmid pOMPAA104 was created by deleting
the 0.08-kb Bell-Snam fragm ent of pOMP AA74-103, after first
adapting the SndSi end with a Bell linker (CTGATCAG) and
cleaving the linker with
Bell.
Insertion of the 0.06-kb
Bell
frag-
ment of pOMPA -1-4-1-3 in its natural orientation into the
Bell
site of pOMPAA104 generated pOMPAA65-104. Plasmids
pOMPAA9-52 and pOMPAA104-114 were constructed by oligo-
nucleotide-directed deletion from pOMPA-f4 of either a 44-bp
fragment encoding th e top of hpl or an 11-bp fragment encoding
the first 11 nucleotides of ss2, respectively.
Plasmid pOMPA-l- 16a was created by cloning a 1.28-kb AccI
(filled-in)-PstI ompA fragment of pTUlOO (Bremer et al. 1980)
between the Bell (filled-in) an d Pstl sites of pJB322 (Belasco et al.
1986),
thereby reconstituting the BeR. s ite. Plasmid pOMPA-h 12a
was constructed by cleaving pOMPA-l- 16a with
Bell,
removing
the protruding ends with mung bean nuclease, and religating
the resulting blunt ends. Plasmid pOMPA-l-21b was con-
structed by oligonucleotide-directed insertion of 17 bp (GATC-
TATACTATAACCG) at a site immediately promoter-distal to
th e
Bell
recognition sequence of pOMPA-l-4, thereby creating a
Bglll
site adjacent to the
Bell
site (TGATCAGATCTATAC-
TATAACCG). Plasmid pOMPA-hl6b is nearly identical to
pOMPA-l-21b but lacks the 5-bp Bell-BglU fragment. Insertion
of a symmetrical DNA linker (GCCCACCGGCAGCTGCCG-
GTGGGC) into the Bell site (filled-in) of pOMPAA64,
pOM PA+ 16a, pOMPA
21b,
pOMPAA104, and pBLA200 (Em-
ory and Belasco 1990) created pOMPAA64», pOMPA-l-12a*,
pOMPA-l- 16b*, pOMPAA104*, and pBLA202, respe ctively.
Plasmid pOMPAA331 was constructed by deleting a 0.33-kb
Bell (filled \n]-Hpal fragment from pOMPA-l- 16a, thereby re-
constituting the Bell site. Plasmid pPBlOl was created by in-
serting a 0.39-kb
BamHl-Pstl
fragment of pOM PA+ 16a be-
tween the BflmHI and Pstl sites of pUC19.
Measurement of mRNA lifetimes
Bacterial culture at 30°C in supplemented LB medium, RNA
extraction, SI analysis, and calculation of mRNA half-lives
were performed as described previously (Emory and Belasco
1990).
All strains grew rapidly under these conditions, with dou-
bling times between 34 and 44 min. Q uantita tion of SI-protec-
tion data was accomplished either by autoradiography and den-
sitometry (Emory and Belasco 1990) or by gel scanning on a
Molecular Dynamics model 400 Phosphorlmager. Half-lives
were calculated by least-squares analysis of semilogarithmic
plots of mRNA concentration versus time. Half-life errors were
estimated from the standard deviation of the slope of each plot.
Three different 5'-end-labeled DNA fragments of pOMPA-l- 16a
were used as probes to monitor the decay of wild-type
ompA
mRNA and most variants
thereof.
SI analysis with any of these
probes (a 1.2-kb SgiII-£coRI DNA fragment, a 0.6-kb BstEII-
£coRI fragment, or a 0.6-kb //indlll-fcoRI fragment), which
have be en described prev iously (Emory and B elasco 1990), gives
the same half-life for the wild-type
ompA
transcript because all
three probes extend beyond the 3' boundary of the compara-
tively stable 5'-terminal segment of ompA mRNA. Decay of
ompA + 16 b and ompA + 16b* mRNA was monitored by SI
analysis with a 0.6-kb Hindlll-EeoRl fragment of a
pOMPA-l-
16b*
point mutan t, which was 5'-labeled at a Hindlll
site created by oligonucleotide-directed mutagenesis at a site
corresponding to ompA codon 94. Differential decay of seg-
ments within the ompA + 16a, ompA + 16b, an d ompAA64 mes-
sages was monitored by simultan eous SI analysis with two seg-
ment-specific probes labeled at the 5' end: a 0.46-kb Avall-
Hindlll
fragment of pOMPA-l-16a complementary to the 5'
UTR plus codons 1-37, and either a 0.86-kb BgiII-£coRI frag-
ment of pOMPAA331 complementary to codons 67-295 or a
0.16-kb BgiII-£coRI fragment of pPBlOl complementary to
codons 249-295. Comparative 5'-end-mapping of ompA,
ompA
+ 16a,
and ompA
+
16b mRNA was performed by SI anal-
ysis with a 0.5-kb
BstEU-EeoW.
fragment of eithe r pOMPA-H 16a
or pOMPA-l- 16b. Decay of blaZOO and bla202 mRNA was mon-
itored by SI analysis with a 5'-labeled
1.03-kb
Hinil-Hindlll
fragment of pBLA200.
Probing of RNA seeondaiy structure w ith DM S
Methyla tion of RNA by DMS in £.
eoli
and in vitro, alkylation
of RNA with CMCT in vitro, and mapping of reactive nucle-
146
GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Presson October 10, 2011 - Published by genesdev.cshlp.orgDownloaded from
http://www.cshlpress.com/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/http://www.cshlpress.com/http://genesdev.cshlp.org/
-
8/16/2019 Emory 1992 E.coli
14/15
mRNA stabilization by a 5' stem -loop
ot ides were performed as descr ibed previously (Moazed et al .
1986;
Ch en e t a l . 1991) . Bacter i a for DMS and CM CT alkyla t ion
(SE600/pOMPA+12a, SE600/pOMPA+16a, SE600/pOMPA+16b,
or SE600/pOMPAA29) were grown and harves t ed under t he
sa m e c o n d i t i o n s a s w e r e e m p l o y e d fo r m e a su r e m e n t s of m R N A
stabi l i t y . Two 5 ' - end- l abeled DNA ol igonucleot ides (AGC-
G A A A C C A G C C A G T G C C A C T G C o r G A T A A C A C G G T T A -
A A T C C T T C A C ) w e r e u se d a s p ri m e r s t o m a p m e t h y l a t i o n s i t es
wi th in the 5 ' UTR of
ompA
m R N A v a r i a n t s . T h e se p r i m e r s
a n n e a l e d t o m R N A se q u e n c e s 2 2 - 4 5 n u c l e o t i d e s d o w n s t r e a m
or 50-72 nucleot ides upst r eam of t he ompA t r ans l a t ion in i t i a -
t i on codon, r espect ive ly , and they were extended a t 55°C wi th
AMV rever se t r ansc r ip t ase (Boehringer) .
A c k n o w l e d g m e n t s
We thank L.-H. Chen, I . Laird-Offr inga, and C. Well ington for
the i r he lpfu l com me nt s and M. Han sen for a p l asm id cons t ruc-
t i on . Thi s work was suppor t ed by U.S. Publ i c Heal th Servi ce
grant GM 35769 to J .G.B. f rom the Na t iona l Ins t i t u t es of Hea l th ,
a Nat ional Sci ence Foundat ion Graduate Fel lowship to S .A.E. ,
and funds f rom the Associa t ion pour l a Recherche sur l e Cancer
to P.B.
The publ icat ion costs of this ar t icle were defrayed in par t by
payment of page charges . Thi s ar t i c l e must t herefore be hereby
m a r k e d a d v e r t i s e m e n t i n a c c o r d a n c e w i t h 1 8 U S C se c t io n
1734 solely to indicate this fact .
R e f e r e n c e s
Api r ion , D. 1978. I so l a t ion , genet i c mapping, and s ome charac-
t er i za t ion of a muta t ion in
Escherichia coh
that af fects the
process ing of r i bonucle i c ac ids .
Genetics
9 0 :
6 5 9 - 6 7 1 .
Babi t zke , P . and S .R. Kushner . 1991. Th e Am s ( a l te r ed mR NA
stabi l it y) pro t e in and r ibonuclea se E are encoded by the sam e
st ructura l gene of
Escherichia coh. Proc. Natl. Acad. Sci.
8 8 :
1-5.
Baga, M., M. Goransson, S. Normark, and B.E. Uhl in. 1988.
Processed mRNA wi th d i f f er ent i a l s t ab i l i t y i n t he r egula t ion
of
E. coh
pi l i n gene express ion .
Cell
52: 197-206.
Bechhofer, D.H. and D. Du bna u. 1987. Induced m RN A st abi l i t y
in
Bacillus subtilis. Proc. Natl Acad. Sci.
84: 498 -502 .
Belasco , J .G. and C.F. Higgins . 1988. Mechani sms of mRNA
decay in bact er i a : A per spect ive .
Gene
7 2 :
15-23 .
Belasco, J.G., J.T. Beat ty, C.W. Adams, A. von Gabain, and S.N.
Coh en. 1985. Di f f er enti a l express ion of photos ynthe s i s
genes in R. capsulata res ul ts f rom seg me nta l di f ferences in
s t abi l i t y wi th in the polyci s t ronic rxcA t r anscr ip t . Cell
40 : 171-181 .
Belasco, J.G., G. Ni lsso n, A. von G abain , and S.N. Coh en. 1986.
The s t abi l i t y of E. coh gene t r anscr ip t s i s dependent on de-
t erminant s l oca l i zed to speci f i c mRNA segment s .
Cell
46 : 245-251 .
Bremer , E. , E. Beck, I . Hindennach, I . Sonntag, and U. Henning.
1980.
Cloned s t ruc tura l gene
[ompA]
for an in t egra l me m-
brane pro t e in of
Escherichia coh
K-12.
Mol. Gen. Genet.
179: 13-20.
Chen, C.-Y.A. and J.G. Belasco. 1990. Degradat ion of
pufLMX
m R N A i n
Rhodobacter capsulatus
i s i n i t i a t ed by non- r an-
dom endonucleoly t i c c l eavage . / .
Bacterial. 172:
4 5 7 8 - 4 5 8 6 .
Chen, C.-Y.A., J.T. Beat ty, S.N. Cohen, and J.G. Belasco. 1988.
An in t er c i s t ronic s t em- loop s t ruc ture funct ions as an
mRNA decay t erminator necessary but i nsuf f i c i ent for
puf
m R N A s t a b i l i t y .
Cell
5 2 : 6 0 9 - 6 1 9 .
Chen, L.-H., S.A. Emory, A.L. Br icker , P. Bouvet , and J.G.
Belasco . 1991. S t ruc ture and funct ion of a bact er i a l mR NA
stabi l i zer : Analys i s of t he 5 ' unt r ans l a t ed r egion of
ompA
m R N A . / .
Bacteriol.
173: 4 5 7 8 - 4 5 8 6 .
Donovan, W.P. and S .R. Kushner . 1986. Polynucleot ide phos-
phorylase and r ibonuclease I I are required for cel l viabi l i ty
a n d m R N A t u r n o v e r i n
Escherichia coli
K-12.
Proc. Natl.
Acad. Sci.
83: 120-124.
Emo ry, S.A. 1991 . An alys is of th e
Escherichia coli ompA
m R N A s t a b i l i z e r. P h . D . t h e s i s . H a r v a r d U n i v e r s i t y , C a m -
br idge , Massachuset t s .
Emory, S.A. and J.G. Belasco. 1990. The
ompA
5 ' u n t r a n s l a t e d
R N A se g m e n t f u n c ti o n s i n
Escherichia coli
as a growth- r a t e-
r egula t ed mRNA st abi l i zer whose ac t iv i ty i s unre l a t ed to
t ranslat ional ef f iciency. / . Bacteriol. 172: 4 4 7 2 - 4 4 8 1 .
Gorski , K., J.-M. Roch, P. Prentki , and H.M. Krisch. 1985. The
st abi l i t y of t he bact er iophage T4 gene 32 mR NA : A 5 ' l eader
sequence tha t can s t abi l i ze mRNA t r anscr ip t s . Cell 4 3 : 4 6 1 -
469.
Jaeger, J.A., D.H . Tur ner , an d M. Zu ke r . 1989. Imp rov ed pred ic-
t i ons of secondary s t ruc tures for RNA. Proc. Natl. Acad. Sci.
86: 7 7 0 6 - 7 7 1 0 .
Kennel l , D .E. 1986. Th e ins t abi l i t y of mess enge r RN A in bac-
ter ia. In
Maximizing gene expression
(ed. W.S. Reznikoff and
L. Gold) , pp . 101-142. But t erwor th s , S toneha m, Mass achu -
se t t s .
Kunkel , T .A. 1985. Rapid and ef fi c i ent s i t e - speci f ic mu tage nes i s
w i t h o u t p h e n o t y p i c s e l e c t i o n .
Proc. Natl. Acad. Sci.
82 :
4 8 8 - 4 9 2 .
Lin-Chao, S . and S .N. Cohe n. 1991. Th e r a t e of process ing and
degradat ion of an t i sense RNAI r egula t es t he r epl i ca t ion of
ColEl - type p l asmids in v ivo .
Cell
6 5 : 1 2 3 3 - 1 2 4 2 .
Lund berg, U., G. Ni lsson , and A. von Ga bain . 1988 . Th e differ-
ent ial stabi l i ty of the
Escherichia coli ompA
a n d
hla
m R N A
at var ious growth rates is not correlated to the eff iciency of
t r a n s l a t i o n . Gene 7 2: 141-14 9.
Lundberg, U., A. von Gabain, and O. Melefors. 1990. Cleavages
in the 5' region of the
ompA
an d
bla
m R N A c o n t r o l s t a b i l -
i t y : S tudies wi th an £ .
coli
m u t a n t a lt e r i n g m R N A s t a b i l it y
and a novel endor ibonuclease .
EMBO J.
9 : 2 7 3 1 - 2 7 4 1 .
Mack ie , G. 1991. Speci fi c endon ucleo ly t i c c l eavage of t he
mR NA for ri bosom al pro t e in S20 of
Escherichia coli
r equi r es
the product of t he
am s
gene in vivo and in vi t ro. / .
Bacteriol.
173: 2 4 8 8 - 2 4 9 7 .
Mayford, M. and B. We i sblum . 1989. Con forma t ional a l t e r -
a t ions in t he
eimC
t r anscr ip t s i n v ivo dur ing ind uct io n .
EMBO J. 8: 4307-4314.
McP heeter s , D.S . , G .D. Stormo, and L . Gold . 1988. Au toge nou s
regula tory s i t e on the bact er iophage T4 gene 32 messenger
RNA. / . Mol. Biol. 2 0 1 : 5 1 7 - 5 3 5 .
Melefor s , O. and A. von Gabain . 1988. S i t e- speci f i c endonucle-
o ly t i c c l eavages and the r egula t ion of s t ab i l i t y of E. coli
ompA m R N A . Cell 5 2 : 8 9 3 - 9 0 1 .
. 1991. Gen et i c s tudies of c l eavage- in i t i a t ed m RN A de-
cay and process ing of r i bosom al 9S RN A sho w tha t t he Esch-
erichia coli ams
and
rne
loc i a r e t he same.
Mol. Microbiol.
5:
8 5 7 - 8 6 4 .
Moazed, D. , S . S t ern , and H.F. NoUer . 1986. Rapid chemical
probing of conform at ion in 16S r ibosom al RN A and 30S r i-
b o so m a l su b u n i t s u s i n g p r i m e r e x t e n s i o n ,
f. Mol. Biol.
1 8 7 : 3 9 9 - 4 1 6 .
Mott , J.E. , J.L. Gal loway, and T. Piat t . 1985. Maturat ion of
Esch-
erichia coh
t ryptophan operon mRNA: Evidence for 3 ' exo-
nucleoly t i c process ing af t er r ho-dependent t e rminat ion .
EMBO J. 4 : 1 8 8 7 - 1 8 9 1 .
M ow a, N.R. , K. Na kam ura , and M. Inouye. 1980. Regula tory
GENES & DEVELOPMENT 147
Cold Spring Harbor Laboratory Presson October 10, 2011 - Published by genesdev.cshlp.orgDownloaded from
http://www.cshlpress.com/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/http://www.cshlpress.com/http://genesdev.cshlp.org/
-
8/16/2019 Emory 1992 E.coli
15/15
Emory et al.
r eg ion of t he gene for t he omp A prote in , a major outer m em -
brane pro t e in of Escherichia coh. Pioc. Natl. Acad. Sci.
77 :
3 8 4 5 - 3 8 4 9 .
Mudd, E.A., A.J. Carpousis, and H.M. Krisch. 1990a. Escherich-
ia coh
RNa se E has a ro l e i n t he decay of bact er iophage T4
m R N A .
Genes Sk Dev.
4 : 8 7 3 - 8 8 1 .
Mudd, E.A., H.M. Krisch, and C.F. Higgins. 1990b. RNase E, an
endor ibon uclease , h as a genera l r o l e i n t he chem ical decay of
E. coh
m R N A : E v i d e n c e t h a t
me
a n d
am s
are t he same ge-
net i c l ocus . Mol. Microbiol. 4 : 2 1 2 7 - 2 1 3 5 .
Na kam aye , K.L. and F . Eckst e in . 1986. Inhib i t i on of r es t r i c t i on
e n d o n u c l e a se Neil c l eavage by pho spho roth ioa t e groups and
i t s appl i ca t ion to o l igonucleot ide-di r ec t ed mutagenesi s . Nu -
cleic Acids Res. 1 4: 9679 -9697 .
Newbury, S.F. , N.H. Smith, E.G. Robinson, I .D. Hi les, and C.F.
Higgins . 1987. S t abi l i za t ion of t r ans l a t ional ly ac t ive mRNA
by prokaryot i c REP sequences . Cell 4 8 : 2 9 7 - 3 1 0 .
Nils son , P. and B.E. Uh l in . 19 91. Different ial decay of a poly-
c i s t ronic Escherichia coh t r anscr ip t i s i n i t i a t ed by RNase
E-dependent endonucleoly t i c process ing . Mol. Microbiol.
5 : 1 7 9 1 - 1 7 9 9 .
Nilsson, G., J.G. Belasco, S.N. Gohen. and A. von Gabain. 1984.
Growth- r a t e dependent r egula t ion of mRNA st abi l i t y i n
Escherichia coh. Nature
3 1 2 :
7 5 - 7 7 .
. 1987. Ef fec t of pr ema ture t e rm inat ion of t r ans l a t ion on
mRNA st abi l i t y depends on the s i t e of r i bosome r e l ease .
Proc. Natl.
Acad.
Sci.
8 4 : 4 8 9 0 - 4 8 9 4 .
O n o ,
M. and M. Kuwan o. 1979. A condi t i onal l e tha l mu ta t io n
in an
Escherichia coh
s t r a in wi th a l onger chem ical l i f e t ime
o f m R N A . f. Mol. Biol. 129: 3 4 3 - 3 5 7 .
Pedersen, S. , S. Reeh, and J.D. Fr iesen. 1978. Funct ional mRNA
half- l ives in
E. coh. Mo l. G en. Genet.
166: 3 2 9 - 3 3 6 .
Sandler , P . and B. We i sblum . 1988. Erythrom ycin- ind uced s t a-
bi l izat ion of ermA m e sse n g e r R N A i n Staphylococcus au-
reus and Bacillus subtilis. J. Mol. Biol. 2 0 3 : 9 0 5 - 9 1 5 .
Tarasevic i en e , L ., A . Miczak , and D . Api r ion . 1991. The gene
specifying RNase E
[rne]
and a gen e affect ing m RN A s tabi l -
i ty [ams] are t he same gene . Mol. Microbiol. 5: 851 -855.
Varshavsky, A., A. Bachmair , D. Finley, D. Gonda, and I . Wun-
ning . 1988. The N-e nd ru l e of se l ec t ive pro t e in turno ver . I n
Ubiquitin ( ed . M. Rechst e iner ) , pp . 287 -324 . P l en um Press ,
New York .
von Gabain, A., J.G. Belasco, J.L. Schot tel , A.C.Y. Chang, and
S.N. Cohen. 19 83. Deca y of mR NA in
Escherichia coh:
In -
vest igat ion of the fate of specif ic segments of t ranscr ipts.
Proc. Natl. Acad. Sci.
8 0 : 6 5 3 - 6 5 7 .
Yam amo to , T . and F . Ima mo to . 1975. Di f f erent i a l s t ab i l i t y of
trp messenger RNA synthes i zed or ig ina t ing a t t he trp pro-
mo ter and PL prom oter of l ambd a trp phage. / . Mol. Biol.
92 : 2 8 9 - 3 0 9 .
Youvan, D.C., E.J. Byl ina, M. Alber t i , H. Begusch, and J.E.
Hear s t . 1984. Nu cleo t ide and deduced polypept ide se-
quences of t he photosynthet i c r eac t ion-center , B870 an-
t enna , and f l anking polypept ides f rom
R. capsulata. Cell
37 : 9 4 9 - 9 5 7 .
Cold Spring Harbor Laboratory Presson October 10, 2011 - Published by genesdev.cshlp.orgDownloaded from
http://www.cshlpress.com/http://www.cshlpress.com/http://genesdev.cshlp.org/http://genesdev.cshlp.org/http://www.cshlpress.com/http://genesdev.cshlp.org/