The relation between translation and mRNA degradation in the lacZ gene

Biochimie ( 199 i ) 73, 1533-1541 © Socirt6 francaise de biochimie et biologie molrculaire / Elsevier, Paris

i533

T h e relation b e t w e e n translation and m K N A degradation in the lacZ gene O Yarchuk, I Iost, M Dreyfus* Laboratoire de Gdndtique mol~culaire de I' Ecole Normale Sup~rieure (CNRS D 1302). 46, rue d' UIm. 75230 Paris Cedex 05, France

(Received 12 September 1991; accepted 1 October 1991 )

Summary - - The technique of gene fusion, in which the gene of interest, severed from its 3' end, is in-phase fused to a reporter gene - usually l a c Z - is widely used to study translational regulation in Escherichia coil Implicit in these approaches is the assumption that the activity of the ribosome binding site (RBS) fused in-phase with lacZ, does not per se modify the steady-state level of the lacZ mRNA. Herein, we have tested this hypothesis, using a model system in which the RBS of the lamB gene is fused to lacZ. Several point mutations affecting translation initiation have been formerly characterized in this RBS, and we used Northern blots to study their effect upon the lacZ mRNA pattern. Two series of constructs were assayed: in the first one, a 51-bp fragment centered around the lamB initiator codon, was inserted in front of lacZ within the natural lactose operon, whereas in the second the lacZ gene was fused to the genuine malK-lamB operon just downstream from the lamb RBS. We observed that in the first series, the concentration and average molecular weight of the lacZ mRNA dropped sharply as the efficiency of the RBS decreased. This apparently arose from a decreased stability of the message, since the mRNA patterns are equalized when the endonuclease RNase E is inactivated. We suggest that in this case the rate limiting step in the decay process is an RNase E cleavaz~e that is outcompeted by translation. In contrast, in the second series, the lacZ mRNA pattern was hardly affected by translation, and the removal of RNase E caused a decrease, rather than an increase, in the amount of the lacZ mRNA. Possible reasons for the different behaviour of the two series are discussed. Finally, we propose a general strategy for characterizing the rate-limiting step in the decay process when translation and degradation are coupled. lacZ gene / mRNA stability / translation initiation / RNase E / gene fusion

Introduction A p o p u l a r technique for s t u d y i n g gene expression, uz.u

lJalLIL,UlC~tl~ U a l l a l t t t l U l l a l iC~UlClLIUII~ C O n s i s t s i n

c o n s t r u c t i n g fusions to the E s c h e r i c h i a coli l a c Z gene [I]. In this approach, the gene o f interest, p r e c e e d e d by its o w n p r o m o t e r or b y an h e t e r o l o g o u s one, is severed f r o m its 3' end and fused in phase with lacZ, f r o m w h i c h the natur~.l transcriptional and translational signals have been r e m o v e d . P r o v i d e d the j u n c t i o n is located upstream f r o m the ca 25th l a c Z c o d o n , the hybrid protein is r e a s o n a b l y stable and retains the s a m e specific activity as the g e n u i n e e n z y m e [2]. This has been c o n v e n i e n t l y exploited to c o m p a r e the strength o f different translation initiation *Correspondence and reprints |PTG: lsopropyl-~-D-thiogalactopyranoside. ONPG: 2-Nitrophenyl-l~-o-thiogalactopyranoside. RBS: ribosome binding site. The RBS is the mRNA region protected against RNases by the bound ribosome. It may not coincide with the translation initiation region, which contains all informations required in cis for translation initiation [8]. Herein, as this distinction is inessential, we use RBS as a unique denomination for both concepts.

Abbreviations:

signals [3], to study specific translational regulation [4], to a n a l y z e translational c o u p l i n g between the successive cistrons o f a polycistronic m R N A [5, 6], and to study the influence o f g r o w t h conditions upon translation initiation [7]. In these strategies it ;,q implicitely a s s u m e d that the expression o f the reporter lacZ gene reflects faithfully the activity o f the fused r i b o s o m e b i n d i n g site (RBS), and, in particular, that the a b u n d a n c e o f the lacZ m R N A is not affected by the nature o r activity o f the RBS. Attempts to verify this point have yielded c o n f u s i n g results. In s o m e constructs and/or genetic b a c k g r o u n d s , the lacZ m R N A c o n c e n t r a t i o n appeared indeed insensitive to the nature o f the R B S (eg [5, 9]), whereas in others it was quite dependent o f it [6, 10, 11]. In the lager studies, variations c o n c e r n e d m a i n l y the probability o f premature transcription termination (polarity; see [ 12]) and not m R N A stability, even t h o u g h a complete block in translation initiation labilizes the iacZ m e s s a g e [13]. S o m e R B S replacements did alter the lacZ m R N A stability, but these c h a n g e s were ur,correlated with translation [ 14]. In this Communication, we re-examine the relationship b e t w e e n translation and m R N A a b u n d a n c e in

1534

0 Y,'u-chuk et al

l a c Z t'rotein fusion g e n e s . To tlais end, w e h a v e c h o s e n a p a r t i c u l a r R B S in w h i c h p o i n t m u t a t i o n s that affect t r a n s l a t i o o h a v e b e e n well c h a r a c t e r i z e d p r e v i o u s l y . T h e w i l d - t y p e o r m u t a t e d R B S s w e r e then f u s e d u p s t r e a m o f lacZ, a n d w e h a v e r e c o r d e d the effect o f the m u t a t i o n s u p o n the l a c Z m R N A a b u n d a n c e , a n d e v e n t u a l l y u p o n the p o l a r i t y a n d m R N A stability. O d d l y , w e o b s e r v e d ~hat the effect o f the m u t a t i o n s is v e r y d e p e n d e n t u p o n the e x a c t s e q u e n c e o f the f u s i o a u p s t r e a m o r d o w n s t r e a m f r o m the R 3 S . M o r e o v e r , the stability o f the l a c Z m R N A in c o n s t r u c t s that r e s e m b l e m o s t c l o s e l y the g e n u i n e l a c Z g e n e , is v e r y s e n s i t i v e to translation. W e d i s c u s s the r e a s o n for these c o n s t r u c t - t o - c o n s t r u c t d i f f e r e n c e s , a n d p r o p o s e a strategy to s t u d y m R N A d e g r a d a t i o n w i t h o u t the i n t e r f e r e n c e o f t r a n s l a t i o n .

Materials and methods Strains and plasmids POP strains The derivatives of the strain MC4100 (or pop3) which harbor either the wild-type lamB RBS or its mutated 701 or 708 alleles in front ef a la::.lB'-'lacZ hybrid gene, have been described [3] and were obtained from Dr O Raibaud. The DNA region encompassiog these fusion genes is depicted in figure 1A. For simplicily, these strains are herein refe~ed to as 'POP' strains. ENS strains The general technique for the replacement of the lacZ RBS by a foreign ore onto the chromosome of HfrG6 [16] has been described [15]. Briefly, we have constructed a plasmid, pEMBLA45, in which the RBS of the 13-galactosidase ~peptide gene has been replaced by a polylinker, allowing RBSs from other genes to be cloned in phase with the ct-peptide sequence. The resultant pla~mids are then introduced into ENS0 (formerly HfrGrAlacl2; [15]), a derivative of HfrG6 obtained by substituting the sequence -13 to + 58 of the lacZ gene (+1 being the transcriptional start) by an unrelated, 17 bplong, synthetic sequence. This substitution removes part of the lac promoter, the lac operator sequence, and the lacZ RBS, and therefore knocks down the expression of the lacZ gene. Lac+Amp~ clones resulting from transfer of the RBS from the plasmid onto the chromosome by homologous recombination, followed by plasmid segregation, are then selected. We have already described the preparation of three 5 I-bp long fragments which encompass the RBS of lamB and its mutated 761 and 708 alleles [17, 18]. These have now been transferred onto the chromosome of HfrG6 (fig IB). IMO strain This strain is very similar to the previously described MOO [ 171, except for the replacement of the sequence +3 to +28 of the gene transcribed by T7 RNA polymerase by the synthetic sequence 5'-AA'Iq'GTGAGCGGATAACAA'IqTG_3' that encompasses the lac operator. A detailed description of this strain will appe&- elsewhere (I lost and M Dreyfus, in preparation). Briefly, I'd0 is a derivative of BL21(DE3) [19~, an E coli B strain harboring the T7 gene 1 coding for RNA polymerase, under the control of the lacUV5 promoter. IM0 harbors a TnlO insertion within the endogenous lacZ gene, and carries an

extra single-copy lacZ gene under the control of the T7 gene I0 promoter and terminator. "lhis gene lacks any RBS and cannot therefore be translated (fig IC). The two conditional rne mutations, ams and rne-3071, were introduced into the POP, ENS, or IM0 strains by PI transduction, using lysates grown on CHI826 (rne-3071) or CHI828 (ams) ([20]; courtesy of Dr P Regnier), and selecting for tetracycline resistance and thermosensifive growth. Prior to the transduction, iM9 was cured of TnlO [21 ]. Bacterial culno'es Cells were grown in MOPS medium, supplemented with glycerol (0.2% w/v) and all amino acids, nucleic acid bases and vitamins [22]. For the preparation of RNA from cells carrying the rne-3071(ts) or ams(ts) mutations at the non-permissive temperature, cultures were grown at 28°C until they reached 0.2 to 0.50D60 o units. They were then shifted to 42--44°C for 30 rain (the non-permissive temperature varied with the genetic context; see legends of figures), before RNA extraction. The ENS and IM0 strains were continuously induced with IPTG (0.2 and 0.7 mM, respectively). Concerning the POP strains, the malK-lamB promoter was induced by adding maltose (0.2% final), together with glycerol. ~-Galactosidase assay and RNA analysis Harvested cells were disrupted by pulsed sonication and 13galactosidase was assayed in the cleared iysate as before [ 17]. For RNA analysis, cells were rapidly chilled, centrifuged and total RNA was extracted with hot phenol [17]. Alternatively, the chill-centrifugation step was omitted, ie the culture was immediately mixed with an equal volume of hot phenol, which allowed instant lysis of the cells. Equal amounts (2-5 lag) of RNA from the strains under comparison were then electrophoresed and blotted on nylon membranes (Amersham H~cbond N) [17]. To visualize lacZ and lacA specific mRNAs, the membranes carrying the immobilized total RNAs were incubated under stringent conditions with 3ZP-labelled restriction fragments encompassing most of the lacZ and lacY sequences (Genebank Ecolac nt 1305 to 5253), or the end of the lacY gene ......... : t~e ,u,~ sequence (Genebartk Ecoiac nt 5254 to 6499), respectively. These fragments were labelled to a specific activity of 109 dpm/lag, using the multiprime labelling kit of BRL. The radioactivity retained on the membranes was quantified using the Phosphor imager system (Molecular Dynamics). Alternatively, the membranes were autoradiographed without amplifying screen and the films were scanned. The two methods usually agreed to within 20%.

Results The m a l K - i a m B ' - ' l a c Z f u s i o n s Hall et a i [3] h a v e c o n s t r u c t e d a h y b r i d m a l K - l a m B ' "lacZ-lacY-lacA o p e r o n in w h i c h the l a c Z g e n e (starting f r o m t h e 20th c o d o n ) r e p l a c e s the l a m b s e q u e n c e a f t e r the 39th c o d o n (fig I A ) . T h e h y b r i d o p e r o n is l o c a t e d in the m a l B r e g i o n o f the c h r o m o s o m e , in place of the genuine malK-iamB operon. Meanwhile, t h e s e a u t h o r s c h a r a c t e r i z e d t w o m u t a t i o n s (701 and 708) w h i c h i n h i b i t e d the t r a n s l a t i o n o f l a m B , the phage lambda receptor structural gene, by creating m R N A s e c o n d a r y structures s e q u e s t r a t i n g the S h i n e D a l g a r n o e l e m e n t [3]. W h e n t r a n s f e r r e d into t h e

Translation affects the lacZ mRNA stability malKp ~

A

IE~o"K,21

malK

""-

lamB" .

.

.

"lacZ

39 th

RBS lacZ

C

50

•lacZ

AU t-

lacA

l

i

20 t l l co(ton

codon

iE coil KI2

lacy

.i

.

1535

,e

lacy

lacA

I

t

,,, J

-

AATf0,rGaO,JGGG~'~.~.~,~,~ AAAAAUG,\CUCAGGA~AUAGA AU6 AUG AUU ACU CUG CGC AAA CUU £TGCASCr._AC.GCTT~-C-~CTG Ir W 1 BamHI VG •A PSl; s t~ coco~ of lacZ

701

708

. . . .

mm

I

I

I

p

I lacl

,, i

I I

tacZ I

.........

7,.

i

Tn lO

T7 RNA pO1 I

|

,,p~"

|

,~'~

RBS

FT7 I=~c~Y

'--"'"

T

lacOq5 ~ [~ _ _

I

ter ,1~.7

laCY" T7

....

,

--I

ma-!O

,,"7,alp

I

IJ

Fig 1. Schematic representation of the chromosomal regions encompassing the different hybrid lacZ genes studied in this work. A. Structure of the hybrid malK-lamB'-'.lacZ-lacY-lacA operGn from the POP strains. The region encompassing the lamb RBS is hatched, whereas the rest of the operon is boxed. The junction between the lamB" and the "lacZ sequences is indicated. B. Structure of the lactose operon in the ENS-45 strains. The enlarged region encompasses the 5' end of the transcribed sequence, including the 51-bp fragment which harbours the lamb RBS. The 701 and 708 mutations are indicated by arrows. The genuine lac sequence is italicized. C. Structure of the chromosome of IM0, aa E coli B strain carrying a single-copy lac cassette transcribed by the T7 RNA polymerase, in the malt~, region of the chromosome (see also [17]). The symbols + and - stand for positive and negative control by ths T7 RNA polymerase and lac repressor, respectively. The endogenous lacZ gene is inactivated by an inserted transposon (vertical hatches). PT7 comprise the sequence of the T7 gene 10 promoter from positio,s - 21 to + 2 with respect to the transcription startpoint. Immediately downstream from this promoter is inserted the lacZ sequence starting from the first nucleotide of the natural transcript, and carrying the same RBS deletion as plasmid pEMBLAa6 [15]. The lac repressor binding site is noted lacO. Closed boxes are genetic element from the T7 phage. 'ter T7' is the major T7 RNA polymerase terminator from bacteriophage T7 (see [ 17] for details). All other symbols are as in B. hybrid operon, these mutations affected the expression o f the lacZ gene in m u c h the same way as they did for the natural lamB gene, at least when cells were g r o w n at low temperature ([3]; see table II). However, when the cells were grown at a higher temperature (43°C), dramatic changes occurred in the expression pattern. First, globally, the I]-galactosidase activities were

much lower at 43°C, compared to 28°C (table II), due possibly to the natural thermosensitivity of the maltose regulon activator protein, the product o f the m a l T gene (O Raibaud, personal communication;. Second, rising the temperature affected differentl3 the activities o f the strains (table II). Elsewhere, we show that this behaviour is not specific to this system: tile

1536

O Yarchuk et al

relative expressions from genes that differ in the sequence of their RBSs, often change markedly following modifications ;,n the growth conditions [ 18]. We next visualized the lacZ mRNA pattern from these strains on NoOhern blots, using a probe encompassing most of the lacZ coding sequence. Since we wanted to study the effect o f conditional rne mutations upon this pattern, cells were shifted to 43°C, a non-permissive temperature for these mutations, 30 rain before harvest. The lacZ m R N A pattern showed several bands, presumably reflecting multiple processing events and/or the existence o f several promoters upstream of the iacZ gene. The reason for this complexity has not been worked out. However, the pattern remained qualitatively the same whatever the RBS, although quantitatively it varied in intensity in oarallel with the ~-galactosidase expressions at 43°C (t~g 2A). The degradation of many individual E coli m R N A s is controlled, directly or not, by the endonuclease RNase E [20, 23, 28]. The concentration o f the hybrid m a l K - l a m B " l a c mRNA, was therefore assayed in the presence o f the allelic ares or rne-3071 mutations, both o f which result in a thermosensitive RNase E activity. As seen in figure 2A, the removal o f RNase E produced no increase in the amount of message: rather, a sharp decrease was noted. The mech-

anism :.,nderlying this decrease is tentatively discussed below. In any case. the rate-limiting step in the decay of this message is unlikely to involve RNase E. The ENS strains

Elsewhere we have described the preparation o f small (51 bp) D N A fragments encompassing the RBS o f lamB and its mutated 701 and 708 alleles [17, 18]. These RBSs have now been inserted within the genuine lactose operon in place o f the ribosome binding site o f the lacZ gene [ 15]. The [3-galactosidase activity o f the strain carrying the wild-type lamb RBS was 0.3- to 0.5-fold that o f the strain carrying the unmodified operon, depending upon the temperature (table II). The effects o f the 701 and 708 mutations upon the ~-galactosidase expression were similar in this new context (hereafter referred to as the ENS strains) as in the POP strains at 28°C (table II). Concerning now the lacZ m R N A pattern, the hybrid gene carrying the wild-type l a m b RBS behaved like the natural lacZ gene, which has been recently studied in this respect [24]. Thus, three discrete bands (ca 5.2, 4.5 and 3.2 kb in length) were obseTved, with the 3.2kb band being by far the most abundant. These

Table I. Bacterial strains used in this work.

Nomenclature

Genotype

Reference or origin

MC4100

AlacU169 araD139 rpsL relA thiA flbB

[1 ]

pop3857

MC4100 (~(lamB-lacZ)52-4)

[31

pop3859

MC4100 lamB 701 (~,~(lamB-lacZ )52-4 lamB 701)

[31

pop3860

MC4100 lamB 708 (~O(lamB-lacZ)52-41amB 708)

[3[

ENS0

HfrG6 Alacl2

[ 15]

ENS3-45

HfrG6 lacZ::RBS tamB

this work a

ENS4-45

HfrG6 lacZ::RBS t,onaTol

this work a

ENS5-45

HfrG6 IacZ::RBS tamBTos

this work a

CH 1826

MC 1061, rne-3071 (ts), -ce-726.':Tn 10

[20]

CH 1828

MC 1061, ams(ts), zce-726.':TnlO

[20]

BL21(DE3)

F- hsdS gal (~dmm21 Anin5 int.":T7 genel)

[19]

IM0

BL2 I(DE3) lacZ::TnlO malPpA534::lacZA46

this work b

aThe exact sequence of the lamB RBS-carrying fragments and of the resulting lamB'-'lacZ fusions, are shown in figure lB. The strains am designated 45 after the plasmid pEMBLA45 [ 15], which has been used in their construction, bThe sequence inserted nomenclat' malP_ ureint°is used.illustratedin figure IC. This insertion harbours the same lacZ RBS deletion as pEMBLA46 [15], hence the

Translation affects the lacZ mRNA stability

tO 04

~0

I

I

1:537

L~ U ~ ~"

I'~" ~ L~

F 3 L~

~

nm

;

~

~i¸¸

~i i ~

~¸~¸¸¸

i ~iTi¸¸¸

uJ~ ¢.,.=..

I

0'3 -~C'q

~4

~ ~ z ~. ~ "~.~ 03*"

~

0

o~ puJ~ ! #Ud

~,.~ ~.~.~ ~

m

!

~ ' ~

i~ ¸

~ . = ~, ~i~. .~ L_~ ~

I ¢q

!

i 538

O Yarchuk et al

species correspond to the full length (ZYA) operon transcript, and to species processed between l a c Y and lacA and l a c Z and lacY, respectively 124, 25]. As concerns the strains harbouring the 708 and, even more so, the 701 RBSs, both the abundance and the average molecular weight o f the l a c Z m R N A decreased dramatically compared to the wild-type. Quantification revealed that a large part of the strainto-strain variations in l]-galactosidase activity were paralleled by variations in the amount o f m R N A (compare tables II and III). The strength o f the promoter is unlikely to change from one gene to the next, as they are identical in sequence down to at least the 39th transcribed nucleotide. However, the strains could still differ by their degree o f polarity and/or their l a c Z mRNA stability. We assayed polarity by reprobing the Northern blots with a DNA fragment corresponding to the end o f l a c y and lacA, the distal genes o f the operon (fig 2C). Aside from the ZYA transcript, this probe revealed a complex set o f poorly resolved bands, in agreement with previous reports [24, 25]. Going from the wildtype to the mutant RBSs resulted in the disappearance o f the ZYA transcript, as anticipated from figure 2B, but the total amount o f material that hybridized with the probe remained constant (table III). We conclude that the change in the l a c Z mRNA pattern seen on figure 2B, is not due to variations in polarity, and

therefore must stem from variations in mRNA stability. This was confirmed by the following experiment. We introduced the ares mutation into the ENS strains, and analyzed again the m R N A pattern at the non-permissive temperature. The result was strickingly different from what was seen with the POP strains: the removal o f RNase E caused an increase, rather than a decrease, in the amount of l a c Z mRNA, and it equalized its average molecular weight whatever the strain (fig 2B). The amount of l a c Z transcript was still lower in strains carrying the 701 and 708 RBSs compared to wild-type, but this was due to strong polarity effects induced by the a m s mutation. Indeed, the amount o f lacA m R N A was now much smaller in the strain carrying the 701 and 708 RBSs, compared to the wild-type RBS (table III). After correction for these polarity effects, the content o f l a c Z m R N A is nearly the same whatever the RBS (see the l a c Z / l a c A mRNA ratios in table III). This suggests that, in the rne- context, the stability of the l a c Z m R N A is the same for all RBSs. Figure 2C reveals yet another interesting phenomenon: in the presence o f the ares mutation, most o f the lacA m R N A accumulates as the full-length 5.2-kb molecule, with very little processing to smaller species. This suggests that somehow this processing is RNase E dependent. Further work is required to determine whether the cleavages between l a c Y and lacA,

Table II. The ~-galactosidase activities were averaged from two to four independent assays and are expressed in nmol ONPG hydrolyzed per rain and per mg of protein. For comparison, the strain HfrG6, which carries the wild type lac operon, produces

4100 and 3300 units of ~-galactosidase at 28 and 43°C, respectively.

RBS 708

WT

701

28°C

43oc

28°C

43oc

28°C

43°C

ENS-45 strains

1170±180

1750 ± 55

213 ± 17

302 ± 25

96 ± 11

81 ± 4

POP strains

4200±600

70 ± 2

620 ± 10

160 ± 10

73 ± 7

11.4 ± 0.2

Table III. Quantification of the lacZ and lacA mRNAs from the Northern blots shown in figure 2B and 2C. RBS 708

WT rne +

701

ares

rne +

ams

me+

ams

lacZ mRNA a.b

100 + l0

167 + 4

50 + 5

95 + 9

28 + 3

45 + I

lacA mRNA b

100 ___7

82 + 2

106 + 13

28 + 6

98 + 7

19 + 1

lacZ/lacA

1.0 ± 0.15

2.0 + 0.1

0.5 + 0.1

3.4 + 1.0

0.28 + 0.05

2.4 + 0.2

aFor this quantification, a probe corresponding to the 3' end of lacZ (Genebank Ecolac 3948 to 4310) was used. For each RBS, the quantification from the lanes wt and c were averaged, bAveraged measurements performea on the same membrane by either autoradiogram scanning, or direct radioactivity counting (see Materials and methods). On each blot, the signal observed with the strain carrying the wild-type rne locus and lamB RBS, was arbitrarily given a value of 100.

Translation affects the lacZ mRNA stability and between lacZ and lao'Y, are directly controlled by RNase E, or whether the dependence is indirect, these cleavages requiring the prior RNase E cleavage of the lacZ mRNA. In summary, our observations fit the following model: in the ENS strains, the stability of the message increases together with the efficiency of translation, due to the fact that the RNase E cleavage which is rate-limiting in the degradation, can be outcompeted by translation. Obviously, this mechanism does not apply to the lacZ constructs harbored by the POP strains, despite their close relatedness.

1539

Wt U

ams

~

~

b

u

;.,

- ~ 4 . 2 kb - ~ 3.2 kb

Decay o f mRNA from untranslated fusions ";?!

We have recently described constructs in which the lacZ gene is fused downstream from the T7 gene 10 promoter and leader sequence [17]. These constructs were introduced in a strain that produced T7 RNA polymerase, and we showed that even those genes that were not translated at all could be transcribed, due to the insensitivity of T7 RNA polymerase to polarity [ 17]. We have modified these constructs by replacing the T7 gene 10 leader by the lac operon leader (fig 1C); hence, the 5' end of the message is now very similar to that of the natural lac mRNA. We then tested whether, when devoid of any RBS, this mRNA would still be stabilized by the removal of RNase E. Indeed as shown in figure 3, both the amount and average molecular weight of the message increased dramatically upon RNase E removal (IM0 strain; the full-length transcript is 4.2 kb long in this construct). Therefore, the degradation of the untranslated lacZ mRNA, like the lamB'-'lacZ mRNA from the ENS strains, is RNase E-dependent. Discussion

The coupling o f translation and degradation in the lacZ gene

Using constructs in which the translation of iacZ is driven by RBSs of variable efficiencies (ENS strains; fig 1B), we have obtained circumstantial evidence that the chemical stability of the lacZ mRNA is dependent upon its translation. Our argument is the following: the abundance of the mRNA drops as translation goes down, but it can be equalized if a particular endonuclease - RNase E - is removed from the cell. Elsewhere, we directly show that, in these strains, the chemical and functional lifetimes of the mRNA decrease together with the RBS efficiency (O Yarchuk et al, in preparation). Oddly, this correlation is not seen with the slightly modified constructs harL-~red by the POP strains (fig 1A). Moreover, tormer studies

F/

• ;;

!~

:,2! i} .2,

Fig 3. Pattern of the lacZ mRNA in the IM0 strain carrying ~,,,,~, ~,,,..,~e2¢i.~ r l ' l ¢ ,v~.uo, or its ares ,,,,~,~. ,,,~ b~,,s were either continuously grown at 28°C, or grown at 28°C and then shifted at 43°C for 30 min before harvest, as indicated. All symbols as in figure 2. have failed to detect any relationship between translation and mRNA decay in the lacZ gene ([10, 11, 141; but see however [13]). These inconsistencies can receive a number of interpretations. First, the crucial RNase E cleavage site may simply be absent from many constructs involving lacZ fusions. Sequence comparisons between the ENS and POP strains would then suggest that this site lies either within the first 12 nucleotides of the lac message, or between the 8th and the 19th lacZ codon, or, perhaps, in the unique polylinker sequence of ENS strains (~f fig IA and IB). Alternatively, the folding of the mRNAin the formerly studied constructs (and in the POP strains) may prevent the RNase E cleavage. While we cannot exclude these possibilities, we presently favor an alternate one. So far, whenever a coupling has been seen between translation and mRNA abundance [6,

1540

O Yarchuk et al

10. I I l, this coupling arose from variations in polarity, not in mRNA stability. In contrast, in ENS strains, we do not detect any variation in polarity, but only in mRNA stability. Polarity is presumably largest for weak RBSs, wheree,s it levels off for efficient ones. Therefore, we suspect that, compared to the formerly used constructs, the hybrid lacZ genes harboured by the ENS strains are endowed with much higher translational efficiencies. This, in turn, may be the reason why the mRNA stability is now sensitive to translation. We are presently testing this idea by comparing the translational efficiencies from the same RBSs in the ENS and POP conte.xts. We stress that, in the ENS strains, the structure, chromosomal location, and expression o f the engineered lamB'-'lacZ genes, are rather similar to those of the genuine lacZ gene. Hence, we believe that the translation-stability coupling which is documented here, might hold for the unmodified lacZ gene as well. Such coupling has indeed been proposed, on different grounds, by Kennell and coworkers [261.

The role of RNase E in determining the steady-state concentration of mRNAs RNase E has been initially implicated in the 9S to 5S ribosomal RNA processing [27], and, more recently in the decay of many individual mRNAs. In particular, the conditional mutation ams [28], now known to map within the rne gene, results in a 3--4-fold bulk stabilisation o f E coli mRNAs at the non-permissive temperature. Herein we have shown that the lacZ m R N A from the ENS strains (fig IB) also behaves in this way. In striking contrast, the lacZ mRNA from the closely related POP sLrains (fig !A), is less abundant in the absence of RNase E than in its presence. This behaviour is unprecedented, except for mRNAs that require RNase E for their processing [29], but we have recently observed it for several other m R N A s such as the galETK or the malK-lamB mRNAs (O Yarchuk, unpublished observation). While this behaviour can receive several explanations, we favor the following one. The synthesis o f mRNAs represents a large part o f the total E coli transcription whatever the growth conditions [30], and we postulate that their rapid turnover is necessary to replenish the precursor pools. Therefore, the bulk stabilization o f mRNAs is likely to cause a shrinkage in these pools, thereby globally reducing transcription. Obviously, then, the concentration of those mRNAg that are not stabilized by the RNase E removal, will decrease. It is also possible that polarity contributes to this decrease. More mRNAs are available to the translational machinery in the absence of RNase E than in its presence, and this greater competition is likely to induce polarity effects, particularly for those mRNAs

that bear weak RBSs. Figure 2C and table III show that indeed the removal of RNase E induces large polarity effects in poorly translated genes.

Uncoupling of translation and degradation How to locate the crucial RNase E cleavage site on the m R N A sequence from ENS strains? A seemingly straightforward strategy is to search for c/s-acting mutations that increase the m R N A stability. However, since translation and degradation are coupled, this approach is ambiguous: such mutations might not only hit the rate-limiting RNase E cleavage, but also affect the efficiency o f translation initiation. Indeed, we noted above that the different behaviour of the ENS and POP constructs can receive two interpretations: either the crucial site is absent from the POP mRNA, or the translational efficiencies are very different in the two series. This ambiguity is encountered whenever stability is dependent upon translation (cf [31 ]). A way to bypass it is to focus on derivatives in which the gene o f interest lacks any RBS, and therefore is completely untranslated. In this case, mutations cannot anymore affect m R N A stability via translation. However such mRNAs are usually not synthesized to their end due to polarity effects, unless the RNA polymerase is itself polarity-insensitive. This is illustrated here with the IM0 strain, in which the lacZ mRNA, though untranslated, is synthesized by T7 RNA polymerase. As expected, the stability o f this mRNA is very sensitive to the removal o f RNase E, even more so that the weakly translated ENS mRNAs (compare fig 2B and fig 3). We are now modifying systematically the sequence o f the lacZ m R N A from the IM0 strain

Acknowledgments Oleg Yarchuk is supported by a long-term fellowship from the French Minist~re de la Recherche et de la Technologie. This work has been funded by the Centre National de la Recherche Scientifique (CNRS), by the Minist~re de I'Education Nationale (DRED), and by grants from the Ligue Franqaise contre le Cancer and the Fondation pour la Recherche M6dicale to M Dreyfus. We thank Drs O Raibaud, J Guillerez and N Jacques for their critical review of the manuscript, Dr P Roux for his help in operating the phosphor imager system, and Drs Raibaud and Regnier for gifts of strains.

References 1 Silhavy TJ, Berman ML, Enquist LW (1984) Experiments with Gene Fusions. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, USA 2 Fowler AV, Zabin I (1983) Purification, structure and properties of hybrid ~-galactosidases. J Biol Chem 258, 14354--14358

Translation affects the la~Z mRNA stability 3 4 5

6

7

8 9

10 11

12 13 14

15 16

17

Hall MN, Gabay J, Ddbarbouill6 M, Schwartz M (1982) A role for mRNA secondary structure in the control of translation initiation. Nature 295,616--618 Lesage P, Truong HN, Graffe M, Dondon J, Springer M (1990) Translated translational operator in E coll. J Mol Biol 213,465-475 Hellmuth K, Rex G, Surin B, Zinck R, McCarthy JEG (1991 ) Translational coupling varies in efficiency between different pairs of genes in the central region of the atp operon of E coli. Mol Microbiol 5, 813-824 Lindahl L, Archer RH, McCormick JR, Freedman LP, Zengel JM (1989) Translational coupling of the two proximal =enes in the S I0 ribosomal protein operon of E coil J Bacteriol 171, 2639-2645 Carter-Muenchau P, Wolf RE (1989) Growth-rate-dependent regulation of 6-phosphogluconate deshydrogenase level mediated by an anti-Shine-Dalgarno sequence located within the E coil grid structural gene. Proc Natl AcadSci USA 86, ! 138-1142 McCarthy JEG, Gualerzi C (1990) Translational control of procaryotic gene expression. Trends Genet 6, 78-85 Aituvia S, Locker-Giladi H, Koby S, Ben-Nun O, Oppenheim AB (1987) RNase III stimulates the translation of the clll gene of bacteriophage lambda. Proc Natl Acad Sci USA 84, 6511-6515 Stanssens P, Remaut E, Fiers W (1986) Inefficient translation initiation causes premature transcription termination in the lacZ gene. Cell 44, 711-718 Petersen C (1987) The functional stability of the lacZ transcript is sensitive towards sequence alterations immediately downstream of the ribosome binding site. Mol Gen Genet 209, 179-187 Adhya S, Gottesman M (1978) Control of transcription termination. Annu Rev Biochem 47, 967-996 Schneider E, Blundell M, Kennell D (1978) Translation and mRNA decay. Mol Gen Genet 160, 121-129 Petersen C (1991) Multiple determinants of functional mRNA stability: sequence alterations at either end of the la~'Z gene affect the rate of mRNA inactivation. J Bacteriol 173, 2167-2172 Dreyfus M (1988) What constitutes the signal for the initiation of protein synthesis on E coli mRNAs? J Mol Bio1204, 79-84 Matney TS, Goldschmidt EP, Erwin NS, Scroggs RA (1964) A preliminary map of genomic sites for F attachment in E coli KI2. Biot'him Biophys Res Comm 17, 278-281 Chevrier-Miller M, Jacques N, Raibaud O, Dreyfus M (I 990) Transcription of single-copy hybrid lacZ genes by T7 RNA polymerase in E coll. Nucleic Acids Res 18, 5787-5792

18

19 20

21 22 23

24 25

26 27

28 29

30

31

i 541

Jacques N, Chevrier-Miller M. Guillerez J. Dreyfus M (1990) Culture conditions affect differently Ihe translation of individual E coli genes. In: Post-Tralt.wriptional Co~trol of Gene Expression (McCarthy JEG. Tuite MF. eds), NATO-ASI series, Springer-Verlag, Berlin, 145156 Studier W, Moffat BA (1986) Use of bacteriophage T7 RNA polymerase to direct selective high-level expres',ion of cloned genes. J Mol Biol 189, 113-130 Mudd EA, Krisch HM, Higgins CF ~1990) RNase E. an endonuclease, has a general role in the ,:heroical decay of E coil mRNA: evidence that rne and ants are the same genetic locus. Mol Mi~'robiol 4, 2127-2135 Maloy S, Nunn WD (1981) Selection for loss of tetracycline resistance by E coli. J Bacteriol 145. 11101112 Neidhardt FC, Bloch PL, Smith DF (19741 Culture media for Enterobacteria..I Bacteriol i 19, 736-747 Taraseviciene L, Miczak A, Apirion D (1991) The gene specifying RNase E (rne) and a gene affecting mRNA stability (ares) are the same gene. Mol Microbiol 5. 851-855 Murakawa G J, Kwan C, Yamashita J, Nierlich DP ( 1991 ) Transcription and decay of the lac messenger: Role of an intergenic terminator. J Bacteriol 173, 28-36 McCormick JR, Zengel JM, Lindahl L (1991) Intermediates in the degradation of mRNA from the lactose operon of E coli. Nucleic Acids Res 19, 27672776 Kennell D (1986) The instability of messenger RNA in bacteria. In: Ma.rimizing gene eapression (Reznikoff W. Gold L, eds), Butterworth, USA, 101-142 Ghora BK, Apirion D (1978) Sequence analysis and in vitro processing to p5 rRNA of a 9S RNA molecule isolated from an rne mutant of E coli. Cell 15. 10551066 Ono M, Kuwano M (1979) A conditional lethal mutation in an E coil strain with a longer chemical lifetime of mRNA. J Mol Biol 129, 343-357 Mudd EA, Prentki P. Belin D, Krisch HM (1988) Processing of uns.table bacteriophage T4 gene 32 mRNAs into a stable species requires E coli RNase E. EMBO J 7, 3601-3607 Bremer H, Dennis PP (1987) Modulation of chemical composition and other parameters of the cell by growth rate. hi: Escherichia coli and Salmonella ~.phimuritmt: Celhdar and molecu!ar biology (Neidhardt F et al. eds) ASM, Washington DC. USA, 1527-1542 Cho KO, Yanofsky C (1988) Sequence changes preceding a Shine-Dalgarno region influence o7~E mRNA translation and decay. J Mol Bio1204, 51-60