In vitro expression of the Escherichia coli nusA-infB operon

(1)

HAL Id: hal-02725409

https://hal.inrae.fr/hal-02725409

Submitted on 2 Jun 2020

HAL

is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire

HAL, est

destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

In vitro expression of the Escherichia coli nusA-infB operon

Yves Cenatiempo, Françoise Deville, Nathan Brot, Herbert Weissbach

To cite this version:

Yves Cenatiempo, Françoise Deville, Nathan Brot, Herbert Weissbach. In vitro expression of the

Escherichia coli nusA-infB operon. Journal of Biological Chemistry, American Society for Biochemistry

and Molecular Biology, 1987, 262 (1), pp.152-157. �hal-02725409�

(2)

Q 1987 by The American Society of Biological Chemists, Inc. Printed in U.S.A.

In Vitro Expression of the Escherichia coli nusA-infB Operon*

(Received for publication, April 28, 1986)

Yves CenatiempoS, Francoise DevilleS, Nathan Brot, and Herbert Weissbach

From the $Laborutoire de Bwlogie Molkculaire, Universitk Lyon I, 43 Boulevard du I 1 Novembre 1918,69622 Villeurbanne, France and the Roche Institute of Molecular Biology, Roche Research Center, Nutley, New Jersey 07110

The expression of the nmA-infB operon has been investigated using an in vitro system based on the formation of the first dipeptide of the gene product. A series of plasmids containing various deletions of the operon were used as templates in this study. Of the four genes coding for protein products, 15Ka, nusA, infB, and 15Kb, only 15K" was not expressed in this dipeptide system. The initial dipeptides for the other gene products, met-Asn (pnusA), met-Thr ( I F - 2 4 , and met-Ala (p15Kb), were synthesized even from plasmids lacking the primary promoters. It appears that secondary (internal) promoters in the operon can efficiently direct the expression of these genes. No regulation of the expression was observed with IF-2a, but pnusA inhibited the expression of the nusA gene (autoregulation) as well as t h e ~ 1 5 ~ gene. Experiments using an uncoupled system indicated that the effect of pnusA on nusA expression was at the level of transcrip- tion, but that both a transcriptional and a post-tran- scriptional effect of pnusA was seen on 15Kb expres- sion.

The nusA-infB operon is located at 69 min on the chro- mosome map of Escherichia coli (1-3). This complex transcriptional unit includes at least five genes. The first (metY) codes for t R N A p , a minor form of initiator tRNA? (4), whereas the four other genes, namely 15Ka, nusA, infB, and 15Kb, code for five proteins: pnusA, originally referred to as L factor (5, 6), two proteins of unknown function, p15K" and p15Kb, characterized by a similar molecular weight of 15,000 (7, 8), and the a and

6

forms of the translational initiation factor IF-2l(9,10). The a form is synthesized from the longest open reading frame on the infB mRNA, whereas the p form has recently been shown to arise from an internal initiation start on the mRNA (11). I n vivo expression of these genes depends on a primary promoter located 20 to 50 bases upstream of the metY gene (8) and the proposed order for transcription is metY-15K"-nusA-infB-15Kb (8, 12). Fig. 1 shows a schematic map of this transcription unit including some flanking genes and 3 potential transcription termination sequences (7, 13). It has been clearly established that the flanking genes, argG, rps0, and pnp, are part of independent transcriptional units (8, 14), but the exact location of the end of the operon is not clearly established. Downstream of the 15Kb gene, there is an open reading frame that corresponds to a protein of 35,000 daltons, but a gene product of this size

~~ ~

* This work was supported by Grant 83 V 0625 from the French M. R. I. and Grant UA 528 from the Centre National de la Recherche Scientifique. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked ''advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

l The abbreviations used are: IF, initiation factor; bp, base pair.

has not been detected in clones spanning this area (8).

The DNA sequence of the operon suggests that all the genes are co-transcribed from the primary promoter. Whether internal promoters become active under some circumstances remains questionable. On the one hand, a study using mini- cells harboring a promoter-deletion plasmid showed that p15K" and pnusA are not formed and that IF-2a, IF-26, and p15Kb are synthesized at only 5% of the level of parental plasmids (8). However, studies in maxicells (12) and in a DNA directed coupled in vitro protein synthesis system, directed by plasmids which were lacking the leader region of the operon (ll), suggested that infB and possibly nusA could be expressed from internal promoters.

Previous in vivo data on the nusA-infB operon showed that there is autogenous regulation of pnusA synthesis and that IF-2 synthesis is affected as well (15, 16). Since pnusA is known to bind to RNA polymerase (17) and modulate transcription at pause, attenuator, and termination sites (5, 18- 26), pnusA could regulate its own synthesis by affecting the two rho-independent terminators, t l and t2, lying between metY and 15K" and/or function at, as yet, undetermined pause sites. There is some evidence that a pnusA-polymerase complex may require a specific DNA sequence termed Box A (27) for interaction with the template (28).

In the present study, we have investigated the in vitro expression of the different protein coding genes of the nusA- infB operon. A simplified DNA-directed in vitro system which measures the synthesis of the initial dipeptide of the gene product has been mainly used (29). The results presented here indicate that internal promoters in the nusA-infB operon are quite active in vitro and that pnusA can autogenously regulate its own synthesis. In addition, expression of the 15kb is strongly inhibited by pnusA.

MATERIALS AND METHODS

Plasmids-The construction and isolation of plasmids pB19-1, pB18-1, pBA-2, pB17-1, and pB16-1 have been described elsewhere (12, 15). These plasmids, as well as pG9-3B (construction to be described elsewhere) were kindly provided by J. A. Plumbridge (In- stitute of Biology and Physico-Chimique, Paris). pG9-3B was used to construct pFY9 by deleting a HindIII-Sal1 fragment, leaving on the vector pGA-39 (30) a large portion of the nusA-infB operon from the PstI site in front of the main promoter to the Hind111 site located 220 bp within the 15Kb gene. Plasmids pFY18-1, pFY17-1, and pFY16-1 were derived from the corresponding pB series by transfer- ring the inserted fragments of the nusA-infB operon from the original vector, pBR322, to another vector, pRLS100, a pBR325 derivative in which the beginning of the /3-lactamase gene is missing (31). For that purpose, SulI-Hind111 fragments carrying the various regions of the nusA-infB operon and a small amount of the tetracycline-resistant genes were excised from pBR322 and religated into pRLS100. Recom- binant plasmids were used to transform E. coli RR1 strain, selecting them for chloramphenicol resistance, and screening for ampicillin and tetracycline sensitivity.

Biochemicals-Purified E. coli tRNA isoacceptor species tRNA?, tRNA1", tRNhnr, tRNbThr, and tRNA? were purchased from Sub- riden RNA, Rolling Bay, WA, and t R N A p from Boehringer Mann-

152

(3)

Expression of the E. coli nusA-infB Operon 153

-

P t,.t, (PI (PI (1,)

1

n I

metY 15Ka nusA inf B 15Kb

FIG. 1. Schematic map of the nusA-infB operon. P, primary promoter operon; t l and t2, two rho-independent terminators; ( P ) putative secondary promoters for nusA and infB; (t3) putative terminator or pause site. The direction of transcription is from left to right.

heim. Unfractionated E. coli tRNA (type X X ) used for tRNAA""

aminoacylation was from Sigma. A 0.25 M DEAE salt eluate (5) was used as the source of enzymes to acylate the tRNA species and to formylate methionyl-tRNA?". RNA polymerase was purified as previously reported (32). DNA restriction endonucleases were from com- described elsewhere (33). Ribonuclease inhibitor (RNasin) was pur- mercial source and used as recommended by the manufacturers or as chased from Genofit, Geneva. NusA protein (pnusA) was kindly provided by M. Grunberg-Manago (Institute of Biology and Physico- Chimique, Fond. Edmond de Rothschild, Paris).

In Vitro Transcription Using a DNA Fragment-A 1600-bp fragment was prepared by digesting pB17-1 with the restriction enzyme AluI. The fragment was separated by electrophoresis on a 1% agarose gel, extracted from the gel using an optimized freeze-squeeze method (34), and further purified on a minicolumn of DEAE-cellulose (35).

The incubation conditions for in uitro transcription based on [3H]

UMP incorporation into RNA were as described previously (36).

I n Vitro DNA-directed Dipeptide and Protein Synthesis-The incubation conditions for in uitro dipeptide synthesis were as previously described (11, 29) except that 4 units of ribonuclease inhibitor were added to each incubation mixture. The assay for the dipeptide product was based on the separation of the 3H-dipeptide from the free 3H- amino acid on a minicolumn of Dowex 5OWX-8 (H' form) as described earlier (37).

In some experiments, translation was uncoupled from transcription, using a preliminary 30-min incubation at 37 "C in the absence of aminoacyl-tRNA and ribosomes. After the first incubation, rif- ampicin (29 pg/ml) was added to stop RNA synthesis and translation was initiated by the addition of the required aminoacyl-tRNAs and 70 S ribosomes.

The in uitro synthesis of proteins in a highly defined system has been described elsewhere (18), except that 4 units of ribonuclease inhibitor were added to the incubation mixture in place of the nuclease inhibitor in the ascites fraction. =S-Labeled proteins synthesized in uitro were separated by electrophoresis on 12% polyacrylamide gels (38) and the proteins were detected by fluorography.

RESULTS

I n Vitro Synthesis of the N-terminal Dipeptides of the Gene Products-Various plasmids carrying different regions of the nusA-infB operon were used throughout this study (see Fig.

2). bB19-1 and pFY-9 contain the primary promoter as well as the metY, 15K", nusA, and infB genes and part of t h e 15Kb gene (220 bp). pB19-1 also contains a segment of t h e argG gene. pB18-1 contains a BamHI-Hind111 fragment that is lacking the genetic information for the primary promoter, the metY gene, and the first 170 bp of the 15K" gene. The end of 15K" (250 bp) and a truncated nusA gene (1370 out of 2330 bp) are present i n pB17-1, whereas the end of nusA, the entire infB gene, and the first 220 bp of 15Kb are present i n pB16- 1. The plasmid pBA-2 contains the end of 15K", the entire nusA gene, and approximately 1000 bp of infB. Its exact junction with pBR322 is not known (12).

The N-terminal dipeptides specified by the genes contained i n the nwA-infB operon are shown i n Table 1. The dipeptides have been determined from either the DNA (4, 7, 13) or protein sequence (11).

Table I1 shows that the various plasmids direct the synthesis of the expected dipeptides with the exception of t h e dipeptide for p15K" which is Met-Ile. With plasmids pBA-2, pB17-1, pB18-1, and pFY16-1 no 15K" expression was ex-

B 5

I i

_pB17-1

OT pFYl7-I

i

,I.:- pBA-2

i

^w pBl6-1

-

or ~ F Y I W

FIG. 2. Map of the bacterial inserts in the different plasmids listed. See text and Refs. 12 and 15 for detailed construction of the plasmids. Thick lines represent vector DNA. The dotted lines in pBA- 2 indicate that the exact junction between the vector and the insert is not known (12). 15Kb' denotes that the 15Kb gene is truncated. pB plasmids, vector pBR322: pFY-9, vector pGA39; other p F plasmids;

vector pRLS100. B , BamHI; E, EcoRI; H, HindIII; P, PstI; S , Sau3A.

TABLE I

N-terminal dipeptides of the gene products of the nusA-infB operon The data compiled in this Table are taken from previous reports (4, 8, 11-13). The tRNA isoacceptor species needed for dipeptide synthesis were determined according to codon-anticodon base pairing (41). tRNkTh"' can also recognize the codon ACA (11).

C I Isoacceptor

J

Gene N-terminal Nucleotide tRNA for second amino

acid

product dipeptide coding

sequence metY t R N A p

1 5 P p15K" met-Ile AUG-AUU tRNA"'"

nusA pnusA fMet-Asn AUG-AAC tRNAA""

infB I F - ~ c Y m e t - T h r AUG-ACA tRNLThr

IF-28 met-Ser GUG-AGC tRNA?

15Kb p15Kb fMet-Ala AUG-GCG tRNA$'"

~ ~ -~ - . ____

TABLE I1

Dipeptide synthesis directed by p k m i d s containing segments of the nusA-infB operon

Incubation conditions for dipeptide synthesis were as described using 1 pmol of plasmid DNA as template (11, 29). tRNA species charged with the appropriate 3H-amino acid were used as listed in Table I. The assay for the dipeptide product is described in the text (see "Materials and Methods" and Ref. 37). See Table I for the gene products and corresponding N-terminal dipeptides

Dipeptide 'lasmid Met-Ile Met-Asn Net-Thr Net-Ala

(p15K') (pnusA) (IF-2a) ^(p15Kb) pmol

pB19-1 0 2.6 4.0 4.4

PBB-1 0 2.3 1.6 4.3

PBA-2 0 1.9 1.2 0

pB17-1 0 2.5 0.6 0

pFY16-1 0 0 1.7 1.3

"

. - .

~

"

. I _

"

pected since these plasmids are lacking either the 15K" gene or its leader region (Fig. 2). However, even pB19-1, which has t h e 15K" gene, does not direct the synthesis of M e t - I l e although there is synthesis of M e t - A s n , M e t - T h r , a n d M e t - Ala, the respective initial dipeptides of pnusA, IF-&, and p15Kb. The lack of 15K" gene expression from pB19-1 could

(4)

be due to an inactive primary promoter, inefficient initiation of translation, or transcription termination at tl and/or t2. If the primary promoter is the only promoter functioning in pB19-1, the lack of expression of p15K" is not due to termination at t l and/or t 2 for, if so, the synthesis of the other dipeptides from the downstream genes would also be expected to be low. It should be stressed that there is no evidence in these in uitro studies that the primary promoter is functioning or that the 15K" mRNA is formed but not translated. How- ever, as will be discussed below, there is reason to believe that secondary promoters are responsible for the bulk of the expression of the nusA, infB, and 15Kb observed in these experiments.

Met-Asn synthesis (pnusA) is observed with all of the plasmids with the exception of pFY16-1 (Table 11). This result is expected since the 5' coding region of nusA is absent in pFY16-1. M e t - T h r (IF-201) synthesis was always obtained with the plasmids tested even when a large distal segment (-2 bp) of infB was absent (pBA-2). It should be noted that a low, but significant synthesis of M e t - T h r was obtained when the infB gene was absent (pB17-1). This is due to background formation of this dipeptide directed by pBR322, as has been observed previously ( l l ) , and corresponds to an unidentified gene product.

Finally, as seen in Table 11, Met-Ala synthesis (p15Kb) is obtainedwhen plasmids pB19-1, pB18-1, or pFY16-1 are used, which contain the 5' proximal coding region of the 15Kb gene.

In contrast, when pBA-2 or pB17-1 is the template there is no Met-Ala synthesis since these plasmids are lacking this gene. The ability to express the nusA and infB genes in plasmids lacking the primary promoters (e.g. pB17-1 and pB18-1) indicates that transcription directed by these plasmids may initiate from secondary promoters within the 15K"

gene and not from the primary promoter. This would also explain why there appears to be no expression of the 15K"

gene from any of the plasmids (see above). The ability to express 15Kb using plasmid pFY16-1, which is lacking the primary and putative secondary promoter, indicates that there may be a promoter for this gene either within the info structural gene or within the 3' end of the nusA gene.

Regulation of the nusA-infB Operon

Effect of IF-2a-Since the formation of Met-dipeptides provides a rapid and gene specific procedure to study regulation of gene expression, this technique was initially used to examine the effect of the gene products, IF-% and pnusA, on the in vitro expression of the various genes of the operon.

Additional experiments were carried out using a highly defined system in which the synthesis of the entire gene product is obtained (18).

IF-2 is required for initiation of translation in vivo and in vitro and shows an absolute dependency in the dipeptide system (29). Experiments were performed to test the effect of IF-2a on the expression of its own gene, infB, as well as the expression of the neighboring genes, nusA and 15Kb. As seen in Table 111,1.2 pg of IF-Sa/assay (34 pg/ml) produces a near maximum synthesis of Met-Asn (pnusA), Met-Thr (IF-2a), and Met-Ala (p15Kb). A &fold excess of the factor in the incubations gives only a 10% stimulation of Met-Asn and Met-Ala synthesis, but a somewhat greater stimulation of M e t - T h r synthesis. These results suggest that besides its requirements for the initiation of protein synthesis, IF-2a does not exert a significant regulatory function on the expression of these three gene products present in the nusA-infB operon. Of particular interest is that there is no autogenous regulation of IF-2a synthesis.

TABLE I11

Effect of IF-Sa on dipeptide synthesis

IF-2a was either omitted or added at the beginning of the incuba- tion. pB18-1 was used as template in this experiment. The results are presented as in Table 11.

Dipeptide

Met-Thr Met-Ala IF-2a Met-Asn

(pnusA) (IF-2a) (p15Kb)

d m l P m l

0 0 0 0

34 2.9 2.6 4.4

170 3.2 3.9 4.8

0 2 4 6 E l 0 1 2 1 4

pw A .

FIG. 3. Effect of pnusA on Net-Asn (pnusA), Met-Thr (IF- 24, and met-Ala (p16Kb) synthesis. DNA-directed dipeptide synthesis was monitored in a transcription-translation coupled sys- tem (11, 29,37) in the presence of varying amounts of pnusA added prior to the beginning of the reaction. Plasmids pB19-1 or pFY-9 were used as template for met-Asn synthesis and plasmid pFY16-1 was used as template for met-Thr and met-Ala synthesis.

Effect of pnusA-Since pnusA plays an important role within the cell in modulating transcription at pause, atten- uation, and termination sites (5,18-26), an attempt was made to evaluate the effect of this factor on its own synthesis and the formation of IF-2a and p15Kb in the dipeptide system.

Fig. 3 shows the effect of pnusA on met-Asn, Met-Thr, and Met-Ala synthesis. In these experiments pB19-1 (or pFY9), which contains the major part of the operon, was used to direct Met-Asn synthesis, whereas pFY16-1, which contains the infB and truncated 15Kb gene was template for M e t - T h r and Met-Ala synthesis.

The addition of 0.3 pg (8.6 pg/ml) of pnusA to the incubations resulted in 70% and 50% inhibition of Met-Ala and Met-Asn synthesis, respectively. Under the same experimen- tal conditions Met-Thr synthesis was decreased only by 20%.

Slightly higher amounts of the factor (0.4 pg) increased the inhibition of Met-Asn and Met-Thr synthesis to 70% and 25%, respectively. Whether the overall effect on Met-Thr is specific remains to be established, since control experiments, either on chloramphenicol acetyltransferase (Met-Glu) and 0-lactamase (Met-Ser) synthesis, showed about a 15% inhibition of dipeptide synthesis in the presence of 0.3 to 0.4 pg of pnusA. A similar effect has been observed previously when measuring met-Ser synthesis directed by a plasmid containing the @-lactamase gene (19).

However, the results on Met-Asn and Met-Ala synthesis indicate that pnusA is inhibiting the expression of the nusA and 15Kb genes. Interestingly, an identical effect of pnusA on Met-Asn synthesis was observed using either pB18-1 or pB17-1 (data not shown), which are both lacking the primary promoter of the operon. Thus, the autogenous regulation of pnusA synthesis does not appear to involve transcription from

(5)

Expression of the E. coli nusA-infB Operon 155 this promoter in the present in vitro system. As stated above,

it appears that a secondary promoter within the 15K" gene may be responsible for expression of the nusA and infB genes.

The inhibition of 15Kh gene expression by pnusA also indicates that pnusA is functioning a t a site downstream from an internal promoter located within the infB or at the 3' end of the nusA gene, since the results shown in Fig. 3 were obtained with pFY16-1 as template. A similar inhibition was obtained with pB16-1 or other plasmids harboring a longer fragment of the operon such as pB18-1 (data not shown).

In order to confirm the results of the dipeptide system, which indicated that there is autogenous regulation of pnusA synthesis, experiments were carried out in highly defined in vitro protein synthesis systems in which the entire gene product was made (18). Fig. 4 shows that pB17-1 directs the synthesis in this system of two major radioactive proteins, p- lactamase and a protein of a M , = 60,000. This latter protein corresponds to the product of the truncated nusA gene (nusA') on pB17-1 (see Fig. 2). According to the DNA sequence, this protein should have a M , of 50,000. However, pnusA itself behaves anomalously during gel electrophoresis, migrating as a 69-kDa protein although its actual mass is -55 kDa. In the presence of pnusA a marked decrease in the synthesis of the truncated pnusA is observed. In contrast, (3-lactamase synthesis is only slightly affected by pnusA under the same conditions.

Other experiments were performed to verify that pnusA was functioning at the level of transcription. In these experiments, the dipeptide system was uncoupled and formation of either Met-Asn (pnusA) or fMet-Ala (p15K") was determined when pnusA was added either at the beginning of transcription or before translation. Table IV shows that a 50% inhibition of N e t - A s n formation was observed when

200-

92 -

68- -

^--pnusA'

45 -

30-

1 2

--Wactamase

FIG. 4. Gel electrophoresis of 36S-labeled proteins synthe- sized from pB17-1 in a highly defined system. Details of the incubation are described elsewhere (18). Lanes I and 2 show the in vitro products synthesized, respectively, in the absence or presence of pnusA (14 pglml). Protein standards (kDa) are shown in the left margin. p n w A ' denotes a truncated form of pnusA (see text for details).

TABLE IV

Effect of p n w A on dipeptide synthesis in an uncoupled system The incubation conditions for dipeptide synthesis in which tran- scription and translation are uncoupled are described in the text (see

"Materials and Methods"). pnusA (0.3 pglincubation or 11.4 pglml) was added either at the beginning of transcription or 30 min later when transcription was stopped and the components required for translation were added.

Dipeptide synthesis pmol (5) Addition of pnusA ^~-Met-Asn Met-Ala

(anusA) (o15Kb)

Omitted 0.58 (100) 1.10 (100)

Before transcription 0.29 (50) 0.23 (21) After transcrbtion 0.55 (97) 0.72 (66)

0 1 2 3 4 5

time. min

FIG. 5. AluI restriction map of pB17-1 and in vitro tran- scription of a 1600-bp AluI fragment in the presence or absence of pnusA. A , vertical arrows indicate A h 1 restriction sites;

E, EcoRI. B , ['HH]UMP. Incorporation into RNA was measured as described elsewhere (36) using 1 pg of the fragment in a 25-pl incubation. Control, 0 +p-nusA, 0.

pnusA was added before transcription, whereas no inhibition could be seen when pnusA was added after transcription, i . e a t t h e beginning of translation. This result demonstrates that pnusA is autogenously regulating its own synthesis at the level of transcription. This was confirmed by measuring RNA synthesis directly. For these experiments ['HIUMP incorporation was directed by a 1600-bp AluI fragment (Fig. 5) encompassing a segment of the tet gene (pBR322) plus the end of 15K" gene (250 bp) and 1000 bp of nusA. As seen in the Figure, pnusA reduces the incorporation of ['HIUMP into RNA by about 50%. This effect was not nonspecific for similar experiments using other templates, such as pBR322 which showed no effect of pnusA on UMP incorporation into RNA.

Surprisingly, the pnusA effect on fMet-Ala synthesis (p15Kh) does not seem to be solely due to an effect on transcription. As shown in Table IV, when pnusA is added immediately before transcription, a 79% inhibition of M e t - Ala synthesis is obtained. However, pnusA also inhibits fMet- Ala synthesis when added after transcription, although to a lesser extent (34%). These results suggest that pnusA exerts a dual effect on 15Kb gene expression, both a t transcription and at a post-transcriptional event.

DISCUSSION

A variety of mechanisms are probably involved in regulating the synthesis of the various gene products of the nusA-infB operon. This operon is somewhat unusual since a tRNA gene is located within this transcription unit and one of the genes, infB, encodes two proteins, IF-2n and IF-2P (11). Different groups have investigated the in vivo effect of the known gene products of this operon on either their own synthesis or the

(6)

expression of neighboring genes (15, 16). In this report we have employed a simplified in vitro system, based on the formation of the first dipeptide of the gene product, to examine the regulation of nusA-infB expression. This technique has previously provided useful information on the regulation of a large number of prokaryotic genes (19,20, 29,31,37,39, 40).

Generally, in multigene operons, the problem of transcription initiation can be quite complex. In most cases, transcription is initiated in vivo at a primary promoter and the nusA- infB operon seems to follow this rule (8). However, our in vitro results indicate that internal promoters can very efficiently direct the synthesis of pnusA, IF-2a and -28, and p15Kb. When plasmid pB19-1, which contains the primary promoter and most of the operon, is used as template, the synthesis of the initial dipeptides of all of the products is seen with the exception of Met-Ile (p15K"). As discussed above, the lack of Met-Ile synthesis in the present system could be due to an inactive primary promoter, termination at the tl and/or t 2 sites that immediately precede the 15K" gene, or the lack of an efficient translational initiation signal. The latter is mentioned since the protein ribosome-binding site sequence of UAG appears to be rather weak. Any of the above are possible, but other results favor transcription termination.

It was demonstrated, in in vitro transcription experiments, that the terminators t l and t 2 are particularly efficient since 90% of the transcripts were terminated at these sites (4). Also, a similar situation is encountered in the MMS operon where the N-terminal dipeptide of the primase (dnaG gene product) is not synthesized in a defined in vitro system, presumably because of the presence upstream of a strong rho-independent terminator (20). A consequence of transcription termination is that downstream internal promoters must be capable of allowing the synthesis of other gene products. This interpre- tation is strengthened in the present experiments by the fact that pB16-1, pB17-1, pB18-1, and the equivalent plasmids in the pFY series, which are lacking the primary promoter, all direct the synthesis of the N-terminal dipeptides of the downstream gene products. Although all of the results could be explained if initiation took place from a plasmid promoter, this does not appear to be the case. The DNA fragments of 1 he rrusA-itrfH ol)eron, presenl, i n 1.hese plasmids, were cloned in an opposite direction to the tet gene of pBR322 or pRLSlOO (12) as determined by restriction enzyme analysis (data not shown). This minimizes the possibility that the cloned genes were transcribed from a vector promoter. It thus appears in our in vitro system that essentially all of the expression observed is due to the activity of secondary (internal) promoter.

The downstream 15Kb gene is efficiently expressed based on Met-Ala synthesis. This suggests transcription initiation either from a promoter upstream of the infB gene, from an internal promoter within the infB gene, or in the intercistronic region between infB and 15Kb. Other constructions will be necessary to test this point. Assuming that transcription is initiated before the t3 stem and loop structure (Fig. l ) , p15Kb synthesis can occur only if this potential terminator is rela- tively inefficient in the absence of any external factor.

The regulation of pnusA, IF-2a, and p15Kb has been further investigated. As expected, the in vitro expression of the three genes is dependent on the presence of IF-2a. Although in the dipeptide system saturating amounts of this factor necessary for maximum synthesis of the three proteins appears to be slightly different, there is no apparent regulation by IF-2.

This result is in agreement with previous in vivo data showing that an overproduction of IF-2, per se, does not lead to

alterations in pnusA, IF-Ba, or IF-ZP synthesis (15, 16).

In contrast, pnusA is involved in the autogenous regulation of its own synthesis and has a negative effect on infB and 15Kb gene expression. In vitro, pnusA inhibits transcription of the nusA gene, very likely at a pause or attenuator site.

Previous in vivo studies (15) led to a similar conclusion, but indicated that the most likely sites of pnusA action were the termination sites t l and t2. The studies presented here appear to eliminate t l and t2 since these terminators are absent in some of the templates showing a pnusA effect. In order to search for a potential pnusA recognition site, 15K" and nusA DNA sequences were screened for the presence of Box A-like sequences (27). One sequence, CGCTGGTTA, similar to the consensus Box A sequence CGCTCTTA, was found starting at nucleotide 184 within the nusA structural gene. Although pausing at this site could explain the mechanism by which pnusA regulates its own synthesis, it does not explain how pnusA inhibits synthesis of the first dipeptide of pnusA which is upstream of the putative pause site.

Finally, pnusA also regulates p15Kb synthesis in vitro. The preliminary data indicate that pnusA inhibits transcription but also functions at a post-transcriptional step. A translational effect of pnusA has not been reported and experiments with isolated mRNA should be performed to address this point directly.

Acknowledgments-We thank M. Grunberg-Manago, J. Plum- bridge, and A. Cozzone for helpful discussions.

REFERENCES

1. Bachman, B. J. (1983) Microbwl. Reu. 4 7 , 180-230

2. Plumbridge, J. A., Howe, J. G., Springer, M., Touati-Schwartz, D., Hershey, J. W. B., and Grunberg-Manago, M. (1982) Proc.

Natl. Acad. Sci. U. S. A . 79,5033-5037

3. Kurihara, T., and Nakamura, Y. (1983) Mol. Gen. Genet. 190, 4. Ishii, S., Kuroki, K., and Imamoto, F. (1984) Proc. Natl. Acad.

5. Kung, H., Spears, C., and Weissbach, H. (1975) J. Biol. C h m . 6. Greenblatt, J., Li, J., Adhya, S., Friedman, D., Baron, L., Redfield, B., Kung. H.. and Weissbach, H. (1980) Proc. Nutl. Acad. Sci.

189-195

Sci. U. S. A . 81,409-413 260,1556-1562

u. s.

^A.1 7 , &I-1994

7. Ishii. S.. Ihara. M.. Maekawa. T., Nakamura. Y.. lJchida, H.. and Imamoto, F.'(1984) Nucl& A&s /it%. 12,3333-3342

8. Nakamura, Y., and Mizusawa, S. (1985) EMBO J. 4,527-532 9. Hershey, J. W. B., Yanov, J., and Fakunding, J. L. (1979) Methods 10. Dondon, J., Plumbridge, J. A., Hershey, J. W. B., and Grunberg-

Manago, M. (1985) Biochmie (Paris) 67,643-649

11. Plumbridge, J. A., Deville, F., Sacerdot, C., Petersen, H. U., Cenatiempo, Y., Cozzone, A. J., Grunberg-Manago, M., and Hershey, J. W. B. (1985) EMBO J. 4,223-229

12. Plumbridge, J. A., and Springer, M. (1983) J. Mol. Biol. 167, 13. Sacerdot, C., Dessen, P., Hershey, J. W. B., Plumbridge, J. A., and Grunberg-Manago, M. (1984) Proc. Natl. Acad. Sei. U. S. A . E n ~ y m l . 60, 3-11

227-243

_ . . .

81, 7787-7761

14. Portier, C., and Regnier, P. (1984) Nucleic Acids Res. 1 2 , 6091- 6102

15. Plumbridge, J. A., Dondon, J., Nakamura, Y., and Grunberg- Manago, M. (1985) Nucleic Acids Res. 13,3371-3388

16. Nakamura, Y., Plumbridge, J., Dondon, J., and Grunberg-Man- ago, M. (1985) Gene (Amst.) 36,189-193

17. Greenblatt, J., and Li, J. (1981) Cell 24, 421-428

18. Zarucki-Schulz, T., Jerez, C., Goldberg, G., Kung, H. F., Huang, K. H., Brot, N., and Weissbach, H. (1979) Proc. Nutl. Acad.

Sci. U. S. A . 76,6115-6119

19. Peacock, S., Cenatiempo, Y., Robakis, N., Brot, N., and Weiss- bach, H. (1982) Proc. Nutl. Acad. Sci. U. S. A . 79,4609-4612 20. Peacock, S., Lupski, J. R., Godson, G. N., and Weissbach, H.

(1985) Gene (Amst.) 33,227-234

21. Friedman, D. I., and Gottesman, M. (1983) in Lambda ZZ (Hen-

(7)

Expression of the E. coli nusA-infB Operon 157

drix, R. W., Roberts, J. W., Stahl, F. W., and Weisberg, R. A., eds) pp. 21-51, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

22. Kingston, R. E., and Chamberlin, M. J. (1981) Cell 2 7 , 523-531 23. Hauser, C. A., Sharp, J. A., Hatfield, L. K., and Hatfield, G. W.

24. Farnham, P. J., Greenblatt, J., and Platt, T. (1982) Cell 29,945- 25. Greenblatt, J., Limont, M., and Hanley, S. (1981) Nature 292, 26. Kassavetis, G. A., and Chamberlin, M. J. (1981) J. Biol. Chem.

27. Olson, E. R., Flamm, E. L., and Friedman, D. I. (1982) Cell 31, 28. Tsugawa, A., Kurihara, T., Zuber, M., Court, D. L., and Naka-

mura, Y. (1985) EMBO J. 4,2337-2342

29. Robakis, N., Cenatiempo, Y., Meza-Basso, L., Brot, N., and Weissbach, H. (1983) Methods Enzymol. 101, 690-706 30. An, G . , and Friesen, J. D. (1979) J . Bacteriol. 140,400-407

(1985) J. Biol. Chem. 260, 1765-1770 951

215-220 256.2777-2786 61-70

31. Shoeman, R., Redfield, B., Coleman, T., Greene, R. C., Smith, A.

A., Brot, N., and Weissbach, H. (1985) Proc. Natl. Acad. Sci.

32. Burgess, R. R., and Jendrisak, J. J. (1975) Biochemistry 14, 33. Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

U. S. A. 82, 3601-3605 4634-4638

34. Tautz, D., and Rem, M. (1983) Anal. Biochem. 132, 14-19 35. Smith, H. 0. (1980) Methods Enzymol. 65, 371-380

36. Cenatiempo, Y., Robakis, N., Reid, B. R., Weissbach, H., and Brot, N. (1982) Arch. Biochem. Biophys. 218,572-578 37. Weissbach, H., Robakis, N., Cenatiempo, Y., and Brot, N. (1984)

Biotechniques 2.16-22

38. Laemmli, U. K. (1970) Nature 227,680-685

39. Peacock, S., and Weissbach, H. (1985) Arch. Biochem. Biophys.

40. Peacock, S., Weissbach, H., and Nash, H. A. (1984) Proc. Natl.

Acad. Sci. U. S. A. 81,6009-6013 41. Ikemura, T. (1981) J . Mol. Biol. 146,l-21

243,315-319