PROTEIN

EXPRESSION

AND PURIFICATION

1,

81-86 (1990)

Overproduction and Purification of the w Subunit of Escherichia co/i RNA Polymerase Daniel

R. Gentry

McArdle

Laboratory

Received

June

and Richard

R. Burgess’

for Cancer Research,

University

of Wisconsin-Madison,

Wisconsin,

53706

25, 1990

This paper reports the construction of plasmids which direct the overproduction of the w subunit of Escherichia coli RNA polymerase and the subsequent purification of w. Useful overproduction is achieved only if the natural ribosomal binding site region of rpoZ is replaced with the ribosomal binding site region of bacteriophage T7 gene 10. Overproduction is directed by T7 RNA polymerase which is provided on a separate plasmid. w is purified by three column steps either from the insoluble inclusion body fraction or from the soluble fractions of lysates. The final yield is approximately 2 mg o per 10 g cells wet wt. Additionally, we found that recombinant w is readily cleaved by an endogenous protease. Sequence analysis of the most prevalent proteolytic fragment suggested that the protease responsible was the product of the ompT gene. Cleavage of (J is o 1890 Academic PWS, IUC. greatly reduced in ompTstrains.

Escherichia coli RNA polymerase is a large, complex enzyme containing multiple subunits (2). E. coli RNA polymerase comes in two forms. Core enzyme, consisting of a&9’, is competent for transcription but is unable to initiate transcription at specific sequences of DNA called promoters. Holoenzyme consists of core enzyme with a u factor bound to it. The presence of a u factor allows the enzyme to initiate transcription at promoters. There are several d factors in E. coli, each of which directs transcription of distinct promoters when bound to core RNA polymerase (12). Most work with proteins which interact with RNA polymerase has been concerned with u factors. Several other polypeptides, however, interact with RNA polymerase but have not been studied in much detail. One of these proteins, denoted w, is present in highly pure RNA polymerase. Very little is known about the function of w. It is a small protein of M, 10,105 with many charged residues (33%) and a predicted net charge of -4 at pH 7.0 (8).

1 To whom

Madison,

correspondence

should

1046-5928/90 $3.00 Copyright 0 1990 by Academic Press, All rights of reproduction in any form

be addressed.

Inc. reserved.

The gene encoding w, rpoZ, is located at approximately 82 min on the E. coli genetic map (9,18). Insertions in rpoZ are viable for cell growth under all conditions tested though RNA polymerase isolated from such insertion mutants lacks detectable o. Additionally, rpoZ is located in an operon that also encodes the gene SPOT (9,18). The SpoT protein is responsible for degrading ppGpp which is the mediator of the stringent response. The stringent response is characterized by an inhibition of stable RNA transcription following amino acid starvation (4). The proximity of rpoZ to SPOT raised the possibility that w is required for transcriptional regulation by ppGpp. Indeed, it has been asserted that w is required for stringent control (13) though results from our laboratory show conclusively that strains in which rpoZ is deleted still exhibit inhibition of stable RNA transcription following amino acid starvation (that is, they are still stringent) (Gentry, Xiao, Burgess, and Cashel, in preparation). Any serious study of the function of a particular protein requires that the protein be available in reasonable amounts. In the case of w, large amounts bound to RNA polymerase are available; however, this source is not satisfactory because w is not readily dissociated by nondenaturing treatments. In contrast to 6” (the most common u factor in E. coli), which can be removed from RNA polymerase by passage over phosphocellulose, there is no known nondenaturing treatment which will remove w. What is needed is a source from which w can be purified away from RNA polymerase and still retain polymerase binding activity. This purified w could then be used to study its effect on in vitro transcription, to raise antibodies, and to study its physical structure. In this paper, we report the construction of a strain which overproduces w, the purification of w from this strain, and the identification of an endogenous protease, OmpT, which greatly affects the yield of intact w. MATERIALS

AND

METHODS

Bacterial strains, plusmids, media, and growth conditions. All strains, plasmids, and their sources are listed 81

82

GENTRY

AND TABLE

Bacterial Strain BL21 C600 CF1528 UT5600 DG2 Plasmid pSP72 pGPl-2 pE3C pE3C-2

Strains

BURGESS 1 and

Plasmids Source

Genotype F-, F-, F-,

hsdS, gal, (E. coli B), OmpTthil, thrl, buB6, lacYI, tonA21, supE44 thi, pyrE60, argE3, hisl, proA2, thr, leuB6, mtll, rpsL, supE44, cdd, gyrA, relA1, A(rna), A(xth-pnc), F-, proC, k-6, trpE38, entA, AompT F-, galK, rpsL, luc-, rpoZ: :knn Relevant

xy15, arall, AompT

galK2,

J. Dunn (Brookhaven) Gentry and Burgess

characteristics

P,,-polylinker, amp’ Pn-T7 gpl0 (T7 RNA polymerase), P,-~1857, P,,-t-rpoZ, amp’, pSP72 derivative P,,-t-rpoZ-spoT, amp’, pSP72 derivative

lacYI,

H. Liao (Univ. Wisconsin-Madison) Laboratory Collection M. Cashel (NIH)

(8)

Source

pBR

in Table 1. Fermenter cultures for large-scale production of overproducing cells were grown in 4X LB medium (14) + 1% glucose. For temperature induction, cells were grown to an OD, of at least 7 at 30°C and then shifted to 42°C for 1 h before being harvested by tangential flow filtration. Cells from fermenter cultures were stored at -80°C until needed. Construction of overproducing plasmids. pE3C was constructed by removing the entire region upstream of rpoZ and replacing it with an oligonucleotide of sequence GTTAACTTTAAGAAGGAGATATACATATGGCA (see Discussion) by cloning the oligonucleotide into the MluI site located within the second and third codons of rpo2. The modified rpoZ was then cloned into the sole EcoRV site of pSP72 (Promega) which is 19 bp from the initiation site of a T7 RNA polymerase promoter. pE3C-2 was constructed by cloning SPOT downstream of rpoZ in its natural orientation into pE3C. The overproducers are shown schematically in Fig. 1. w purification. Ten grams of frozen induced BL21 (pE3C-2, pGPl-2) is thawed by mixing with 30 ml lysis buffer (50 mM Tris-HCl, pH 7.9, 100 mM NaCl, 2 mM EDTA, 1 mM phenylmethylsulfonyl fluorophosphate (PMSF, Sigma), and 10 mM benzamidine (Sigma)). PMSF and benzamidine are added just before use to inhibit proteases. Lysozyme is added to 0.5 mg/ml and the mixture is stirred at room temperature for 20 min. Triton X-100 is then added to 1% and stirring is continued for 20 min. To lower the viscosity and to ensure complete lysis, the mixture is passed twice through a cell disruption bomb (VWR). The insoluble pellet, which contains most of the w, is removed from soluble material by centrifugation at 30,OOOg for 30 min. The pellet is washed once with lysis buffer + 1% Triton X-100 and twice with lysis buffer alone by vigorous mixing in a tissue homogenizer. Insoluble w is solubilized by adding N-lauroylsarcosine (Sarkosyl) (Sigma) to 0.25% and

compatible,

kan’

Promega Corp. (Madison, WI) Tabor and Richardson (17) This work This work

stirring for 1 h at 4°C. Following centrifugation at 30,OOOg for 30 min, the mixture is dialyzed with four changes against 50 mM Tris, pH 7.9, 1 mM PMSF, and 10 mM benzamidine. After dialysis the mixture is, again, centrifuged at 30,OOOg and loaded at 4°C onto a 50-ml Q-Sepharose Fast Flow column (Pharmacia) equilibrated with 50 mM Tris, pH 7.9, and eluted with a linear O-O.5 M NaCl gradient in the same buffer. w elutes from this column at about 100 mM NaCl. Fractions containing w, identified by SDS-PAGE, are pooled and loaded directly onto a 20-ml Red-Sepharose dye affinity column equilibrated with 50 mM Tris, pH 7.9, and 0.2 M NaCl. After extensive washing with at least 5 column volumes, w is eluted with 50 mM Tris, 1 M NaCl. Fractions from the 1 M NaCl step are dialyzed until the conductivity is at or below the conductivity of 50 mM Tris, 50 mM NaCl and then loaded onto a Mono-Q 515 FPLC column (Pharmacia) and eluted with a O-O.5 M NaCl gradient. w elutes from this column at 100 mM NaCl. Because w’ lacks tryptophan and tyrosine, it does not absorb at 280 nm. It is therefore necessary to monitor the column either with protein assays or by absorbance at 210 nm. It is sometimes useful to monitor the columns with a dual wavelength detector because w is usually the only peak which absorbs strongly at 210 nm but not at all at 280 nm (this characteristic is useful in assessing the purity of a given sample of w). Fractions containing w are pooled and dialyzed against 20 mM Tris, pH 7.9,lOO mM NaCl, 0.2 mM EDTA, and 50% glycerol. Substantial but variable amounts of w are often present in the soluble fraction. This is recovered by Q-Sepharose, Red-Sepharose, and Mono-Q chromatography as outlined above. w purified from the soluble fraction behaves on columns identically to w purified from the solubilized pellet. Proteolytic fragment purification. C600 (pE3C-1, pGPl-2) is grown and induced as above. Lysis is carried out as above except lysozyme, Triton X-100, and benz-

PURIFICATION

GTTAACTTTAAGAAGGAGATATACAT

OF

w SUBUNIT

OF

T

pTkp

pE3C

pE3C-2

FIG. 1.

Schematic of w overproducing plasmids. The upstream region of rpoZ was modified by cloning an oligonucleotide encoding the T7 gpl0 ribosomal binding site in place of its natural sequences. S/D refers to the Shine-Dalgarno sequence (ribosomal binding site). t refers to the sequence determined by Olins et al. (15,16) to exhibit translational enhancer activity. The boxed ATG indicates the initiating codon of rpo.2’.The site of transcription initiation from the T7 promoter (PT7) is located 65 bp upstream of the ATG. The 3’-end of the fragment cloned into pSP72 to make pE3C is at the Hid site located 44 bp from the end of rp0.Z (includes the first 21 bp of spoT). The 3’-end of the fragment cloned to make pE3C-2 is at a PstI site located downstream of SPOT(18).

amidine are omitted. After solubilization with Sarkosyl and dialysis, the mixture is loaded directly onto a TSK DEAE 5PW (4.6 X 100 mm) HPLC column equilibrated with 20 mM Tris, pH 7.4, and eluted with a O-O.5 M NaCl gradient. Assay for RNA polymerase binding. To determine if purified w will bind RNA polymerase, an excess of w is incubated with 20 I.cg of w-free RNA polymerase (purified from an rpoZ null mutant, DG2 (9), by the method of Hager et al. (11)) in a buffer consisting of 20 mM Tris, pH 7.9, 0.25 M NaCl, and 0.1 mM DTT for 30 min at 37°C. The mixture is then mixed with Q-Sepharose equilibrated with the same buffer for 30 min. Free w will not bind the Q-Sepharose at the ionic strength of the buffer. The resin is washed several times and then RNA polymerase is eluted off the resin by adding ammonium acetate to 1 M. The resin is removed by centrifugation and the supernatant containing RNA polymerase is dried overnight in a Speed Vat and run on a 16% SDSpolyacrylamide gel. Protein sequencing. Protein sequencing was performed by the University of Wisconsin Biotechnology Center using a gas-phase sequencer. The proteolytic fragment was purified as above followed by further purification on a C, reverse-phase HPLC column prior to sequencing. Protein assay. Protein content was measured by using either the Bradford dye binding assay (Pierce) or the BCA assay (Pierce). Both assays agreed when measuring w concentrations using BSA as a standard.

Escherichiu

coli

RNA

a3

POLYMERASE

RESULTS

The plasmids illustrated in Fig. 1 Overproduction. direct the expression of w at a high level (around l-5% of total cell protein or more) when T7 RNA polymerase was provided by the plasmid pGPl-2. T7 RNA polymerase expression is under P, control in this plasmid and is induced by shifting the culture from 30 to 42°C (24) (see Table 1). Surprisingly, overproduction was not apparent when T7 RNA polymerase expression was provided by the prophage DE3 using the Studier system (22) where T7 RNA polymerase is under the control of the lac promoter. The level of overproduction tapers off after about 90 min (Fig. 2) when cells are grown in a shaking water bath. It appears that pE3C-2 overexpresses w slightly better than pE3C although no overproduction of SpoT is apparent (SPOT, 80,000 Da). Purification. The purification scheme developed for w is presented under Materials and Methods. Figure 3 shows the steps of w purification schematically as well as an SDS-gel analysis of fractions from key steps along the purification. As mentioned previously, the majority of w was insoluble when overproduced. The amount of soluble w is dependent on several factors. Surprisingly, if lysozyme is omitted, nearly all detectable w is insoluble. Although it is advantageous for w to be entirely insoluble using the purification scheme outlined under Materials and Methods, omitting lysozyme results in less efficient lysis even after passage through the cell disruption bomb. Lysis by passage through a MantonGaulin homogenizer, in the absence of lysozyme, results in complete lysis and nearly all detectable w is found in the pellet. We chose to solubilize w with Sarkosyl because it is the least denaturing, most effective reagent tested. Other detergents tried were 0.05 and 0.5% sodium deoxycholate (which worked slightly at both concentrations), 1% Triton X-100 and Triton X-114 (which were totally ineffective), and 0.05% hexadecyltriethylammonium bromide (which was better than sodium deoxy-

Minutes 0

FIG. 2.

30

post

60

inductlon

90

120

N

Time course of u overproduction from pESC-1 in C600 (pGPl-2). The culture was grown at 30°C until OD, = 0.4 when it was shifted to 42°C. Time points were taken at times after upshift indicated above each lane. Molecular weight marker positions are indicated on the right.

84

GENTRY A

AND

BURGESS

Lysate (IO g whole cells) CdTitUg~ (30,ooo I g, 30’) I I supernatant (soluble fraction) t

I

B 1

wash, sob with 0.25’

JMono-QJ

2

3

4

5

6

7

8

9

10

pure w (US mg)

t

pure 0 (1.5 mg) FIG. 3. Purification of w. (A) Steps of w purification from both the soluble and the insoluble fraction. (B) 16% SDS-polyacrylamide gel showing fractions from individual steps of the w purification. Lane 1, markers; lane 2, whole cell lysate; lane 3, pellet fraction (insoluble fraction); lane 4, Q-Sepharose column (insoluble fraction); lane 5, Red-Sepharose (insoluble fraction); lane 6, Mono-Q (insoluble fraction); lane 7, supernatant fraction (soluble fraction); lane 8, Q-Sepharose (soluble fraction); lane 9, Red-Sepharose (soluble fraction); lane 10, Mono-Q (soluble fraction).

cholate hut not as effective as Sarkosyl). Four molar urea, 4 M GuHCl, and 0.2 M NaOH were slightly less effective than Sarkosyl. Eight molar urea and 6 ‘M GuHCl, not surprisingly, solubilize all the o, nearly all of which reprecipitates upon dialysis. w binds the two triazine dyes most commonly used for protein purification, Procion red and Cibacron blue. For purification, Procion red columns (such as Red-Sepharose) are superior to Cibacron blue columns for two reasons. First, fewer proteins bind Procion red under the conditions used. Second, w quantitatively elutes from Red-Sepharose at lower ionic strength (1 M NaCl as opposed to 2 M NaCl). Both Procion red and Cibacron blue are notorious for binding nucleotide binding proteins (20). Often such nucleotide binding proteins can be removed from dye columns using the nucleotides to which they bind as an eluting agent. w will not be removed from either Cibacron blue or Procion red columns by the following nucleotides: ATP, UTP, CTP, GTP, ADP, GDP, CAMP, NADH, NADPH, S-adenosylmethionine, and ppGpp (all at 10 IrIM). The yield of w from the insoluble fraction is typically 1.5 mg per 10 g wet wt of frozen cells. The yield from the soluble fraction is variable but is often about 0.5 mg. Estimating from Fig. 3, the purity of w purified from the insoluble fraction using the method presented is greater than 95%. o purified from the solubilized fraction is slightly less pure but typically is also at least 95% pure. Proteolytic cleavage of recombinant w. Early attempts at purifying (LI resulted in a very pure peptide which cross-reacted with anti-w antibody but was noticeably smaller than the o found bound to RNA polymerase. Sequence analysis revealed that the purified material matched w starting at position 26 of w (see Fig. 4).

This indicated that w was cleaved by a protease which was not efficiently inhibited by PMSF. The protease cleaved efficiently at an internal Arg-Arg (residues 24 and 25) but not the Arg-Arg at the C-terminus. These characteristics, lack of inhibition by PMSF and the tendency to cleave internal Arg-Arg residues, are diagnostic of the OmpT protease (23). The OmpT protease, identified as protease VII (also known as membrane proteins a and 3b), has been implicated in the cleavage of T7 RNA polymerase (lo), the Ada protein (21), the UvrB protein (3), colicins (5), a P-lactamase-protein A fusion protein (l), recombinant human y interferon (22), and the ferric enterobactin receptor (7). It is an abundant outer membrane protein involved in the maturation of other outer membrane proteins such as ferric enterobactin receptor (7). Interestingly, OmpT expression is extremely temperature dependent and its levels

1

50

+

NH> -~ARVTVODAVEKIGNRFDLVLVAARRAR~~~~GGKDPLVPEENDKTT~I~

0

NH2-RAROMOVGGKDPLVPEENDKTTVlA 2;

fragment

;o

91 LRElEEGLlNNOlLDVREROEOOEQEAAELOAVTAlAEGRk-COOH

LREIEEGLINNQILDVREROEOOEOEAAELOAVTAIAEGRR-COOH

(D

fragment

9’1

FIG. 4. Sequence of w and of its major proteolytic fragment. The sequence of w was taken from Gentry and Burgess (8). The sequence of the major proteolytic fragment, which starts to the immediate right of the arrow, was determined by microsequencing purified protein as described under Materials and Methods. The protein sequence of w depicted in this figure is based on the protein encoded by rpo.Z. Mature w lacks the initiating methionine (8).

PURIFICATION 1

234

OF

w SUBUNIT

OF

56

FIG. 5. Assay for RNA polymerase binding. Pure w was incubated with w-free core enzyme or w-free holoenzyme as described under Materials and Methods. After incubation, Q-Sepharose was added to the mixture under ionic conditions where RNA polymerase, but not w, binds the resin. After washing, the bound RNA polymerase was eluted, dried, and run on a 16% SDS-polyacrylamide gel. The presence of w in the eluate indicates that it was bound to RNA polymerase. Lane 1, w-free core enzyme treated as described above; lane 2, w-free core enzyme incubated with w and treated as described above; lane 3, w-free holoenzyme treated as described above; lane 4, w-free holoenzyme incubated with w and treated as described above; lane 5, w treated as described above; lane 6, untreated w run as a marker.

are greatly reduced in cells grown at 30°C (17). When w is overproduced in strain BL21, which is OmpT-, very little proteolysis of w occurs upon lysis in contrast to overproduction in C600 (Fig. 6). Additionally, using the assay devised by Grodberg and Dunn (lo), ompT+ cells, but not omp T- cells, readily cleave purified w (data not shown). Proteolysis is still apparent at a low level when w is purified in an ompT mutant. We have observed variable amounts of a fragment of identical size to the largest product of the reaction with OmpT as well as another fragment of slightly smaller size. These fragments are easily detected by Western blots. The amount of these fragments is greatly reduced if PMSF and benzamidine are included in all dialysis buffers. w, but not the proteolytic fragment, will bind RNA polymerase. With the binding assay described under Materials and Methods, w purified by the procedure above will bind to RNA polymerase (Fig. 5). The proteolytic fragment does not appear to bind under these conditions (data not shown).

Escherichiu

coli

RNA

85

POLYMERASE

front of rpoZ matches the sequence of the ribosomal binding site region of T7 gene 10. In addition to having a very strong ribosomal binding site, it also contains what has been described as a translational enhancer sequence, termed t (15,16). This sequence has been shown to increase translation of a gene in a semi-orientationindependent manner. The apparent lack of overproduction when T7 RNA polymerase is provided by DE3 is puzzling. A trivial explanation is that pGPl-2 makes more T7 RNA polymerase. While this is certainly true, T7 RNA polymerase is not thought to be limiting in this system (22). Further, quite dramatic overproduction of other proteins is possible using DE3 (22). It is possible that w inhibits T7 RNA polymerase expression so that the higher initial level of T7 RNA polymerase produced by pGPl-2 is necessary for successful w overproduction. Other explanations are possible, however. Purification of w from the strains described here is straightforward. Impediments to devising its purification were its insolubility and its cleavage by OmpT. It was surprising that w was insoluble because it has none of the characteristics which are thought to cause insolubility, namely, disulfides, extensive hydrophobic regions, and many prolines (19). Additionally, small size and high polarity are thought to be characteristics of proteins which will not precipitate when overproduced (19). The behavior of w is evidence of the extreme variability of proteins and our inability to predict how a protein will behave upon overproduction. The cleavage by OmpT, although initially very frustrating, was easily rectified by overproducing w in ompT

6L31 ompT -1

36000

2

CWO -

cunpr+ 3

4

5

-

43mo33000

-

16400

-

14300

-

DISCUSSION The overproduction of w required some manipulations. Simply cloning rpoZ downstream of strong, inducible E. coli promoters did not result in useful overproduction (data not shown) nor did driving transcription from a T7 RNA polymerase promoter where T7 RNA polymerase was provided by either a plasmid or a prophage. This suggested either that sequences upstream inhibited translation or that the ribosomal binding site of rpoZ was not efficient. By replacing the upstream region with an optimal ribosomal binding site both the ribosomal binding site and the possible inhibitory sequences were changed. The oligonucleotide cloned in

-co “(

FIG.

6.

“in (1

“c”1

-

fmgmm

Comparison of proteolysis of w overproduced in C600 BL21 (ompr). Cultures were grown to mid-log phase at 30°C and induced for 1 h at 42°C. Following pelleting and resuspension in lysis buffer, each culture was split. One half was lysed by the addition of SDS sample buffer and the other half of each culture was lysed by passage through a cell disruption bomb two times followed by the addition of SDS sample buffer. Lane 1, molecular weight markers; lanes 2 and 4, lysis by SDS sample buffer only; lanes 3 and 5 lysis by nassaee through the cell disruntion bomb followed bv SDS sample bufferY -

(ompT+) and

86

GENTRY

AND

mutants. We have tried overproducing w in three ompT mutant strains, BL21, CF1528, and UT5600, and found that BL21 was superior due to its very fast growth rate. There is little difference in the amount of overexpression in C600, UT5600, and BL21. Expression in CF1528 is quite a bit lower probably due to its extremely slow growth rate (note the number and type of genetic lesions in CF1528 listed in Table 1). Grodberg and Dunn (10) have recommended the use of BL21 as a host for overexpressing proteins noting that it has, in addition to a high growth rate and being OmpT-, a deficiency in the ion protease which is thought to be responsible for the degradation of some foreign proteins in E. coli. The binding of w to Cibacron blue and Procion red is intriguing because such dyes often bind nucleotide binding proteins. This, coupled with the fact that rpoZ, the gene that encodes o, is in the same operon which also contains SPOT,whose product is involved in the metabolism of ppGpp, prompted us to test if w does indeed bind a nucleotide. Elution from the dye columns using a variety of nucleotides proved unsuccessful. Additionally, o does not bind GTP, GDP, or CAMP agarose (Sigma) at 0.1 M NaCl. w will not bind DNA cellulose, singlestranded DNA agarose, or heparin agarose, making it unlikely that it binds DNA. Even though nearly all nucleotide binding proteins will bind Cibacron blue and a smaller number Procion red, quite a few non-nucleotide binding proteins will bind these dyes, most notably bovine serum albumin (20).

thank Dayle Hager for providing RNA polymerase. Additionwe thank Mark Knuth for his help and advice in purifying w. work supported by NIH Grants GM28575 and CA07175 and Grant CHE-8509625.

REFERENCES 1. Beneyx, 2.

F., and Georgiou, G. (1990) In vivo degradation of secreted fusion proteins by the Escherichiu coli outer membrane protease OmpT. J. Bacterial. 1’72,491-494. Burgess, R. R. (1969) Separation and characterization of the subunits of RNA polymerase. J. Biol. Chem. 244, 2168-2176.

3. Caron, 4.

R., and Grossman, L. (1988) Potential role of proteolysis in the control of uurABC incision. Nucbic Acids Res. 16,964l. Cashel, M., and Rudd, K. E. (1987) The stringent response in “Escherichia coli and Salmonella typhimurium Cellular and Molecular Biology” (Neidhardt, F., Ed.), pp. 1410-1438, Amer. Sot. Microbiol., Washington, DC.

5. Cavard,

D., and Lazdunski, during both entry into and Bacterial. 172, 648-652.

6. Earhart, 7.

C., Lundrigan, M., Pickett, C., and Pierce, J. (1979) Escherichia coli K-12 mutants that lack major outer membrane protein a. FEMS Microbial. L&t. 6, 277-280. Fiss, E., Williams, C., Hollifield, J., and Neilands, J. (1979) Absence of ferric enterobactin receptor modification activity in mutants of Escherichia coli lacking protein a. Biochem. Biophys. Res. Commun. 91,29-34.

8. Gentry, 9.

D. R., and Burgess, R. R. (1986) The cloning and sequence of the gene encoding the omega subunit of Escherichia coli RNA polymerase. Gene 48,33-40. Gentry, D. R., and Burgess, R. R. (1989) rpoZ, encoding the omega subunit of Escherichia coli RNA polymerase, is in the same operon as SPOT. J. Bacterial. 171,1271-1277.

10. Grodberg,

J., and Dunn, J. (1988) ompT encodes the Escherichia coli outer membrane protease that cleaves T7 RNA polymerase during purification. J. Bacterial. 1’70, 1245-1253.

11. Hager,

D., Jin, D., and resolution ion exchange and active Escherichia press.

12. Helmann, of bacterial

Burgess, R. (1990) Use of Mono& chromatography to obtain highly coli RNA polymerase. Biochemistry,

high pure in

J., and Chamberlin, M. (1988) Structure and function sigma factors. Annu. Reu. Biochem. 54, 171-204.

13. Igarashi,

Promoter selecfactor is respon-

14.

Genetics,” NY.

K., Fujita, N., and Ishihama, A. (1989) tivity of Escherichia coli RNA polymerase: Omega sible for the ppGpp sensitivity. Nucleic Acids Res. Miller, J. H. (1972) “Experiments in Molecular Spring Harbor Laboratory, Cold Spring Harbor,

15. Olins,

P., and Rangwala, rived from bacteriophage lation of the lacZ gene 16,973-16,976.

17,8755-8765. Cold

S. (1989) A novel sequence element deT7 mRNA acts as an enhancer of transof Escherichia coli. J. Biol. Chem. 264,

16. Olins,

P., Devine, C., Rangwala, S., and Kavka, K. (1988) The T7 gene 10 leader RNA, a ribosome-binding site that dramatically enhances the expression of foreign genes in Escherichia coli. Gene

73,227-235. 17. Rupprecht, K., Gordon,

ACKNOWLEDGMENTS We ally, This NSF

BURGESS

C. (1990) Colicin cleavage by OmpT release from Escherichia coli cells. J.

G., Lundrigan, M., Gayda, R., Markovitz, A., and Earhart, C. (1983) ompT: Escherichia coli K-12 structural gene for protein a (3b). J. Bacterial. 153,1104-1106. 18. Sarubbi, E., Rudd, K., Xiao, H., Ikehara, K., Kalman, M., and Cashel, M. (1989) Characterization of the spoT gene of Escherichia coli. J. Biol. Chem. 264, 15,074-15,082. 19. Schein, C. (1989) Production of soluble recombinant proteins in bacteria. BiolTechnology 7, 1141-1149.

20. Scopes, 21.

R. (1987) “Protein Purification,” 2nd ed., Springer-Verlag, New York. Sedgwick, B. (1989) In vitro proteolytic cleavage of the Escherichia coli Ada protein by the ompT gene product. J. Bacterial.

171,2249-2251. 22. Studier, F. W., and Moffatt, 23.

24.

B. (1986) Use of bacteriophage T7 RNA polymerase to direct selective high level expression of cloned genes. J. Mol. Biol. 189, 113-130. Sugimura, K., andNishihata, T. (1988) Purification, characterization, and primary structure of Escherichia coli protease VII with specificity for paired basic residues: Identity of protease VII and OmpT. J. Bacterial. 170, 5625-5632. Tabor, S., and Richardson, C. (1985) A bacteriophage T7 RNA polymerase/promoter system for controlled exclusive expression of specific genes. Proc. Natl. Acad. Sci. USA 82, 1074-1078.