Pii: s1050-3862(98)00007-2

Genetic Analysis: Biomolecular Engineering Viability of E. coli cells containing phage RNA polymerase and promoter: interference of plasmid replication by transcription Young-Soo Kwon, Jinsuk Kim, Changwon Kang * Department of Biological Sciences, Korea Ad6anced Institute of Science and Technology, Taejon 305-701, South Korea Received 2 January 1998; received in revised form 8 June 1998; accepted 17 June 1998 Abstract
Strong transcription of phage promoters often renders the host E. coli cells containing the phage RNA polymerase inviable.
When expression of the phage SP6 RNA polymerase gene in one plasmid was induced in the E. coli JM109 cells, cells that bearan active SP6 promoter were inviable. When it was not induced (the polymerase was still produced in low level), viability of thehost cells and stability of the promoter-bearing plasmids depended on the orientation of the promoter with respect to that of thereplication origin and on the sequence of the origin. A group of SP6 promoter-bearing plasmids (group I plasmids) that had thepromoter directed towards the ColE1 replication origin, rendered the polymerase-containing host cells inviable in selective media.
When the sequence of the origin was different (group II plasmids), this adverse effect was not observed. When the promoterdirection was same as the replication origin and the ampicillin-resistant gene (group III plasmids), many satellites formed aroundthe colonies on ampicillin-containing agar plates. These effects were caused by strong transcription of the phage SP6 promoter byits RNA polymerase, since they were reduced or eliminated by inserting an active terminator just downstream of the promoter.
The viability of host cells and copy number of the promoter/terminator-bearing plasmids appear to be quantitatively related withefficiency of initiation and termination of the phage transcription. These systems may be useful for in vivo screening for mutantvariants of the phage promoter, polymerase and terminator that are affected in their efficiency. 1998 Elsevier Science B.V. Allrights reserved.
Keywords: Cell viability; Phage RNA polymerase; Phage promoter; ColE1 replication; Plasmid curing 1. Introduction
been reported, however, that some plasmids carrying aT7 promoter are unstable in E. coli cells producing the Cloning of a gene in E. coli cells sometimes evokes phage T7 RNA polymerase [1,2]. It was suggested that frustration, when an expression product of the cloned the highly efficient transcription with the T7 RNA gene is harmful, or toxic to the host cells. Also, it is polymerase/promoter system might use up the reservoir sometimes hard to clone a small piece of DNA such as of the ribonucleoside triphosphates [3,4], probably due strong transcription promoters. The bacteriophage T7, to multiple rounds of transcription around the entire T3 and SP6 RNA polymerases of single subunit tran- scribe their promoters with high processivity and speed.
It is also possible that strong transcription interferes Thus, the phage transcription systems have been widely with replication of the plasmids. The cloning of a used for over-expression of genes in E. coli cells. It has strong promoter into an E. coli plasmid vector hasoften been inefficient. Such strong promoters as thoseof the bacteriophage T5 and of Streptococcus pneumo- * Corresponding author. Present address: Professor C. Kang, c/o niae could only be cloned in conjunction with strong Sequenom, Incorporated, 11555 Sorrento Valley Road, Suite ‘C’, SanDiego, CA 92121, USA. Tel: terminators [6,7]. It was suggested that excessive tran- 619 3500344; e-mail: ckang@sorak.kaist.ac.kr scription into the replication origin may result in en- 1050-3862/98/$ - see front matter 1998 Elsevier Science B.V. All rights reserved.
PII S1050-3862(98)00007-2 Y.-S. Kwon et al. / Genetic Analysis: Biomolecular Engineering 14 (1998) 133 – 139 hanced levels of Rom and RNA I and thereby decrease When needed, ampicillin, tetracycline or isopropyl-i- in the plasmid copy number [8,9]. Replication of thiogalactoside (IPTG) was added to LB medium to the ColE1-type plasmids is mainly regulated by two com- final concentration of 100 vg/ml, 25 vg/ml or 0.1 mM, plementary RNA species, RNA I and RNA II, encoded respectively. Liquid cultures were grown in a shaking by upstream regions of the replication origin. A part of the RNA II transcript forms a stable hybrid with thetemplate DNA near the origin [10 – 12]. Then, RNase 2.2. Determination of plasmid copy numbers and H-mediated cleavage of the hybrids at the origin yields a primer for the initiation of unidirectional DNA syn-thesis [10,13]. On the other hand, RNA I can form a Plasmid copy number was determined as previously complex with RNA II and prevents it from serving as a described [20] with minor modifications. The cell cul- In this study such interference of plasmid replication stop solution (5% phenol in ethanol) and centrifuged in by strong SP6 transcription has been observed in a a microcentrifuge tube at 12000 rpm for 3 min. The group of SP6 promoter-bearing plasmids, derived from pellets were washed with 0.2 ml of Mg2+-free H1 the ColE1 replicon, in the presence of the phage RNA minimal media and re-centrifuged. Then, the procedure polymerase. This interference resulted in curing, or previously described by Lin-Chao et al. (1986) was reduction in the copy number of the promoter-bearing plasmids. It was also demonstrated that this interfer- per well and electrophoresed with 0.8% agarose gel in ence is relieved by insertion of an effective terminator Tris – acetate buffer (4.84 g Tris, 1.14 ml glacial aceticacid and 2 ml of 0.5 M EDTA per liter). After the gels to the promoter-bearing plasmids in an appropriate were stained with ethidium bromide and destained in location. Furthermore, the copy number of the pro- water, DNA bands were photographed using Polaroid moter/terminator-bearing plasmids appears to reflect at film of type 667. The negative image of the agarose gel least semi-quantitatively the efficiency of initiation and picture was scanned and the bands were quantified with termination of the SP6 transcription.
the aid of ImageQuant Version 3.3 (Molecular Dynam-ics). When needed, plasmids were prepared by thealkaline lysis method as previously described [21].
2. Materials and methods
3. Results and discussion
The plasmids pSP64, pGEM3, pGEM4, pGEM3Z, 3.1. Viability of E. coli cells that contain both phage pGEM4Z, pGEMEX-1 and pGEMEX-2 were available from Promega. They all have the same SP6 promoter.
The plasmids that have a suffix ‘S’ in their names, Viability of E. coli cells containing various SP6 pro- moter-bearing plasmids in the presence of the SP6 structed to contain a copy of the phage T7 terminator RNA polymerase was tested. The polymerase gene was Tf in the direction of SP6 transcription. The termina- inserted in the tetracycline-resistant plasmid pA- tor-containing BamHI/BglII fragment of pET3 [16] was CYC184 that has the replication origin of p15A, result- inserted into the BamHI site of the parent plasmids, ing in pACSP6R [17]. Then, a second plasmid bearing pGEM3, pGEM3Z and pGEM4Z in the direction and an SP6 promoter was introduced into JM109 cells that downstream of the SP6 promoter. The plasmid already contained the pACSP6R. Various commercially pKSP6CAT [17] was previously constructed by insert- available, SP6 promoter-bearing plasmids, pSP64, ing an SP6 promoter-containing synthetic oligonucle- otides into the SmaI/BamHI site of pKK232-8 [18].
and pGEMEX-2 were tested. They all have an ampi- These SP6 promoter-bearing plasmids were intro- cillin-resistant gene and the replication origin of ColE1 duced into E. coli JM109 cells that already contained and are compatible with p15A. The transformants were the plasmid pACSP6R [17] by the standard transforma- spread on agar plates containing both ampicillin and tion method previously described [19]. The plasmid pACSP6R was previously constructed by inserting the When the SP6 RNA polymerase gene under the SP6 RNA polymerase gene at the P6uII/ScaI site of control of E. coli promoter lacUV5 was induced with pACYC184 downstream from the lac promoter. The E.
IPTG, all the cells containing the above plasmids sepa- coli JM109(DE3) cells contain the gene for phage T7 rately did not survive at all in the selective media with RNA polymerase in the chromosome as previously ampicillin and tetracycline (Table 1). Similar phenom- ena were reported for the phage T7 expression system Y.-S. Kwon et al. / Genetic Analysis: Biomolecular Engineering 14 (1998) 133 – 139 Table 1Three groups of SP6 promoter-bearing plasmids affecting viability of their host cells differently Competent E. coli JM109/pACSP6R cells containing the SP6 RNA polymerase gene under an IPTG-inducible promoter were transformed witheach plasmid and spread on agar plates with LB, ampicillin and tetracycline (without IPTG) or those with LB, ampicillin and IPTG (with IPTG).
with a single copy of the polymerase gene in the E. coli tion of replication. Thus, convergent transcription chromosome [1], or its multiple copies in plasmid [2].
would reduce the plasmid copy number by interfering Thus, strong expression of the phage RNA polymerase with the production of RNA II or by increasing the gene makes the host cells that contain an SP6 promoter inviable under selective pressure. These phenomena Likewise, the SP6 transcription of the group I plas- have been utilized for screening inactive variants of the mids would produce long RNA that is antisense also to the ampicillin resistant gene (Fig. 1). Thus, one can They were also tested in the absence of IPTG induc- argue that suppression of i-lactamase gene expression tion, in which case the SP6 RNA polymerase was still could have caused inviability of the cells in the presence produced in much less quantity (Table 1). Without the of ampicillin. This did not appear to be the case, IPTG induction, a group of plasmids, pSP64, pGEM3 however, because the host cells containing the group II and pGEM4, still rendered the host cells inviable in a selective medium with ampicillin and tetracycline were viable (Table 1), although the orientation of the (group I). However, a second group of plasmids, promoter was the same as in the group I plasmids (Fig.
pGEM4Z, pGEMEX-1 and pGEMEX-2 did not render 1). The group II plasmids could also produce the the host cells inviable (group II) in the selective media.
anti-lactamase RNA but did not render the host cells On the other hand, the presence of plasmid pGEM3Z inviable under the same conditions. The difference be- resulted in formation of many small satellite colonies tween the group I and II plasmids lies in the sequence under selective pressure (group III).
of replication origin. The fourth base upstream of the The difference in the viability of the above promoter- replication start site is C in the group I plasmids (and containing cells without induction of the polymerase pBR322) and G in the group II plasmids. The one base gene expression can be explained by the difference in difference in RNA transcripts could produce different the extent of interference of plasmid replication by secondary structures, as predicted by various RNA transcription, as previously suggested for the E. coli folding programs (data not shown). The alternative system [8,9]. The three groups differ in the orientation secondary structure could impair the inhibitory anti- of the phage promoter versus replication origin and in the sequence of the replication origin (Fig. 1). In the The group III plasmid pGEM3Z has an SP6 pro- group I plasmids (pGEM3, pGEM4 and pSP64), the moter in the same orientation as the replication origin.
direction of the SP6 transcription is opposite to that of When the transformants of JM109/pACSP6R with DNA replication. In such cases the SP6 RNA poly- pGEM3Z (cultured without IPTG induction) were merase can read through the replication origin in the spread on agar plates with both antibiotics, numerous opposite orientation. Around the ColE1 origin RNA II satellites grew around small colonies (Table 1). This is produced by transcription in the same direction as plasmid has the same sequence of replication origin as the replication and serves as a primer for the DNA the group II plasmids. A derivative of a group I plas- synthesis. On the other hand, the inhibitory RNA I is mid, pSP64, was also constructed, where the SP6 produced by transcription in the opposite orientation promoter was inverted. The corresponding pSP64- and plays the role of a repressor of replication. The derivative (pSP64der) transformants also formed satel- RNA I sequence, contained in long transcripts pro- lites on selective agar plates (Table 1), while the pSP64 duced from SP6 transcription directed towards the transformants did not form any colony. Thus, regard- replication, might still be capable of repressing initia- less of the sequence of replication origin, when Y.-S. Kwon et al. / Genetic Analysis: Biomolecular Engineering 14 (1998) 133 – 139 Fig. 1. Three different groups of the SP6 promoter-bearing plasmids. All the plasmids have the ampicillin-resistant gene (Amp) in the samedirection as the replication origin (Ori and Ori%). The Ori is the same as the origin of pBR322, but different from Ori% of the group II plasmids.
An SP6 promoter (SP6) is directed towards the Ori and Ori% in the group I and II plasmids and away from the Ori (and Ori%) in group III plasmids.
the SP6 promoter is in the same orientation with the promoter and the replication origin. The observation replication origin, such plasmid-containing transfor- with pKSP6CAT was the same as the case with mants cultured without IPTG induction form satellites on selective agar plates. Since the i-lactamase gene is As the group II plasmid pGEM4Z did not render the located in the same direction and downstream of the host cells inviable on selective agar plates, its derivative SP6 promoter in the group III plasmids (Fig. 1), even pGEM4ZS that has the terminator did not either low level of the SP6 RNA polymerase (produced in the (Table 2). The transformants formed hundreds of absence of IPTG induction) increases expression level colonies on selective agar plates. On the other hand, of the ampicillin-resistant gene. This would lead to many satellites that formed around the colonies con- deprivation of ampicillin in the surrounding areas of taining the group III plasmid pGEM3Z disappeared, initially formed colonies on the agar plates and thus when a terminator was inserted (pGEM3ZS, Table 2).
allow satellites to form around them.
It is likely that due to efficient termination, the SP6transcription did not extend into the i-lactamase gene.
3.2. Effects of a terminator on 6iability of host cells Furthermore, the insertion of a terminator in the promoter-bearing plasmids rendered the host cells vi- The group I plasmid-containing transformants of able on selective agar plates, even when the transfor- JM109/pACSP6R cultured without IPTG induction did mants were cultured with IPTG induction of the not form colonies on agar plates with antibiotics. If it is polymerase gene expression (Table 2). The colony sizes caused by the SP6 transcription going around the plas- were smaller than those from uninduced cultures were.
mid, inserting an effective terminator to the plasmid The terminator-inserted group III plasmid, pGEM3ZS downstream of the promoter would eliminate or reduce caused satellite formation when the transformants were the problem. Therefore, a copy of the phage T7 termi- cultured with IPTG induction (Table 2). It was proba- nator Tf was inserted downstream of the SP6 promoter bly caused by the same reason described for the termi- in one plasmid of each of the three groups (Table 2).
nator-lacking counterpart, pGEM3Z without IPTG The terminator Tf was previously found to terminate the transcription of SP6 RNA polymerase in vitro [22].
The plasmid pGEM3S has the terminator inserted in 3.3. Selecti6e loss of promoter-bearing plasmids the group I plasmid pGEM3. (The suffix S indicatesthat the terminator is inserted in the direction of the The E. coli JM109/pACSP6R cells that were trans- SP6 transcription rather than that of the T7 transcrip- formed with the group I and II plasmids (pGEM3 and pGEM4Z, respectively) and their terminator-inserted JM109/pACSP6R cultured without IPTG induction derivatives (pGEM3S and pGEM4ZS, respectively) now formed regular colonies on selective agar plates were tested for their sustained viability on selective (Table 2). This suggests that the terminator effectively media with only one of the two antibiotics, tetracycline suppresses the SP6 transcription from extending into and ampicillin. The polymerase gene-bearing plasmid the replication origin. It was tested with another termi- pACSP6R has the tetracycline-resistant gene and the nator. The plasmid pKK232-8 has two sets of E. coli other plasmids containing promoter/terminator have rrnB terminators T1 and T2 [18]. Since the SP6 RNA the ampicillin-resistant gene. Thus, the assay identifies polymerase recognizes the terminator T1 [23], an SP6 which plasmid remains stable in the cells. All the trans- promoter was inserted upstream of the chloramphenicol formants that were cultured in LB broth for 45 min acetyltransferase gene in pKK232-8. The resulting plas- were streaked and grown on agar plates containing mid pKSP6CAT [17] has the terminators between the both ampicillin and tetracycline (transformation plates) Y.-S. Kwon et al. / Genetic Analysis: Biomolecular Engineering 14 (1998) 133 – 139 Table 2Viability of the E. coli cells that contain the terminator-inserted derivatives of the group I, II and III plasmids All the plasmids, except pKSP6CAT, contain a copy of the T7 terminator Tf just downstream of and in the same direction as the SP6 promoter.
Thus, it is located between the promoter and Ori/Ori’ in the group I-T and II-T plasmids, while between the promoter and the ampicillin resistantgene in the group III-T plasmids. The plasmid pKSP6CAT [17] was constructed by inserting an SP6 promoter in the polycloning site ofpKK232-8. Thus, it has two sets of E. coli rrnB terminators T1 and T2 downstream of and in the same direction as the SP6 promoter. CompetentE. coli JM109/pACSP6R cells were transformed as described in Table 1.
for 24 h. Then, 100 colonies were randomly picked ampicillin plates, while the transformants with its termi- from each transformation plate and duplicated on agar nator-inserted derivative pGEM4ZS all survived on the plates with LB media only or those with LB plus IPTG selection plates (Table 3). Strong transcription by en- (growth plates) using toothpicks. After they were grown hanced copies of the SP6 RNA polymerase must have for 24 h again, each colony on the growth plates was cured the promoter-bearing plasmid selectively over the duplicated on agar plates with LB plus ampicillin or polymerase-producing plasmid. Thus, strong SP6 tran- those with LB plus tetracycline (selection plates).
scription impairs the replication of the plasmid that is From the transformation culture with the group I being transcribed by the SP6 RNA polymerase. It plasmid pGEM3, no colonies were shown on the first argues against a possibility that the inviability might transformation plates that contained ampicillin and have resulted from general constraints imposed on tetracycline, as described above. On the other hand, all metabolism or growth by strong transcription due to the other three transformants formed hundreds of enhanced level of the phage RNA polymerase.
colonies on the transformation plates and they wereassayed further (Table 3). All the transformant colonies 3.4. Copy number of the group II plasmids that were transferred to the tetracycline selection platewere viable. Thus, the polymerase gene-bearing plasmid Based on all the above results the group II plasmids pACSP6R was not cured at all in any of the six cases as (for example, pGEM4Z) appear to provide an excellent shown in Table 3. On the other hand, the numbers of system for in vivo screening of active or inactive pro- colonies formed on the ampicillin selection plates were moters, terminators and polymerases of the phage SP6.
The group II plasmids are stable in the absence of the While transformants could not be obtained with IPTG induction, whether they have an effective termi- pGEM3, its terminator-inserted derivative pGEM3S nator or not. When the transformant cultures are in- transformed the polymerase gene-containing cells and duced by IPTG, the terminator-lacking plasmids render 92 and 83% of the transformants were viable on the the host cells inviable on selective plates. This is due to ampicillin selection plates after growing in the absence loss of promoter-bearing plasmids resulting from active and presence of IPTG, respectively (Table 3). Also SP6 transcription. This was reconfirmed by measuring when the SP6 promoter was inverted in another group I plasmid pSP64, the plasmid remained stable in The copy number of pGEM4Z in JM109 cells with- JM109/pACSP6R cells, whereas pSP64 itself was very out the SP6 RNA polymerase gene and that of a unstable. Thus, the inviability of the group I plasmid- containing cells was due to the loss of the promoter- pACSP6R cells with the polymerase gene were both bearing plasmid and it can be prevented by insertion of measured to be about 400 per cell in the presence of an effective transcription terminator or by inversion of IPTG induction. However, the copy number of pGEM4Z in JM109/pACSP6R decreased to 72 per cell, The group II plasmid-containing cells that grew in by 82% (Table 4). This reduction was also observed the absence of IPTG were viable on the ampicillin with the phage T7 promoter-polymerase system. The plates, regardless whether the plasmid had a terminator plasmid pGEM3Z has a T7 promoter in the opposite or not (100 and 98% of the cells transformed with orientation of the SP6 promoter and thus resembles the pGEM4ZS and pGEM4Z, respectively, were alive). Af- group II plasmids of SP6. The number of pGEM3Z in ter growing in the presence of IPTG, however, the JM109 cells was again about 400 per cell. When it was pGEM4Z transformants formed no colonies on the introduced to JM109(DE3) cells that contained the T7 Y.-S. Kwon et al. / Genetic Analysis: Biomolecular Engineering 14 (1998) 133 – 139 Table 3Assay for selective curing of the promoter-bearing plasmids The JM109/pACSP6R cells were transformed with the group I and II plasmids and their terminator-inserted derivatives (group I-T and II-T). Thetransformants were obtained individually on the agar plates with both ampicillin (amp) and tetracycline (tc). Then, 100 colonies were selected fromeach case and grown on agar plates with a growth medium, either LB alone or LB plus IPTG. Each colony on the growth plates was transferredto agar plates with a selection medium, either LB plus amp or LB plus tc. The numbers of grown colonies out of the 100 were averaged fromthree sets of experiments.
a NT indicates that no transformants could be obtained.
Table 4Copy numbers of the group II plasmids per JM109 transformant cell RNA polymerase gene in the chromosome under an screening for the terminator variants of higher and IPTG-inducible promoter, its number dropped to 5 per lower efficiency and for the polymerase mutants that cell (99% disappeared) (Table 4). Also when a lac are affected in termination efficiency.
promoter-containing P6uII fragment was inverted inpUC19, the copy number decreased by 70% in thepresence of IPTG, compared with the case without the Acknowledgements
induction. Thus, the loss of promoter-bearing plasmidsin the presence of strong transcription can be observed This work was supported by grants from the Repub- with not only the SP6 but also the T7 and E. coli RNA lic of Korea Ministry of Education (Genetic Engineer- ing Program) and Korea Advanced Institute of Science In the case where an active terminator (Tf) was inserted to the group II plasmid (pGEM4Z) at anappropriate location, the copy number of pGEM4ZS inthe IPTG-induced JM109/pACSP6R cells increased References
back to 283 per cell, which was about 71% of its copynumber in JM109 cells (Table 4). Thus, the presence of [1] Schneider TD, Stormo GD. Nucleic Acids Res 1989;17:659 – 74.
an active terminator saves the plasmid even in the cells [2] Ikeda RA, Ligman CM, Warshamana S. Nucleic Acids Res where the polymerase was amplified. The incomplete recovery may reflect limited efficiency of the T7 termi- nator Tf in vivo for the SP6 RNA polymerase. Its [4] Studier FW, Moffatt BA. J Mol Biol 1986;189:113 – 30.
efficiency has been measured in vitro to be 50 – 60% for [5] McAllister WT, Morris C, Rosenberg AH, Studier FW. J Mol the SP6 polymerase [22] and 70 – 80% for the T7 poly- merase [24,25]. Although the termination efficiency [6] Gentz R, Langner A, Chang AC, Cohen SN, Bujard H. Proc might be higher in vivo than in vitro [26], it is likely to Natl Acad Sci USA 1981;78:4936 – 40.
be less than 100% in vivo. It is tempting to suggest that [7] Rhee DK, Morrison DA. J Bacteriol 1988;170:630 – 7.
[8] Stueber D, Bujard H. EMBO J 1982;1:1399 – 404.
the 65% recovery ((283 – 72)/(400 – 72)) reflects the Tf’s [9] Dillard JP, Yother J. J Bacteriol 1991;173:5105 – 9.
in vivo efficiency of 60 – 70% for the SP6 RNA poly- [10] Itoh T, Tomizawa J. Proc Natl Acad Sci USA 1980;77:2450 – 4.
merase. At least this assay should be useful for in. vivo [11] Masukata H, Tomizawa J. Cell 1984;36:513 – 22.
Y.-S. Kwon et al. / Genetic Analysis: Biomolecular Engineering 14 (1998) 133 – 139 [12] Masukata H, Tomizawa J. Cell 1986;44:125 – 36.
[20] Lin-Chao S, Bremer H. Mol Gen Genet 1986;203:143 – 9.
[13] Selzer G, Tomizawa J. Proc Natl Acad Sci USA 1982;79:7082 – 6.
[21] Sambrook J, Fritsch EF, Maniatis T. editors. Molecular cloning: [14] Tomizawa J, Itoh T. Proc Natl Acad Sci USA 1981;78:6096 – a laboratory manual. Cold Spring Harbor, NY, Cold Spring [15] Tomizawa J. Cell 1984;38:861 – 70.
[22] Kim S, Kang C. Korean Biochem J 1988;21:389 – 95.
[16] Rosenberg AH, Lade BN, Chui D, Lin S, Dunn JJ, Studier FW.
[23] Kwon Y-S, Lee JT, Kang C. Recent Prog Mol Biol Genet Eng [17] Jeong W, Kang C. Biochem Mol Biol Int 1997;42:711 – 6.
[24] McAllister WT, McCarron RT. Virology 1977;82:288 – 98.
[18] Brosius J. Gene 1984;27:151 – 60.
[25] Carter AD, Morris CE, McAllister WT. J Virol 1981;37:636 – 42.
[19] Hanahan D. J Mol Biol 1983;166:557 – 80.
[26] Yoo J, Kang C. Mol Cells 1996;6:352 – 8.

Source: http://biochem.kaist.ac.kr/publications/pdf/research_47.pdf