J. Mol. Biol. (1990) 215, 31-39

Identification of a Region of the Bacteriophage T3 and T7 R N A Polymerases that Determines Promoter Specificity Keith E. Joho~f, Lyndon B. Gross, Nancy J. McGraw Curtis Raskin and William T. McAllister$ Department of Microbiology Morse Institute of Molecular Genetics S U N Y-Health Science Center at Brooklyn 450 Clarkson Avenue, Box 44 Brooklyn, N Y 11203-2098, U.S.A. (Received 29 December 1989; accepted 16 April 1990) Bacteriophages T7 and T3 encode DNA-dependent RNA polymerases that are 82?/0 homologous, yet exhibit a high degree of specificity for their own promoters. A region of the RNA polymerase gene (gene 1) that is responsible for this specificity has been localized using two approaches. First, the RNA polymerase genes of recombinant T7XT3 phage that had been generated in other laboratories in studies of phage polymerase specificity were characterized by restriction enzyme mapping. This approach localized the region that determines promoter specificity to the 3' end of the polymerase gene, corresponding to the carboxyl end of the polymerase protein distal to amino acid 623. To define more closely the region of promoter specificity, a series of hybrid T7fr3 RNA polymerase genes was constructed by in vitro manipulation of the cloned genes. The specificity of the resulting hybrid RNA polymerases in vitro and in vivo indicates that an interval of the polymerase that spans amino acids 674 to 752 (the 674 to 752 interval) contains the primary determinant of promoter preference. Within this interval, the amino acid sequences of the T3 and T7 enzymes differ at only 11 out of 79 positions. It has been shown elsewhere that specific recognition of T3 and T7 promoters depends largely upon base-pairs in the region from - l 0 to - 1 2 . An analysis of the preference of the hybrid RNA polymerases for synthetic T7 promoter mutants indicates that the 674 to 752 interval is involved in identifying this region of the promoter, and suggests that another domain of the polymerase (which has not yet been identified) may be involved in identifying other positions where the two consensus promoter sequences differ {most notably at position -15).

specificity; the T3 RNA polymerase will not efficiently utilize a T7 promoter, nor will the T7 RNA polymerase efficiently utilize a T3 promoter. The consensus promoter sequences for the T3 and T7 RNA polymerases differ significantly only over the three base-pair region from - 10 to - 12 within a common 23 base-pair sequence (Klement et al., 1990, and references therein). In studies utihzing a series of synthetic T7 promoter variants it has been shown that the base-pairs in this region (and in particular the base-pair at - 1 1 ) are the primary determinants of promoter specificity (Klement et al., 1990). Changing the base-pair at - 1 1 in the T7 consensus sequence to the corresponding T3 residue prevents utilization of the variant T7 promoter by the T7 RNA polymerase and simultaneously enables transcription by the T3 RNA polymerase. This result suggests that relatively minor differences

1. Introduction Bacteriophages T3 and T7 each encode DNA-dependent RNA polymerases that consist of a single subunit polypeptide of ~ 98,000 M r (Chamberlin et al., 1970; Dunn et al., 1971). The genes that encode these RNA polymerases have been cloned and sequenced and it has been found that the predicted amino acid sequences of the proteins exhibit 82% identity (Fig. 1). Despite this high degree of structural homology the phage RNA polymerases exhibit strikingly different promoter 1"Present address: Carnegie Institution of Washington, Department of Embryology, 115 W. University Parkway, Baltimore, MD 21210, U.S.A. :~Author to whom all correspondence should be addressed. 0022-2836/90/170031-09 $03.00/0

31

~) 1990 AcademicPress Limited

32

K . E . Joho et al.

T7 T3

I0 20 30 40 50 60 70 80 90 MNT~-NIAKNDF~DIELAAIPFNTLADH~GERLAREQLAL~ESYEMGEARFRKMFERQLKAGEVADNAAAKPLITTLLPKMIARINDWFEEVKAK~GK~ i e e e sA K l R L 1 A i la lit Ve i YAS K rK

T7 T3

I00 110 120 130 140 150 160 170 180 190 PTA~FL~EIKPEAVAY~T~KTTLA~LTSADNTTVQAVASAIGRAIEDEARFGRIRDLEAKHFKKNVEEQLNKRVGHVYKKAFMQVVEADM~SKGLLGGE s yap L1 S f I vi s tnM i a gMl k h H q £gr

T7 T3

200 210 220 230 240 250 260 270 280 290 AWSS~KEDSI~GVRCIEMLIE~TGMV~L~RQNAG~GQD~ETIELAPEYAEAIATRAGALAGI~PMFQPCVVPPKPWTGI~GGGYWANG~RPLALVRT d tim i L le q h Na S S alq q vdvl k va 300

T7 T3

310

320

330

340

350

360

370

380

390

~SKKALMRYEDVY~PEVYKAINIA~NTAW~INKKVLAVAN~ITKWK~CPVEDIPAIEREELPMKPEDIDMNPEAL~A~K`RAAAAVYRKDKARK~RRISLE g

v 1

V E vn

n

a

sl

q

P

d

TEa

Ke

K

gl

L

V

T7 T3

400 410 420 430 440 450 460 470 480 490 FMLEQANKFANHKAIWFPYNMDWRGRVYAVSMFNPQGNDMTKGLLTLAKGKP IGKEGYYWLKIHGANCAGVDKVPFPERIKFI EENHENIHACAKSPLEN sk p e f A khVdd 1 d in

T7 T3

500 510 520 530 540 550 HpGI 570 580 590 TWWAEQDSPFCFLAFCFEYAGVQHHGLSYNCSL~ LAFDGSCSGIQHFSAMLRDEVGGRAVNLLP SETVQDIYGIVAKKVNEIL~ADAINGTDNEV~,TVTD T q Kq P ml

T7 T3

600 610 620 630 640 650 ENTGEI S E K V K L G T K A L A G Q W L A Y G V T R S V T K R S V M T L A Y G S K E F G F R Q Q V L E D T I kd 1 st Q d

T7 T3

700 710 720 730 740 AvGT 760 770 780 790 K S A A K L L A A E V K D K K T G E I L R K R C A V H W V T P D G F P V W Q E Y K K P I Q T R L N L M F L G Q F R L Q P TINTNKDSEI D A H K Q E S G I A P N F V H S Q D G S R L R K T V V W A H K k t r 1 k dmi m y L g

T7 T3

800 820 830 840 850 860 870 EKYGIESFALI H D S F G T I P A D A A N L F K A V R E T M V D T Y E S C D V L A D F Y D Q F A D Q L H E S Q L D K M P A L P A K G N L N L R D I gk I nN s t p K q

8sflrr

660 670 680 690 QPAI D S G K G L M F T Q P N Q ~ G Y M A K L I W E S V S V T W A A V E A M N W L da

Hind117 880 LESDFAFA k

Figure 1. Comparison of the T7 and T3 RNA polymerases. The amino acid sequence of the T7 RNA polymerase is shown in the upper line (Moffatt el aL, 1984). The corresponding amino acid residue in the T3 RNA polymerase is shown below the line only if it differs from the T7 residue; upper and lower case letters indicate non-conservative or conservative substitutions, respectively (McGraw et al., 1985). The relative positions of the restriction sites used in the construction of hybrid RNA polymerase genes are indicated in bold-faced type (see Fig. 3). in enzyme structure could be responsible for altered promoter recognition, and that the specificity-determining region of the enzyme might be localized to a small region of the protein. Previous attempts to identify the region(s) of the phage RNA polymerase that determines promoter specificity utilized classical genetic techniques to generate T7XT3 recombinant phage in which crossover events had occurred within the RNA polymerase gene (gene 1). Genetic analyses of the recombinants suggested that the sequences involved in promoter specificity are located in the 3' (carboxyl-terminal) end of the gene, and are likely to lie between 0-70 amd 0-78 of gene length (Beier & Hausmann, 1974; Hausmann & Tomkiewicz, 1976; Beier et al., 1977). Subsequent analyses of similar recombinants by restriction-enzyme mapping also indicated that the primary determinant of promoter specificity is located in the C-terminal one-third of the gene, but suggested that a second region that modulates the degree of specificity maps between 0"25 and 0"59 gene length (Ryan & MeConnell, 1982). The analyses of the hybrid phage genomes were complicated by the observation that multiple crossover events appeared to have occurred within the polymerase gene in many of the recombinants (Ryan & MeConnell, 1982). It was not clear from those studies whether polymerase specificity might involve two separate regions of the protein, or whether multiple rearrangements of the structural gene are necessary to maintain the functional integrity of a hybrid enzyme, or the viability of the recombinant phage. It is now known that the phage R N A polymerases are multifunetional enzymes that are involved not only in transcription of phage

DNA, but also in DNA replication (Hinkle, 1980; Saito et al., 1980), transport of the phage DNA into the cell (Zavriev & Shamykin, 1982; Moffatt & Studier, 1988), interaction with phage lysozyme (Moffatt & Studier, 1987), and packaging of the DNA into phage particles (Dunn & Studier, personal communication). If the RNA polymerase interacts with other proteins in a phage-specific manner in performing these functions, rearrangements elsewhere in the phage genome during the generation of the T3XT7 recombinants could place additional constraints on the nature of the polymerase genes in the hybrid phages. These considerations tend to obscure the interpretation of data obtained with recombinant phage RNA polymerase genes. To localize further the specificity region, and to ensure that no hidden mutations or crossovers were present, we constructed a series of hybrid RNA polymerase genes by in vitro manipulation of the cloned structural genes. In these hybrids, various regions of the T7 gene were replaced by the corresponding portion of the T3 gene. Analysis of the properties of the resulting T7/T3 enzymes has permitted the identification of a 79 amino acid region that is involved in promoter specificity. Within this region there are only l l amino acid substitutions between the T3 and T7 RNA polymerases. 2. Materials and Methods (a) Bacteria, phages and plasmids

Plasmids were propagated in Escherichia coli HB101 and manipulated by standard recombinant DNA techniques, as described by Sambrook et al. (1989).

Specificity-determinant of T 3 and T7 R N A Polymerase Bacteriophage T3 and T7 were propagated and purified according to Studier (1969). T7 D4107 (a deletion mutant ofT7 that lacks all of gene 1; Studier & Moffatt, 1986) and T3 1-am2 (a T3 gene 1 amber mutant; Studier & Movva, 1976) were obtained from Dr F. W. Studier. Phages Rc38, Rb9 and Rb37 (Ryan & MeConnell, 1982) were obtained from Dr D. J. McConnell. Phage B04 (Molineux el al., 1983) was a gift from Dr I. J. Molineux. Plasmids that carry phage RNA polymerase genes under the control of an inducible lacUV5 promoter were propagated in E. coli BL21, which is defective in a surface protease known to cleave the phage enzymes (Grodberg & Dunn, 1988). Plasmids pRKD256, pRKD247 and pRKD243 carry T7 -10C, T7 - l l C and T7 - 1 2 A variant promoters (respectively) as well as consensus (wild-type) T3 and T7 promoters. These plasmids were constructed by subcloning HincII to BamHI fragments from pJ-2-29, pJ-2-27, and pK-3-19 (Klement el al., 1990) into the HincII to BamH1 sites of pBluescript S K + (Stratgene). Cleavage of these plasmids with EcoRV and SspI gives rise to 297 ntt, 243 nt and 164 nt run-off transcripts from the T7 WT, T3 WT and T7 variant promoters, respectively. Plasmid pJFK17 consists of the 924 base-pair PstI fragment of the neomycin resistance gene of pSV2nco (Mulligan & Berg, 1981) cloned into the PstI site of p J F K 9 (Ling el al., 1989). The neomycin gene (neo) fragment is located in a polylinker segment between opposing T3 and T7 WT promoters. Cleavage of this plasmid at the unique NcoI site within the neo sequences and subsequent transcription by T7 and T3 RNA polymerases gives rise to 400 and 586 nt run-off transcripts from the T7 and T3 promoters, respectively. (b) Reagents Nucleotides and DNA restriction and modification enzymes were purchased from Boehringer-Mannheim, Pharmacia, or New England Biolabs. Oligonucleotides were synthesized by the phosphoramadite method on an Applied Biosystems model 380B synthesizer, and purified by reverse-phase high pressure liquid chromatography (Becket et al., 1985). (c) Construction of T7/T3 hybrid RNA polymerase genes Plasmids pARl219 (Davanioo et al., 1984) and pCM56 (Morris et al., 1986) contain the wild-type T7 and T3 RNA polymerase genes, respectively. Where indicated (Fig. l) new restriction enzyme sites were introduced into the RNA polymerase genes by oligonucleotide-directed mutagenesis (Morinaga et al., 1984). Reverse translation of the predicted amino acid sequences of the two proteins and a subsequent analysis of the degenerate sequences (Mount & Conrad, 1986) permitted the identification of regions into which restriction enzyme sites could be introduced without altering the amino acid coding potential. A BstXl site was introduced into codon 674 of T7 gene 1 using the oligo 5' ACTCAGCCGAACCAAGCGGCTGGATACATG 3, an AvaI site was introduced into codon 752 using the oligo 5' CCTGATGTTCCTCGGGCAGTTCCGCTTACAGC 3', and a HindIII site was introduced into codon 805 using the oligo 5' GTACGGAATCGAAAGCTTTGCACTGATTC 3'. (Bold letters indicate positions where the oligonucleotide sequences differ from the wild-

t Abbreviations used: nt, nucleotides; WT, wild-type; DTT, dithiothreitol.

33

type target gene.) Similarly, an AvaI site was introduced into codon 753 of T3 gene I using the oligo 5' CGATATGATTTTCCTCGGGCAATTCCGTCTGC 3', and a H i n d I I I site was introduced into codon 806 using the oligo 5' GTATGGCATTGAAAGCTTTGCGCTCATCC 3'. (T3 gene 1 contains a naturally occurring BstX1 site in codon 675.) As a substrate for mutagenesis, fragments of the T3 or T7 RNA polymerase gene that extend from the common HpaI site at ~ 63°/o ofgene length (Fig. 1) to the 3' end of the cloned sequence present in pCM56 or pAR1219 were subcloned into a pBR322 derivative. The sequence of the altered DNA in the target region was confirmed by the method of Maxam & Gilbert (1979). After site-directed mutagenesis and subsequent manipulations to substitute T3 for T7 sequences, the T7fr3 subgenie hybrid RNA polymerase gene fragment was built into a full-length gene 1 (Joho, 1988). The T7/T3 hybrid gene 1 construct was then moved into a/acUV5 expression vector (pCM53; Morris et al., 1986) for induction and analysis. This seemingly circuitous strategy was necessary because of the presence of target restriction sites in the 5' two-thirds of gene 1 and in the lacUV5 expression vector. (d) Purification and c~aracterization of T7]T3 hybrid RN A polymerases To prepare crude lysates (Fig. 5), bacterial cultures (10 ml) were propagated in LB medium containing ampicillin (50 gg/ml; Sambrook et al., 1989) to an absorbanee at 600nm (A6oo) of 0+8, whereupon isopropylthiogalactoside (IPTG) was added to 0.4 mM. After 4 h, the cells were harvested, washed with lysis buffer (50 mM-Tris'HCI (pH 8), 0"2 mM-EDTA, 20 mm-NaCl, 1 mM-dithiothreitol (DTT)) and resuspended in 0"5ml lysis buffer containing 200 mm-NaCl. The cells were lysed by sonication, and the suspension was clarified by centrifugation (10,000g, l0 rain). One microliter of lysate (containing ~0-5#g of RNA polymerase) was used directly in transcription assays, as described below. In the case of the hybrid RNA polymerase gene encoded by plasmid pKJ33, the crude extract was further purified by chromatography over a Mono-Q column. The clarified lysate (500#!) was injected onto a Mono-Q HRS/5 column (Pharmaeia) equilibrated with 20 mm-KHPO+, 1 mm-EDTA, 1 mM-DTT, 5% (v/v) glycerol, 200 mm-NaC1. The column was eluted with a linear gradient from 0"2 to 1"0 M-NaCI over a period of 60 min. Individual fractions (0"5 ml) were analyzed by in vitro transcription assays using pJFK17 as template, as described below. Peak fractions were pooled and dialyzed against buffer containing 50% glycerol. For the template competition experiments presented in Fig. 4, the enzyme preparations were partially purified by chromatography over phosphocellulose. The clarified lysate was diluted with 7 vol. of buffer lacking NaCl and applied to a C10/10 column (Biorad) containing phosphocellulose (bed height of 5 cm). The column was washed with 5 mi of lysis buffer containing 15 mM-NaCl and then with 10 ml of buffer containing 200 mM-NaCl. RNA polymerase was eluted with buffer containing 400 mM-NaCl. Peak fractions were pooled and used directly in transcription assays, as described below.

(e) Transcription assays Transcription reactions were carried out in a volume of 50/~1 containing 2/~g template DNA, 1/~l cell lyaate, 20 mm-Tris'HC1 (pH 7"9), 20 mM-MgCl2, 1 mM+DTT,

K . E . Joho et al.

34

2mM-spermidine-HCl, 0"4mM-ATP, GTP, UTP, and 0-1 mrd-[~-a2P]CTP (spec. act., 80/~Ci/#mol). The reactions shown in Fig. 4 contained, in addition, 30 units RNase inhibitor (Boehringer-Mannheim). After 15 min of incubation at 37 °C, KCI was added to a concentration of 150 mrd, and incubation was continued for an additional 5 min. An equal volume of stop buffer (0"1M-EDTA, 100 #g yeast tRNA/ml) and 200 #l ethanol was added and the samples were chilled on ice. The precipitate was collected by centrifugation (10,000 g, 5 rain), washed with 70% ethanol, dried in vacuo, and resuspended in 20 #l sample buffer (4 mM-Tris-acetate (pH 7-5), 2 mM-sodium acetate, 1% (w/v) sodium dodeeyl sulfate (SDS), 2 mM-EDTA, 0-14M-fl-mercaptoethanol, 10% glycerol) and heated to 90°C for 2 rain. The samples were cooled and resolved by electrophoresis in 4% polyaerylamide gels in a Tris-acetate buffer (Studier, 1973).

(b) Construction of hybrid T7/T3 R N A polymerase genes

3. Results (a) Analysis of D N A from T 7 X T 3 recombinant phaqe particles Previous efforts to identify the promoter specificity determinant of the phage RNA polymerases involved restriction mapping of the gene I region of a number of recombinant TTXT3 phage (Ryan & McConnell, 1982). The restriction mapping data were obtained at a time when detailed sequence information for the two polymerase genes was not yet available. We have repeated the mapping on a few of the original hybrid phage and on one phage (B04; Molineux et al., 1983) that had not been previously characterized. The results of this analysis (Fig. 2) have allowed us to refine the crossover points in the various hybrids, and to define more accurately the region of polymerase specificity. Five of the phage tested appear to have a single crossover event that could be mapped with reasonable

Phage

Template specificity

Rb9

T5

Rb37

T3

B04

T5

0 ~ ,,

Percentage of gene length 50 ~

I

["

,......

l

I

I00 l

I

I

II

I o~r

Rc58

T7

i

certainty. The data are consistent with previous results, and suggest that the promoter-specificity domain lies in the carboxyl terminal 30% of the protein, distal to amino acid 623. The hybrid phage designated Re38 is of particular interest. In all other hybrid phages tested, the 5' (amino-terminal) portion of the hybrid gene arises from T7 sequences and the 3' (carboxyl-terminal) region arises from T3 sequences. The hybrid gene in Rc38 reverses this situation, having a 5' region that arises from T3 sequences and a 3' region from T7. As with the other hybrid polymerases the specificity of the enzyme correlates with the 3' end of the gene, which in this case is T7-specific.

I I

Figure 2. Structure of the gene 1 region in hybrid phage DNA. Physical maps of the gene 1 region of the hybrid phage were determined by comparative restriction enzyme mapping (data not shown). DNA from the gene 1 region of each hybrid phage was digested with a variety of restriction enzymes, and the digestion patterns were compared with those obtained with wild-type T3 or T7 DNA. Open bars indicate regions that arise from T3 sequences, continuous lines from T7 sequences. The boundaries of the crossover region are given as pereentage of gene length. The specificity of the RNA polymerase encoded by each hybrid phage was determined in vitro using a plasmid DNA template that contains both T3 and T7 promoters, as described for Fig. 4.

Analysis of hybrid genes by restriction enzyme mapping cannot reveal all of the potential crossovers and/or mutations that may have occurred during the generation and selection of the recombinant phage particles. To confirm and localize further the specificity domain a series of hybrid T3/T7 polymerase genes was constructed in vitro, taking advantage of the availability of the cloned T3 and T7 RNA polymerase genes (Davanloo et al., 1984; Morris et al., 1986). In this manner the precise location of the crossover event could be controlled. Furthermore, since the cloned RNA polymerase genes are not required to perform a biological function in their bacterial hosts, selective pressures that might introduce secondary mutations elsewhere in the polymerase gene are eliminated. The T3 and T7 RNA polymerase genes both contain a HpaI site at 63% of gene length (amino acid position 560, see Fig. 1). By utilizing this common site it was possible to replace the region that specifies the carboxyl 323 amino acids of the T7 RNA polymerase with the corresponding region from the T3 RNA polymerase gene (Fig. 3). The hybrid gene was then inserted into an expression vector under control of the lacUV5 promoter to facilitate the expression and characterization of the polymerase. The resulting plasmid (pNM44) encodes an enzyme that exhibits T3 specificity in an in vitro transcription assay (Fig. 4). Although the amino acid sequences of the T3 and T7 RNA polymerases are highly conserved, their DNA sequences are only 67% identical. Consequently, there are few unique restriction sites that are common within the two genes. To construct additional hybrid RNA polymerase genes it was necessary to introduce restriction sites into the cloned genes without altering the coding potential. Three sites that lie within the carboxyl-terminal one-third of both genes, and which divide the region approximately equally, were chosen. A BstXI site was introduced at amino acid 674, an AvaI site at amino acid 752, and a HindIII site at amino acid 805 (see Fig. 1). These new restriction sites were utilized to exchange defined regions of the T7 RNA polymerase

Specificity-determinant of T3 and T7 R N A Polymerase Hybrid gene

Plasmid

I)CM56

I

NONE

T7

ALL

T3

560-883

T3

674 - 6 8 3

T5

752-883

T7

CZ]

806-883

T7

P,.-

560-805

T3

560-673

T7

j

pNM44

levels of activity on their preferred t e m p l a r , and the specificity of these enzymes was readily determined in vitro using crude extracts from induced cultures. The pKJ33 hybrid RNA polymerase exhibited low, but detectable, levels of activity, and a partial purification of this enzyme was required in order to determine its specificity.

T 5 residues Specificity

pARI219

~

pLG2

~

pKJ46

~ ~

pKJ43 pKJ33

,q

pK J 31

[~)

(c) Specificity of the hybrid R N A polymerases in vivo The specificity of the hybrid polymerases in vivo was determined by means of a complementation assay in which the ability of the cloned polymerase gene to support the growth of gene I - phage particles was tested (Table 1). The experiment was performed using T7 phage that were entirely deleted in the gene 1 region (T7 D4107; Studier & Moffatt, 1986) or T3 phage that carry an amber mutation within gene 1 (T3 1-am2; Studier & Movva, 1976). This analysis was carried out in order to confirm that the specificity observed in vitro was reflected in vivo, and to determine whether other functions of the RNA polymerase not involved in transcriptional activity might be identified in the hybrid genes. For example, a T7fr3 hybrid polymerase might be transcriptionally active on T3 DNA but unable to interact correctly with T3 proteins that are involved in DNA packaging or replication, and thus be unable to complement T3 gene 1- phage. The cloned wild-type T3 and T7 RNA polymerases exhibited tight specificity, complementing only their respective phage types. Similarly, all T7fr3 hybrid enzymes that retained T7 specificity complemented T7 gene 1- phage but not gene 1phage. The T7/T3 hybrid polymerases that had altered (T7-*T3) specificity fully complemented T3 gene 1- phage, but retained a weak ability to complement T7 gene 1- phage, as evidenced by the

F i g u r e 3. Structure of hybrid T3/T7 RNA polymerase genes. Starting with the cloned wild-type RNA polymerase genes, hybrid T7fF3 polymerase genes were constructed in vitro and cloned into a pBR322-based expression vector under control of the/acUV5 promoter. T7 and T3 sequences are indicated by continuous lines and open boxes, respectively. The interval in the T7 enzyme that is replaced by the corresponding amino acids from the T3 RNA polymeraee is indicated under "T3 residues". The specificity of the hybrid enzyme, as determined in an in vitro transcription assay (Fig. 4), is indicated in the right-hand column.

gene with the corresponding region of the T3 gene, and the resulting hybrid genes were inserted into an expression vector as described above (Fig. 3). An analysis of the specificity of the resulting hybrid RNA polymerases (Fig. 4) reveals t h a t promoter preference is determined by the interval of the gene that encodes amino acids 674 to 752 (designated the 674-752 interval). All of the hybrid RNA polymerases, except that encoded by pKJ33, exhibited normal (wild-type)

c~J C~.

Q.

-Ca

CL

35

if)

ro

~.

~1"

u3

O.

Q.

¢X

Q-

~

_ 0.

TS-

TT-

--

T3

--

T7

F i g u r e 4. Specificity of hybrid RNA polymerases. The specificity of the hybrid polymerases was determined by competition run-off assays. Extracts from cells that carry the indicated plasmid were incubated with a plasmid DNA template (pJFK17) that carries both T3 a n d T7 consensus promoters. The plasmid had been digested with NcoI so as to give a 400 nt transcript from the T7 promoter or a 586 nt transcript from the T3 promoter, as indicated in the margin. See Fig. 3 for the structure of the RNA polymerase gene in each plasmid.

K . E . Joho et al.

36

Table

1

Ability of T3/T7 hybrid R N A polymerases to complement T3 and T7 gene 1- phage T7 gene 1 - (D4107) Host BL21 BBW]I BL21[pARI219I BL21[pCM56] BL21[pNM44] BL21[pKJ31] BL21 [pKJ33] BL21[pKJ43] BL21[pKJ46] BL21[pLG2]

T3 gene 1- (am 2)

Titer

e.o.p.t

Titer

e.o.p.~"

Specificity of RNAP•

< l0 s < 105 2-1 x 10 l° 4"0 > 4.0 0"0 0.0 0.0

T7 - 12A 0.6 0.0 0"0 0-1 0.7 0.6 1.0

J" Individual lanes in Fig. 5 were scanned with an L K B densitometer and the density of each band was normalized to take into account the sizes of the transcripts. The density of the band arising from the variant promoter is expressed relative to t h a t arising from the preferred consensus promoter.

Specificity-determinant of T3 and T7 R N A Polymerase

pA.,2 I -,2 -,, -,o1-,2 -,,

I

p...4 -,,

pKJ46

- , o - , 2 -,, - , o

37

pKJ43 I pKJ31 I

-12 -H - I 0 -12 - I I -101-12 -I! - I 0 I ,

ii

T7 T3

VARIANT

Figure 5. Utilization of T7 promoter variants by hybrid RNA polymerases. The preference of RNA polymerases encoded by the plasmids indicated was determined by means of a promoter competition assay. Partially purified preparations of RNA polymerase were incubated with a plasmid that contains three promoters: a consensus T3 promoter (T3 WT), a consensus T7 promoter (T7 WT) and either a T7 -10C, a T7 - I IC or a T7 -12A variant promoter. The template had been digested with EcoRV and SspI so that transcription would give rise to characteristically sized run-off products from each promoter, as indicated in the margin. The products were resolved by electrophoresis in polyacrylamide gels, followed by autoradiography. The particular T7 promoter variant contained within the plasmid is identified above the lane by indicating the position in the T7 promoter at which a T3 base has been substituted (i.e. - 10, - 11 or - 12). See Fig. 3 for the structure of the RNA polymerase gene in each plasmid. The observation that the pLG2 and pNM44 hybrid enzymes preferentially utilize a T7 - l l C variant over the T3 consensus promoter suggests that there is a separate region of the polymerase {which has not been switched in theseexperiments) that is responsible for identifying bases at other positions where the T3 and T7 promoter sequences differ. From previous work, the only other positions at which the two consensus sequences differ are at - 2 and - 1 5 . Whereas changing the base-pair at - 2 in a T7 promoter (a T7 - 2 A variant) had relatively little effect on recognition by the T7 RNA polymerase, switching the base-pair at - 1 5 (a T7 - 1 5 T variant) resulted in a 60~/o loss of promoter activity (Klement et al., 1990). For this reason, it is likely that the modulating effect of the second specificity determinant of the polymerase {which we have not yet identified) involves interactions with basepairs in the - 1 5 region of the promoter. 4. Discussion

In this work, two sets of experiments t h a t implicate a region near the carboxyl terminus of the T3 and T7 RNA polymerases as the primary determinant of template specificity have been described. In recombinant phage generated in vivo the specificity determinant correlates with the region of the RNA polymerase distal to amino acid 623, and in recombinant RNA polymerase genes generated in vitro the specificity of the enzyme correlates with the region of the gene that encodes amino acids 674 to 752. Promoter-competition experiments involving synthetic T7 promoter variants indicate that the region encompassing amino acids 674 to 752 (the 674-752 interval) is concerned with recognizing base-pairs in the - 1 1 region of the promoter, and

that another domain of the RNA polymerase (which we have not yet identified) is involved in identifying the base-pair at - 1 5 . These results are consistent with those of previous investigators, who found that the primary determinant of polymerase specificity was likely to lie in the C-terminal third of the enzyme, but that there might be other regions of the RNA polymerase that modulate the degree of polymerase specificity (Beier & Hausmann, 1974; Hausmann & Tomkiewitz, 1976; Beier et al., 1977, Ryan & McConnell, 1982). On the basis of the promoter preference of pNM44 for the synthetic T7 promoter variants, it is likely that the second specificity determinant lies to the left of the HpaI site at 0-63 gene length (see Fig. 3). There are a number of ways to view our results. Perhaps most attractive is the notion t h a t amino acids in the 674-752 interval make base-specific contacts with the promoter in the - - l l region. In this view, one or a few of the amino acids that differ between T3 and T7 in this interval would determine the specificity of the enzyme by contacting the basepairs in the - 1 1 region in a different manner. Alternatively, the differing residues may not themselves be in contact with the - l l base-pairs, but their presence may change the orientation of {other) conserved residues in the same region, allowing these residues to make alternate contacts with these base-pairs. Lastly, it is possible that none of the amino acids in this interval is in contact with the DNA but, rather, this region serves to position some other interactive domain, and that the amino acid substitutions in the 674-752 interval merely shift the orientation of the interactive domain. In the absence of crystallographic data, none of these possibilities may be excluded. However, it may be possible to infer structure by homology to other sequence-specific DNA binding proteins. The

38

K.E.

Joho et al.

(o) EcoRI

183

~ S G I L ~ R U ~ L :::

T7 RNP 730 Gene ,4

P

D

:

::..

G

~

TAANY~

: .:: . . . . . . . .

I ~ ..: .... : . . : .

:

INLQAASZf~D~a~m~'~MFDIST~L : ............ :...:

...............

FIDQFRID_Fr~EIDAHEDESG . ::.: . .: : .:.::..:..

GIESFAL .::

43

(b)

-i0

+i

T

G

~ . G G - G A C . A ~ G A

tta

ac ga

a

~6

aagag

c

c

~ttaataaCA~OCAC [

]

re~x~lition /cleavage

#X174 gene A [

]

binding

Figure 6. Comparison of amino acid sequences and DNA binding sites for EcoRI, T7 R,NA polymerase, and the bacteriophage ¢X174 gene A protein. (a) The sequences of EcoRI, T7 RNA potymerase, and the (bX174 gene A protein were aligned by use of the program FASTP (Lipman & Pearson, 1985). The start of the alignment in each protein is indicated by the residue number at the left. The significance of the alignments with T7 RNA polymerase are 3"8 and 4"5 standard deviations above the random mean for EcoRI and the gene A protein, respectively, as determined by the program RDF (Lipman & Pearson, 1985). (b) Comparison of promoter sequences for the phage RNA polymerases (Klement et al., 1989) with the binding and cleavage sequences for the ¢X174 gene A protein (Fluit et al., 1984). In the promoter consensus sequence, uppercase letters indicate that the base is conserved in 3 out of 4 promoters, lowercase letters indicate non-conserved bases. program F A S T P (Lipman & Pearson, 1985) was used to search the N B R F database for proteins t h a t might be related to the T7 and T3 RNA polymerases, and in particular, those t h a t might be involved in sequence-specific binding to DNA. Two proteins of particular interest were identified in this manner: the gene A protein of bacteriophage CX174, and the restriction enzyme E c o R I (Fig. 6(a)). Both of these proteins are sequence-specific DNA binding proteins that, like the phage RNA polymerases, are believed to perturb the helical structure of the DNA during or after binding, I m p o r t a n t l y , the alignment of these proteins with T7 RNA polymerase overlaps the 674 to 752 interval. The ¢ X 1 7 4 gene A protein, which is involved in the replication of ¢ X 1 7 4 DNA, binds to a recognition sequence a t the ¢ X 1 7 4 origin, melts open the DNA helix, and then cuts one strand of the duplex DNA in an adjacent region to initiate DNA replication (Fluit et al., 1984). Interestingly, the binding sequence t h a t is utilized by the gene A protein contains a portion of the core promoter consensus sequence t h a t is utilized by all of the T7-1ike phage R N A polymerases (Fig. 6(b)). The DNA sequence recognized by E c o R I (GAATTC) is, in contrast, quite distinct from the phage promoter sequence. The region of E c o R I t h a t aligns with the T3 and T7 RNA polymerases contains one of the specificity

determinants of E c o R I , the outer recognition helix, centered on Arg200 (McClarin et al., 1986). This homologous region may represent a common structural motif t h a t is i m p o r t a n t in sequence-specific contacts, and which m a y also be involved in perturbation of the helical structure of the DNA, perhaps by providing torsional constraint. Experiments to test this hypothesis are in progress, and involve sitedirected mutagenesis to change individual T7 amino acid residues in this region to the corresponding residues found in the T3 RNA polymerase. This work was supported by grants GM21783 and GM38147 from the National Institutes of Health. Part of the work described was performed as fulfillment of the requirement of the Ph.D. degree of K.E.J. at Rutgez~ University and the University of Medicine and Dentistry of New Jersey. We are grateful to Dr Russell K. Durbin for constructing the RKD series of plasmids, and for helpful discussions.

References Becker, C. R., Efcavitch, J. W., Heiner, C. R. & Kaiser, N. F. (1985). J. Chronmt. 326, 293-299. Beier, H. & Hausmann, R. (1974). Nature (Lo~uton), 251, 538-540. Beier, H., Golomb, M. & Chamberlin, M. (1977). J. Virol. 21, 753-765.

Specificity-determinant of T 3 and T 7 R N A Polymerase

Chamberlin, M., McGrath, J. & Waskelt, L. (1970). Nature (London), 228, 227-231. Davanloo, P., Rosenberg, A. H., Dunn, J. J. & Studier, F.W. (1984). Proc. Nat. Acad. Sci., U.S.A. 81, 2035-2039. Dunn, J. J., Bautz, F. A. & Bautz, E. K. F. (1971). Nature New Biol. 230, 84-96. Fluit, A. C., Baas, P. D., van Boom, J. H., Veeneman, G. H. & Jansz, H.S. (1984). Nucl. Acids Res. 12, 6443-6454. Grodberg, J. & Dunn, J. J. (1988). J. Bacteriol. 170, 1245-1253. Hausmann, R. & Gomez, B. (1967). J. Virol. 1,779-792. Hausmann, R. L. & Tomkiewicz, C. (1976). In R N A Polymerase, pp. 731-743, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Hinkle, D. C. (1980). J. Virol. 34, 136-141. Joho, K. (1988}. Ph.D. thesis, Rutgers University, New Brunswick, NJ. Klement, J. F., Moorefield, M. B., Jorgenson, E., Brown, J. E., Risman, S. & McAllister, W. T. (1990). J. Mol. Biol. 215, 21-29. Ling, M.-L., Risman, S. R., Klement, J. F., McGraw, N. & McAllister, W.T. (1989). Nucl. Acids Res. 17, 1605-1618. Lipman, D. J. & Pearson, W. R. (1985). Science, 227, 1435-1441. Maxam, A. M. & Gilbert, W. (1979). In Methods in Enzymology (Grossman, L. & Moldave, K., eds), vol. 65, pp. 499-560, Academic Press, NY. McClarin, J. A., Frederick, C. A., Wang, B. C., Greene, P., Boyer, H. W., Grable, J. & Rosenberg, J. N. (1986). Science, 234, 1526-1541. McGraw, N. J., Bailey, J. N., Cleaves, G. R., Dembinski,

39

D. R., Gocke, C. R., Joliffe, L. K., MacWrigh~, R. S. & McAllister, W.T. (1985). Nucl. Acids Res. 13, 6753--6766. Moffat, B. & Studier, F. W. (1987). Cell, 49, 221-227. Moffatt, B. & Studier, F. W. (1988). J. Bacteriol. 170, 2095-2105. Moffatt, B., Dunn, J. J. & Studier, F. W. (1984). J. Mol. Biol. 173, 265-269. Molineux, I. J., Mooney, P. Q. & Spenee, J. L. (1983). J. Virol. 46, 881-894. Morinaga, Y., Franceschini, T., Inouye, S. & Inouye, M,. (1984). Biotechnology, July, 636-639. Morris, C. E., Klement, J. F. & McAllister, WT (1986). Gene, 41, 193-200. Mount, D. W. & Conrad, B. (1986). Nucl. Acids Res. 14; 443-454. Mulligan, R. C. & Berg, P. (1981). Proc. Nat. Acad. Sci., U.S.A. 78, 2972-2976. Ryan, T. & McConnell, D. J. (1982). J. Virol. 43,844-858. Saito, H., Tabor, S., Tamanoi, F. & Richardson, C. C. (1980). Proc. Nat. Acad. Sci.. U.S.A. 77, 3917-3921. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning, A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Studier, F. W. (1969). Virology, 39, 562-574. Studier, F. W. (1973). J. Mol. Biol. 79, 237-248. Studier, F. W. & Moffatt, B. A. (1986). J. Mol. Biol. 189, 113-130. Studier, F. W. & Movva, N. R. (1976). J. Virol. 19, 136-145. Zavriev, S. K. & Shemyakin, M. F. (1982). Nud. Acids Res. I0, 1635-1652.

Edited by M . Gottesman

Note added in proof. We have recently determined that switching the three amino acids at positions 749, 750 and 751 from T7 residues (Ash, Leu, Met) to T3 residues (Asp, Met, Ile) is sufficient to switch the specificity of the RNA polymerase. The preference of the altered T7 RNA polymerase containing T3 residues at these positions is identical to that of the hybrid LG2.