Point Mutations Which Drastically Affect the Polymerization Activity

TXE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 267, No. 14, Issue of May 15, pp. 101~10176,1992 Printed in U. S. A.

Point Mutations Which Drastically Affect the Polymerization Activity of Encephalomyocarditis Virus RNA-dependent RNA Polymerase Correspond to the Active Site of Escherichia coli DNA Polymerase I*

(Received for publication, December 5, 1991)

Sabita Sankar and Alan G . Porter$ From the Protein Engineering Group, Institute of Molecular and Cell Biology, National University of Singapore, 10 Kent Ridge Crescent, Singapore 0511

The inhibitor sensitivity and functional domains of recombinant encephalomyocarditis (EMC) virus RNA- dependent RNA polymerase (3DP0’) have been exten- sively analyzed. The inhibitor profiles of EMC virus 3Dm1 and Escherichia coli DNA-dependent RNA polymerase are distinct, and experiments with substrate analogs indicate that EMC virus 3DP” lacks reverse transcriptase activity. Twenty amino acid substitutions were engineered in EMC virus 3Dp0’ based on sequence alignments of viral RNA-dependent RNA polymerases that identified conserved amino acid residues within motifs. Ten out of 17 conservative substitutions within the four most conserved motifs reduced the RNA polymerase activity of the mutants to 0-6% of the activity of the wild-type enzyme, demonstrating the importance of these amino acids in the structure and/or function of EMC virus 3Dp0’. Remarkably, 5 of the 10 mutations in EMC virus 3DPo1 which had the most drastic effect on its RNA polymerase activity (D240E, S293T, N302Q, G332A, and D333E) were found to correspond to active site residues in E. coli DNA-dependent DNA polymerase I (Klenow). Our results reveal that a basic structural and functional framework is conserved in the most distantly related classes of nucleic acid polymerases and demonstrate the validity of modeling the active site of an RNA- dependent RNA polymerase on the known structure of a DNA polymerase.

Encephalomyocarditis (EMC)l virus and poliovirus, which are classified in different genera of the picornavirus family, have a single-stranded RNA genome of positive polarity that is replicated in the cytoplasm of infected cells by the virus- encoded RNA-dependent RNA polymerase (3DP”’) (Palmen- berg et al., 1984; Flanegan and Baltimore, 1977; Nicklin et aL, 1986). A template-dependent form of poliovirus 3DP0’ was originally isolated by taking advantage of a sensitive assay for poly(U) polymerase activity using a poly(A) template and an oligo(U) primer (Flanegan and Baltimore, 1977; Dasgupta et al., 1979). The poliovirus enzyme exhibits M e and Zn2+ ion dependence and RNA elongation activity that requires some

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom correspondence and reprint requests should be sent. The abbreviations used are: EMC, encephalomyocarditis; SDS,

sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; HPLC, high pressure liquid chromatography; Hepes, 4-(2-hydroxy- ethyl)-1-piperazinethane sulfonic acid; kb, kilobase pair(s); HIV-1, human immunodeficiency virus, type I.

form of a primer or host factor (Dasgupta et al., 1980; Baron and Baltimore, 1982; Van Dyke et al., 1982).

Much less work has been done on EMC virus 3Dp0’, although a partially purified preparation of the enzyme was isolated which exhibits poly((=)-dependent poly(G) polymerase activity (Rosenberg et al., 1972; Traub et aL, 1976). More recently, a partially purified preparation of EMC virus 3DPo1 with an associated RNA helicase activity was found to exhibit poly(U) polymerase activity (Dmitrieva et al., 1991).

Owing to the difficulties in obtaining stable, pure viral RNA-dependent RNA polymerase, several groups have pro- duced catalytically active EMC virus or poliovirus 3DP” from cloned cDNA using various Escherichia coli expression systems (Morrow et al., 1987; Rothstein et al., 1988; Plotch et al., 1989; Sankar and Porter, 1991). The recombinant EMC virus 3DPo1 has been expressed in large quantities and resembles the poliovirus enzyme in its in uitro requirement for a primer (e.g. oligo(U)) to copy poly(A) or viral RNA into a full-length RNA (Sankar and Porter, 1991).

The availability of a growing number of complete sequences from the genomes of positive, negative, and double-stranded RNA viruses of plant and animal origin has enabled many groups to carry out sequence alignments of the RNA-dependent RNA polymerases in an attempt to identify regions essential for polymerase function, which should appear as most conserved (Kamer and Argos, 1984; Argos, 1988; Poch et al., 1989; Bruenn, 1991; Koonin, 1991). There is general agree- ment that there are at least three or four conserved motifs which contain a few absolutely conserved amino acids (motifs IV-VI1 in Figs. 1 and 2), and these have been suggested to form the catalytic center of this class of polymerases (Poch et al., 1989; Bruenn, 1991; Koonin, 1991). Support for this idea comes from mutagenesis studies, showing that some of the conserved amino acids within motifs IV and VI are required for the catalytic activity of brome mosaic virus RNA polymerase and &/I replicase, respectively (Inokuchi and Hirashima, 1987; Kroner et al., 1989). Furthermore, motifs IV and VI align with similar motifs in the reverse transcriptases and various DNA-dependent DNA and RNA polymerases (Poch et al., 1989; Delarue et al., 1990) which also contain residues important for polymerase activity (Larder et al., 1987; Bernad et al., 1990; Dorsky and Crumpacker, 1990; Marcy et al., 1990; Polesky et al., 1990). In addition, motifs IV and VI have been aligned with part of the active site in the known three- dimensional structure of E. coli DNA-dependent DNA polymerase I (Klenow) (Ollis et al., 1985; Delarue et al., 1990). The combined data for RNA-dependent RNA polymerases from computer alignments and from genetic and mutagenesis studies have not, however, revealed the overall structure of the active site and its role in catalysis.

10168

Putative Active Site of EMC Virus

In this paper, we report an extensive analysis of the functional domains in EMC virus 3DPo1 by engineering point mutations in amino acids which lie within the four most conserved motifs of viral RNA-dependent RNA polymerases. The results allow us to identify some of the amino acids in EMC virus 3DPo1 which are likely to participate in RNA template binding and catalysis. In addition, we have characterized EMC virus 3DPo1 biochemically with respect to its sensitivity toward substrate analogs and various inhibitors of DNA and RNA polymerases to better understand how closely related this enzyme is to the other known polymerases.

EXPERIMENTAL PROCEDURES

Materials-Oligonucleotides were synthesized using a Pharmacia gene assembler and purified by HPLC. Aphidicolin, cerulenin, N- ethylmaleimide, dactinomycin, heparin, and periodate-oxidized UTP (UTP 2',3'-dialdehyde) were purchased from Sigma, and phosphonoacetic acid was obtained from Aldrich. Rifampicin, dTTP, and the nucleotide analogs ddTTP, 3'-dUTP, UDP, and UMP were purchased from Boehringer Mannheim. Guanidine hydrochloride was obtained from BRL, and brefeldin A from Epicentre Technologies. Restriction enzymes, T4 DNA ligase, and T4 polynucleotide kinase were either from Amersham Corp. or Boehringer Mannheim. E. coli DNA-dependent RNA polymerase was purchased from Pharmacia LKB Biotechnology Inc.

Site-directed Mutagenesis and DNA Sequencing-Mutagenesis was carried out essentially as described by Kunkel et al. (1987) using the Muta-gene phagemid in vitro mutagenesis kit from Bio-Rad, following the protocols described by the manufacturer. First, a phagemid recombinant was constructed by subcloning a 1.904-kb XhoI-BamHI fragment of the EMC virus cDNA encoding most of the EMC virus 3Dw1 gene from pT12 (the expression vector containing the glutathione S- transferase gene fused to the EMC virus 3DPo1 gene; Sankar and Porter (1991)) into the multiple cloning site of the Bluescript KS+ vector (Stratagene) as shown in Fig. 3. The resultant recombinant phagemid pSS700 was transformed into Dut- Ung- E. coli strain CJ236. Phagemid particles containing single-stranded, uracil- enriched DNA template for mutagenesis were obtained by superin- fection of E. coli CJ236 carrying pSS700 with the helper phage M13K07. The uracil-containing DNA was purified and used as a template for in vitro mutagenesis.

The mismatch oligonucleotides were phosphorylated in vitro prior to their addition to the mutagenesis reaction. (Their nucleotide sequences are available on request.) The mutagenic oligonucleotides were annealed to the template DNA, closed circular DNA was synthesized by T4 DNA polymerase in the presence of phage T4 gene 32 protein, and the DNA was transformed into an Ung' E. coli strain (XL-1 Blue, Stratagene). Four to eight colonies from each mutagenesis reaction were screened for the correct mutation by double- stranded dideoxy sequencing using the Sequenase kit version 2.0 (U. S. Biochemical Corp.). The primers used for the sequencing reactions span three regions within the carboxyl-terminal region of EMC virus 3Dw1. The sequences of these primers are shown below.

SS128 S'AACGGCAAGCATAGTGATCGA3' SS129 5'AGTTATCTTATATCCTGTCTTGC3' SS130 5'TGAAATGGCTAATGACTCAAG3'

To generate the plasmids expressing the mutant polymerase genes, the mutant derivatives of pSS700 were digested with XhoI and BamHI and the 1.904-kb fragment ligated to the 4.796-kb XhoI-BamHI fragment obtained from the parental expression plasmid pT12 from which the wild-type polymerase gene had been excised (Fig. 3). The DNA was transformed into E. coli strain HB101, and the recon- structed expression plasmids carrying the mutated polymerase gene (pSD series) were identified by extensive restriction enzyme analysis.

Expression and Purification of Recombinant EMC Virus 30"' and Site-directed Mutants-Overnight cultures of E. coli transformed with pT12 (the plasmid containing the wild-type EMC virus 3DPo1 gene) or the pSD series of plasmids (which contain the mutated EMC virus 3DP"' gene) were diluted 1:50 in 1 liter of L-broth containing ampicillin (100 pg/ml) and grown at 30 "C to an Am0 of 0.45 when isopropyl-l- thio-8-D-galactoside was added to 0.3 mM. After 4-6 h of growth at 30 "C, the cells were harvested by centrifugation, and the pellet stored a t -20 "C or resuspended in 60 ml of 50 mM Tris-HC1 (pH 8.0), 150 mM NaCl, 0.25 mM EDTA (buffer 1).

RNA-Dependent RNA Polymerase 10169

All subsequent steps were carried out at 4 "C. The suspension was sonicated in 10-ml aliquots using a 2O-kilocycles/s sonicator (Heat Systems Ultrasonics). The supernatant, recovered after centrifugation at 12,000 rpm for l h, was loaded under gravity onto a 2-ml glutathione-Sepharose 4B column (Pharmacia) equilibrated in buffer 1. The eluate was retained and reapplied to the column. Following absorption of the fusion protein to the glutathione affinity matrix, the column was washed several times with buffer 1. After draining off the final wash, the gel slurry was covered in 400 pl of buffer 1 containing 480 ng of bovine thrombin (Sigma) and incubated at 4 "C for 3-4 h. The purified recombinant EMC virus 3Dw' was recovered by applying 10 ml of buffer 1 to the column and collecting the eluate. An aliquot (-150 pl) was set aside for sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis. Glycerol and MgCl, were added to the remainder of the eluate to give a final composition 40 mM Tris (pH 8.0), 138 mM NaC1,0.23 mM EDTA, 10 mM MgC12, and 5% glycerol.

Enzymatic Assays-The poly(A)-dependent oligo(U)-primed poly(U) polymerase assay was performed as described by Sankar and Porter (1991). Briefly, the reaction was carried out at 30 "C for 60 min in 50 pl containing equivalent amounts of the purified wild-type and mutant enzymes (600 ng), 50 mM Hepes, pH 8.0, 10 p~ UTP, 4 mM dithiothreitol, 3 mM magnesium acetate, 6 p~ zinc chloride, 20 pg/ml rifampicin, 1.0 pg of oligo(U), 2.5 pg of poly(A), and 5 pCi of [w3'P]UTP (Amersham, 400 Ci/mmol). The in vitro synthesized product was precipitated by 10% trichloroacetic acid with 100 pg of carrier tRNA in 0.2 M sodium pyrophosphate, collected on 0.45-pm Whatman GF/C filters, vacuum dried, and solubilized in scintillation fluid. The radioactivity was determined in a scintillation counter.

Gel Electrophoresis and Protein Concentration Determination- SDS-PAGE was performed as described by Laemmli (1970). The protein concentration was determined by Bio-Rad protein assay.

RESULTS

Sensitiuity to Inhibitors of DNA and RNA Polymerases- We have recently purified EMC virus 3DPo1 from recombinant E. coli and quantitated its polymerization activity using an oligo(U)-primed poly(U) polymerase assay (Sankar and Por- ter, 1991). This assay permits the sensitivity of the enzyme toward a range of inhibitors to be determined. As shown in Table IA, the EMC virus 3DPo1 poly(U) polymerase activity was resistant to rifampicin and heparin at concentrations which completely inhibit the activity of E. coli DNA-dependent RNA polymerase. Dactinomycin, an inhibitor of eukaryotic DNA-dependent RNA polymerases also had no effect on EMC virus 3DP0', which agrees with the findings for poliovirus (Morrow et al., 1987; Rothstein et al., 1988). Conversely, EMC virus 3DPo1 was highly sensitive to cerulenin at 100 PM and N-ethylmaleimide at 5 mM, while E. coli RNA polymerase was resistant to these compounds (Table IA).

Cerulenin, a specific inhibitor of fatty acylation and sterol biosynthesis, was tested because at 0.1 mM it selectively inhibits poliovirus plus strand RNA synthesis in poliovirus- infected cells (Guinea and Carrasco, 1990). Because cerulenin fails to inhibit the synthesis of poliovirus RNA in a cell-free system consisting of membrane-bound replication complexes, the suggestion was made that continuous phospholipid synthesis was required for poliovirus replication (Guinea and Carrasco, 1990). The inhibition of EMC virus 3DPo1 polymerization activity in a lipid-free assay is therefore surprising; the value of 1.25 p~ is well within the concentrations used in the experiments with poliovirus-infected cells (Guinea and Carrasco, 1990).

In order to provide evidence that the inhibition by cerulenin is specific, another compound which affects cellular biosynthesis was tested. Brefeldin A, a drug which blocks protein secretion and disrupts the Golgi structure (Misumi et al., 1986) was found to have no effect on the in vitro activity of EMC virus 3DP' (Table IA). In addition, millimolar concentrations of guanidine HCl, which inhibits poliovirus RNA synthesis in cell culture, probably at the initiation step (Cal-

10170 Putative Active Site of EMC Virus RNA-Dependent RNA Polymerase TABLE I

Effects of inhibitors and substrate analogs on the poly(U) polymerase activity of EMC virus 3Dpd All assays were performed in duplicates as described under “Experimental Procedures,” with the inhibitor or

analog being added before the enzyme. Under B, the concentrations of analogs were varied (as shown) in the presence of a constant amount of UTP (10 p ~ ) in every assay. The activity of EMC virus 3DP’ in the absence of inhibitors or analoes is taken as 100%. ND. not determined.

~ ~ ~ _ _ _ _ _ _ ~ ~~

A. Chemical inhibitors

EMC virus 3D”’ poly(U) E. coli DNA-dependent Inhibitor Concentration polymerase activity RNA polymerase

remaining activity remaining % %

Rifampicin 24 p M 100 0 Heparin 24 pglml 100 0 Cerulenin 100 pM <3 100 N-ethylmaleimide 5 mM <3 100 Dactinomycin 8 pglml 100 ND Brefeldin A 0.36 pM 100 ND Guanidine HCl 3 mM 100 ND 1,lO-Phenanthroline 1 mM 52 ND

10 mM 0 ND Aphidicolin 20 rglml 100 ND Phosphonoacetic acid 10 pglml 100 ND

B. Analogs of nucleoside 5”triphosphates

EMC virus 3 P ’ poly(U)

remaining Analog Concentration polymerase activity

P M % dTTP 40 100 ddTTP 160 100 3”dUTP 10 0 UTP-Z‘,B’-dialdehyde 100 48

10 100 UDP 100 100 UMP 100 100

guiri and Tamm, 1968; Tershak, 1982), had no effect on the in vitro activity of EMC virus 3Dpo1 (Table I). Similar results were reported for POliOVirUS 3Dp0’ in cell-free systems (Guinea and Carrasco, 1990).

Both poliovirus and EMC virus 3DP1 require Zn2+ ions for full enzymatic activity, indicating the involvement of Zn2+ ions in the polymerization reaction (Baron and Baltimore, 1982; Sankar and Porter, 1991). Many DNA and RNA polymerases examined to date have been found to be sensitive to 1,lO-phenanthroline (Auld et al., 1975; Kornberg, 1980; Mo- dak and Srivastava, 1979), and EMC virus 3DP1 is also sensitive to 1,lO-phenanthroline (Table IA), raising the possibil- ity that the enzyme may be a zinc metalloenzyme. However, the role of Zn2+ ions in catalysis needs to be clarified before classification of EMC virus 3DP’ as a zinc metalloenzyme can be considered.

Sensitivity to Nucleoside 5’-Triphosphate Analogs-It is well known that DNA polymerases are strongly inhibited by substrate analogs that function as chain terminators. For example, 3’-azido-2’,3’-dideoxythymidine inhibits HIV-1 reverse transcriptase and dideoxynucleoside 5 ’ -triphosphates are used as chain terminators in DNA sequencing with reverse transcriptases and DNA polymerases. Neither dTTP nor ddTTP in large excess of UTP had any effect on the poly(U) polymerase activity of EMC virus 3DP’, and UDP and UMP did not block the incorporation of ribonucleotides (Table IB). These findings, indicating that EMC virus 3DP0’ has a strict specificity for ribonucleoside 5’-triphosphates are strength- ened by the observation that the enzyme is inhibited by ribonucleoside 5’-triphosphate analogs which have their 3’- OH group either removed or modified, such as 3’-dUTP and UTP-2‘,3‘-dialdehyde (Table IB). These ribonucleoside 5’-

triphosphate analogs presumably compete for the substrate binding site and terminate RNA chains.

Rationale for Selection of Amino Acids for Mutagenesis-In the past 7 years, a number of investigators have performed sequence alignments of RNA-dependent RNA polymerases of bacterial, plant, and animal virus origin and have identified conserved motifs with varying degrees of conservation of amino acids (Kamer and Argos, 1984, Argos, 1988; Poch et al., 1989; Bruenn, 1991; Koonin, 1991). Some of these motifs (e.g. YGDD) have also been located in open reading frames coding for enzymes suspected of having RNA-dependent RNA polymerase activity (e.g. Argos, 1988, Lundblad and Black- burn, 1990). Moreover, two of the motifs (IV and VI; Fig. 1) have been found in the reverse transcriptases and all the members of the DNA-dependent DNA polymerase family (Delarue et al., 1990; Blanco et al., 1991).

Fig. 1 shows the positions within EMC virus 3DP1 of the various motifs conserved among the RNA-dependent RNA polymerases of positive strand, negative strand, and double- stranded RNA viruses. Motifs IV, V, and VI (Fig. 1) contain the most highly conserved amino acids, including 4 aspartic acids, 2 glycines, and a threonine, which are present in all known positive strand RNA virus RNA polymerases (Fig. 2). When the entire family of RNA-dependent RNA polymerases is aligned, three of these aspartic acids (D235, D333, and D334 in EMC virus 3Dp”; Fig. 2) and a glycine (G294 in EMC virus 3DP0’, Fig. 2) are almost invariant (Poch et al., 1989; Bruenn, 1991; Koonin, 1991).

We had previously predicted that the sequence LKRKFKKEGPLY, which overlaps motif VII, is the most antigenic region of EMC virus 3DP’, and antibodies raised against this peptide immunoprecipitate both native and re-

Putative Active Site of EMC Virus RNA-Dependent RNA Polymerase 10171

1

51

101

151

201

251

301

351

401

451

10 20 30 40 50

GALERLPDOP RIEVPFWU RpTVAWVFQ PAYAPAVLSK F D P R W

FIG. 1. Amino acid sequence of EMC virus 3P'. A summary of the published conserved amino acid sequence motifs in RNA- dependent RNA polymerases is shown in the context of the EMC virus 3Dp1 sequence (Motifs Z-VZZZ, Koonin (1991); Motif A-C, Poch et al. (1989); Motifs 1-4 and 6, Bruenn (1991). Highly conserved amino acids within motifs IV-VI are highlighted in boldface type (Fig. 2). Conserved motifs IV-VI1 (indicated by arrows) contain amino acids which were chosen for mutagenesis of EMC virus 3Dp1 (Fig. 2).

-7 215 2fO

E m : 3 D p o 1 R V Y D V P Y S N F P S T A S

RNA dep.RNA pol D S/T 0

(consensus)

"

M c 3 D p o 1 T G G L P S E C A A T S M L N T I M N N I L 290 2?5

370 305

RNA dep.RNA pol S/T G T N/E

(consensus)

"

EElc3Dpo1 V L S Y E P P L L Y

RNAdep.RNApo1 e e G D D m e (consensus)

310 375

RNA dep.RNA pol (consensus)

F/S K/R

FIG. 2. Summary of conserved and consensus amino acids within four RNA-dependent RNA polymerase motifs. Amino acids which are invariant or almost invariant in RNA-dependent RNA polymerases are given in boldface type (see also Fig. l ) , whereas preferred amino acids are given in the form of S/T, N/E , etc. 0 indicates that the consensus is hydrophobic residues, whereas W indicates that a majority of residues are hydrophobic. Underlined amino acids in EMC virus 3DPl were chosen for mutagenesis. These include 3 non-conserved amino acids to serve as controls (residues 364, 371, and 376; double-underlined). RNA dep. RNA pol, RNA- dependent RNA polymerase.

combinant EMC virus 3 P ' (Sankar and Porter, 1991). Therefore, this sequence probably occupies a surface location in the protein.

Amino acid residues within motifs IV-VI1 that are highly conserved or conserved in character were therefore chosen for mutagenesis (Fig. 2, underlined), and conservative substitutions were made in order to retain the character of the side chain (Table 11). As controls, substitutions of non-conserved amino acids (N371 and D376) were made close to motif VII, and a drastic change (T364K)' was also engineered within motif D (double-underlined in Fig. 2). Interestingly, according to two published alignments (Poch et al., 1989; Bruenn, 1991), this latter mutation replaces a non-conserved amino acid (T) with a highly conserved one (K).

Expression and Purification of Mutant Enzymes--Twenty mutant plasmids each capable of expressing a fusion protein consisting of glutathione S-transferase and a modified EMC virus 3DW' with a single amino acid substitution, were gener- ated as described under "Experimental Procedures" (Fig. 3). One construct, designed to change the partially conserved Tyr-331 to Phe in motif VI, was not obtained as the plasmid was apparently unstable.

E. coli HBlOl cells harboring the parental plasmid pT12 or mutant plasmids were induced with isopropyl-l-thio-fl-D-galactoside and cell-free extracts prepared. The glutathione S- transferase-polymerase fusion proteins were absorbed onto glutathione-Sepharose 4B, and the modified polymerases were purified to greater than 95% homogeneity as previously described (Sankar and Porter, 1991). Fig. 4 shows that the mobility of each mutant enzyme in SDS-polyacrylamide gels was indistinguishable from that of the wild-type recombinant enzyme.

Poly(U) Polymerase Assay of EMC Virus 3LF" Mutants- The sensitive poly(A)-dependent oligo(U)-primedpoly(U) polymerase assay was used to quantitate the effect of each mutation on RNA chain initiation and/or polymerization activity (Sankar and Porter, 1991). Equivalent amounts of purified wild-type and mutant enzymes were assayed in the presence of rifampicin, which would inhibit any contribution to RNA polymerase activity from E. coli DNA-dependent RNA polymerase (Table IA). The activities of the mutant polymerases are presented in Table I1 and expressed as a percentage of the activity of the wild-type enzyme.

Of the 17 mutant polymerases with amino acid substitutions in conserved motifs IV-VII, eight exhibited significantly lower enzymatic activity, and a further five were devoid of detectable activity (Table IIA). The 17 mutants with conservative amino acid changes can be arbitrarily grouped into three classes based on the observed reduction in poly(U) polymerase activity. One class has no change in activity, the second has 13- 43%, and the third has 0-6% of the activity of the wild-type enzyme (Table IIA). With the exception of the L380I mutant, none of the mutant polymerases with 0-6% activity involve substitution of hydrophobic amino acids, and all the mutations in the 13-43% class are conservative substitutions of hydrophobic residues.

When a particular amino acid is substituted, is there a correlation between the extent of reduction in polymerase activity and its degree of conservation among the RNA- dependent RNA polymerases? There are 7 residues which are invariant or almost invariant: these are D235, D240, G294,

Protein mutations are abbreviated using the following convention; the residue number is preceded by the symbol (in the one-letter code) for the wild-type amino acid and followed by the symbol for the mutant amino acid. Thus D235E denotes a mutation from Asp to Glu at position 235.

10172 Putative Active Site of EMC Virus RNA-Dependent RNA Polymerase

TABLE I1 Effects of single amino acid substitutions on the poly(U) polymerase activity of EMC virus 3LP'

Equivalent amounts of each mutant polymerase (0.6 pg) were added to the poly(U) polymerase assay. The polymerase activity of wild-type EMC virus 3DW' is taken as loo%, and the values given in column 3 are the mean values of at least five independent, duplicate determinations. Conserved and non-conserved refers to comparisons made with the RNA-dependent RNA polymerases of positive strand viruses (Poch et al., 1989; Bruenn, 1991; Koonin, 1991).

Plasmid (pSD series, Fig. 3)

EMC virus 3DP1 Location and POlY(U)

nature of polymerase mutant activitv

Motif no. Equivalent region in E. coli (Figs. 1 and 2) Klenow (Fig. 5)

remainine

% ~ ~~~ ~ ~ _ _ _ _ _ _ _ _ _

A. Mutagenesis of conserved residues pSD701 D235E 2 IV D705 pSD702 D240E 3 E710: M e binding" pSD703 S293T 6 V Y766 dNTP binding (primer

pSD704 G294A 0 G767 pSD705 T298S 0 F771 pSD706 M300I 43 P847 pSD707 N302Q 6 Q849 catalysis/DNA template

pSD708 I304L 30 T851 pSD709 M305L 30 A852 pSD710 I308L 100 I855 pSD711 I309L 13 I856 pSD712 G332A 0 VI H881: M e - d N T P binding" pSD713 D333E 3 pSD714 D334E 0 E883 pSD715 V337L 100 F886 pSD717 L380I 0 VI1 pSD718 K381R 100

? ?

3'-end?)"

binding"

D882: catalysis"

B. Mutagenesis of non-conserved resi-

pSD716 dues

T364K 100 ? ? pSD719 N371Q 100 ? ? pSD720 D376E 100 VI1 ?

Klenow residues for which a role in catalysis, such as dNTP, M e , or template binding has been established.

T298, G332, D333, and D334 (Fig. 2). When these amino acids were substituted, the remaining polymerase activity was in the range of 0-3%. As previously mentioned, D235, G294, D333, and D334 are highly conserved not only among the RNA-dependent RNA polymerases but also among the RNA- dependent DNA polymerases (Poch et al., 1989; Delarue et al., 1990). Surprisingly, the isomeric substitution L380I within motif VI1 resulted in undetectable polymerase activity. How- ever, the existence of this motif is controversial as the published alignments differ in this region (Poch et al., 1989; Bruenn, 1991; Koonin, 1991).

Substitutions at the less well conserved amino acids S293 and N302 resulted in 6% of the activity of the wild-type enzyme, and substitution of 4 of the 6 residues which are conserved only in hydrophobic character resulted in 13-43% activity.

As controls, one non-conservative and two conservative changes were made in or close to motif VI1 (Table IIB). None of these changes had any effect on polymerase activity.

DISCUSSION

There has been renewed interest in RNA-dependent RNA polymerases. Firstly, the enzyme is unique to viruses, and there is no known cellular homolog; therefore, this class of enzyme is an attractive target for antiviral chemotherapy. However, the three-dimensional structure of an RNA-dependent RNA polymerase remains to be determined, and the few mutagenesis studies which have been performed on RNA template-dependent polymerases have yet to yield a model of their active sites and a definition of the critical residues for catalysis and template binding (Inokuchi and Hirashima,

1987; Larder et al., 1987; Jablonski et al., 1991). Secondly, from an evolutionary point of view, RNA-dependent RNA polymerases were probably among the first enzymes, and it would be interesting to see how the structure and function of the DNA-dependent DNA and RNA polymerases and reverse transcriptases have evolved.

Inhibitor Profiles-Our inhibitor and substrate analog studies show that EMC virus 3DPo1 is in a class distinct from the reverse transcriptases and bacterial DNA-dependent RNA polymerases. The fungal antibiotic cerulenin was surprisingly found to be a moderately potent inhibitor of EMC virus 3DPo1 but not E. coli RNA polymerase. The structure of this compound, comprising both reactive hydrophilic and hydrophobic domains, does not, unfortunately, provide any indication of its mode of action. In addition, using a variety of substrate analogs, it was shown that EMC virus 3DwL lacks reverse transcriptase activity under the conditions of the poly(U) polymerase assay, although curiously oligo(dT) was as effi- cient as oligo(U) in priming the synthesis of poly(U) when added at the same concentration (Sankar and Porter, 1991).

Inhibitors can also distinguish EMC virus 3DPo1 from the DNA polymerase family. As expected, aphidicolin and phosphonoacetic acid, which are inhibitors of eukaryotic DNA polymerase a and 6 as well as DNA polymerase a-like viral polymerases, had no effect on the polymerization activity of EMC virus 3Dw' (Table I).

Site-directed Mutagenesis-In order to purify and charac- terize a large number of mutants, we exploited a rapid purification procedure for EMC virus 3DPo1 in which a glutathione S-transferase-3DPoL fusion protein is isolated by affinity chromatography, and EMC virus 3DPo' released by thrombin diges-

Putative Active Site of EMC Virus RNA-Dependent RNA Polymerase

Barn HI

10173

hol

EcoR V \

1 .Xho /-Barn HI digestion 2. Isolate 1.904kb fragment

containing EMC virus 3DPo1

, peI"..C.lpl us.

/" I- Barn HI digestion

I Ligate

t Site-Directed Mutagenesis an Sequence Analysis

Xho I-Barn HI digestion of pSS 9- 00 derivatives to release 1.904 kb mutated fragment of EMC virus 3DPOl cDNA

Ligate to 4.796 kb fragment derived from parental plasmid pT12 digested with Xho I and Barn HI

Generation of expression plasmlds containing the mutated EMC virus 3DPo1 cDNAs (pSD series)

$.

$.

FIG. 3. Construction of expression plasmids with mutations in EMC virus 3D"'. The subcloning strategy for generating single point mutations in EMC virus 3Dm1 is shown. Oligonucleotide-directed mutagenesis was performed with single-stranded DNA of plasmid pSS700, and the resultant mutant 3Dm1 genes were subcloned to give the pSD series of plasmids (Table 11), which are equivalent to pT12. Ori, origin of replication; Amp r, ampicillin resistance gene; GST, glutathione S-transferase; tac, trp-lac hybrid promoter; MCS, multiple cloning site.

tion (Smith and Johnson, 1988 Sankar and Porter, 1991). This procedure was successful for all 20 mutant genes that were constructed, enabling an accurate determination of the poly(U) polymerase activity of each mutant.

Eighteen amino acid positions within EMC virus 3DP0' were chosen for mutagenesis based on sequence alignments of all known RNA-dependent RNA polymerases that identified cer- tain residues as being highly or well conserved. Of the 17 mutants constructed, expressed, and purified, only three re-

tained full polymerase activity (I308L, V337L, and K381R) and four had 13-43% of the activity of the wild-type enzyme (M3001, I304L, M305L, and 1309L). Most strikingly, 10 functionally conservative substitutions within motifs IV-VI1 (Figs. 1 and 2) either abolished enzymatic activity or reduced activity by more than 10-fold, and only one of these substitutions (L380I in motif VII) involved mutation of a hydrophobic amino acid (Table 11). Mutation of 1 non-conserved to a conserved residue and substitution of 2 non-conserved

4 8 2 kd

FIG. 4. SDS-PAGE analysis of purified mutants of EMC virus 3D"'. The purity of each mutant was estimated by electrophoresis in a 0.1% SDS, 10% polyacrylamide gel followed by staining with Coomassie Blue. WTis unmodified EMC virus 3DP"'. All mutants except the controls N371Q and D376E are shown. The positions of molecular mass markers in kilodaltons are shown on the left.

residues close to motif VI1 had no effect. Overall, there is a good, direct correlation between the sensitivity of amino acids to mutation and their degree of conservation among the RNA- dependent RNA polymerases, strongly suggesting that the highly conserved residues fulfill important roles in polymerase structure and function.

This assertion is supported by several mutagenesis studies on other RNA-dependent RNA polymerases, and HIV-1 reverse transcriptase and various DNA polymerases, in con- junction with recent alignments of RNA and DNA polymerase coding regions. In particular, it has been speculated that two of the loop regions which together form part of the active site of E. coli DNA polymerase I (Klenow) correspond to motifs IV and VI of the RNA-dependent RNA polymerases (Delarue et al., 1990). Together, these studies allow the identification within these diverse groups of polymerases of functionally important counterparts of some of the conserved amino acids in motifs IV-VI of EMC virus 3Dp' as follows.

MotifIV-This motif (Fig. 1) has been found in the reverse transcriptase group as well as in various DNA and RNA polymerases from prokaryotes and eukaryotes (Poch et al., 1989; Delarue et al., 1990). Both the mutations D235E and D240E within this region of EMC virus 3DW1 resulted in low levels of enzyme activity. In HIV-1 reverse transcriptase, the analogous sequence is DVGDAY, where some mutations, including substitutions of the 2 Asp residues and the Tyr residue significantly reduced reverse transcriptase activity and pro- foundly altered the sensitivity of the mutants to drugs that act as substrate analogs (Larder et al., 1987, 1989). Addi- tionally, mutation of an Asp residue in Brome mosaic virus 2a polymerase (corresponding to D240 in EMC virus 3Dw') resulted in complete loss of activity (Kroner et al., 1989).

In EMC virus 3DP0', the sequence (235)DYSNFD(240) within motif IV aligns with (705)DYSQIE(710) in E. coli DNA polymerase I (Klenow), a protein of known three- dimensional structure (Ollis et al., 1985; Delarue et al., 1990; Blanco et al., 1991). Both Q708 and E710 in Klenow have been reported to bind the M e ion during substrate binding (Blanco et al., 1991), and lie in a loop (turn) region between a @-sheet and an a-helix (Ollis et al., 1985; Fig. 5). A similar loop region is strongly predicted in the viral RNA-dependent RNA polymerases, including EMC virus 3DP0' (Poch et al., 1989; Delarue et al., 1990): D240 in EMC virus 3DP0' corre-

S. Sankar and A. G. Porter, unpublished data.

RNA-Dependent RNA Polymerase

FIG. 5. Localization of amino acids in the structure of E. coli DNA polymerase I (Klenow) which correspond to substitutions affecting the polymerase activity of EMC virus 3Dw'. The representation of the three-dimensional structure of Klenow is from Joyce and Steitz (1987) with minor alterations. Regions which form the a-helices are represented by lettered cylinders, and those forming ,%sheets are given as stippled, numbered arrows. Connecting turns and unstructured regions are represented by doubled lines. Filled triangles (V) and sqwres (W) mark the positions of Klenow residues which correspond to conservative changes in EMC virus 3DW1 resulting in 0-6% and 13-30% of wild-type poly(U) polymerase activity, respectively (Table 11) (reproduced with permission from Elsevier Science Publishers, London).

sponds to E710 in Klenow, and D235 rather than N238 may be analogous to Q708 in Klenow, because N238 is not conserved among RNA-dependent RNA polymerases (Poch et al., 1989). Thus, 1 or both Asp residues in EMC virus 3DW1 may bind the Mg2+ ion during catalysis. Other enzymes of known tertiary structure where an Asp or Glu (or Gln) residue or both bind a metal ion directly are phosphoglycerate kinase (Watson et al., 1982), concanavalin A (Hardman et dl., 1972), E. coli elongation factor EF-Tu (Cour et al., 1985), and ras p21 (McCormick et al., 1985).

The well conserved YGDTD motif present in many a-like viral and cellular DNA polymerases was originally aligned with the GDD sequence of motif VI (Delarue et al., 1990). More recently, a convincing general structure for all classes of DNA-dependent DNA polymerases has been proposed, and YGDTD now aligns with (D)YSNFD of motif IV in EMC virus 3DW' and (D)YSQIE in Klenow, such that the 2 Asp residues (in YGDTD) correspond to Q708 and E710 of Klenow (Blanco et al., 1991). Good evidence for the participation of this region in Mg2+ ion and substrate binding comes from analysis of site-directed mutants of adenovirus DNA polymerase, phage 629, and herpes simplex virus DNA polymerases in vitro or in uiuo (Bernad et al., 1990; Dorsky and Crum- packer, 1990; Marcy et al., 1990; Joung et al., 1991). In general, both conservative and non-conservative substitutions within this core sequence reduced polymerase activity, particularly replacement of either Asp residue. In the case of the herpes simplex virus polymerase, some mutations in this region conferred altered sensitivity to substrate analogs (Marcy et al., 1990), which is consistent with an important role for this site in binding the M e - d N T P complex. Furthermore, the YGDTD sequence is consistently predicted to lie in a turn region as in Klenow (Marcy et al., 1990; Delarue et al., 1990).

Motif VI-The GDD core of motif VI is almost completely conserved in the RNA-dependent RNA polymerases, and most 3D polymerases of picornaviruses have the sequence YGDD (Poch et al., 1989). In EMC virus 3Dw1, replacements

Putative Active Site of EMC Virus RNA-Dependent RNA Polymerase 10175

within the GDD core either destroyed or markedly reduced polymerase activity (Table 11). When the homologous Gly residue of two other RNA-dependent RNA polymerases (QP replicase and poliovirus 3DPO') was mutated, polymerase activity was also reduced or lost (Inokuchi and Hirashima, 1987; Jablonski et al., 1991). In HIV-1 reverse transcriptase, motif VI corresponds to region E (YMDD) where mutation of any of these amino acids greatly affected reverse transcriptase activity, and a Tyr to Ser mutant was less sensitive to the substrate analogs AZT triphosphate and phosphonoformate (Larder et al., 1987, 1989; Le Grice et al., 1991).

In EMC virus 3Dp0', (331)YGDD(334) aligns with (880)VHDE(883) of Klenow, which occurs in a tight turn between two @-sheets (Ollis et al., 1985; Fig. 5). Secondary structure predictions of the region containing the GDD core in both RNA and DNA polymerases suggest that a tight turn exists in the same location as in Klenow (Poch et al., 1989; Delarue et al., 1990).3 In EMC virus 3Dp1, G332 may be important in forming this turn, since an alanine residue has a lower probability of participating in turn structures (G332A; 0% activity). G332 aligns with H881 of Klenow, a residue which may be involved in dNTP binding (Pandey et al., 1987). In Klenow, D882 fulfills an essential role in catalysis independent of DNA template binding (Polesky et al., 1990). Since D882 aligns with D333 of EMC virus 3DPo1 (D333E; 3% activity), a residue which is totally conserved in all polymerases (Poch et al., 1989; Delarue et al., 1990; Blanco et al., 1991), it is likely that D333 in EMC virus 3DP' likewise fulfills an essential role in catalysis. The role, if any, of E883 in Klenow remains to be established. E883 corresponds to D334 in EMC virus 3DPo1 where substitution with a Glu residue completely destroyed activity (Table 11). E883 also corresponds to Dl86 in the YMDD motif in HIV-1 reverse transcriptase, where substitution with an Asn residue almost completely destroyed activity (Larder et al., 1989; Le Grice et al., 1991). We speculate that D334 in EMC virus 3DP0' either participates in binding the M e - N T P complex, perhaps by interacting with the metal ion, or else it is important for turn formation in the RNA template-dependent polymerases (Poch et al., 1989).

Motif V-This motif is characterized by SGXXXT in the RNA-dependent RNA polymerases, where the SG is predicted to lie in a turn region preceded and followed by @sheets (Poch et al., 1989; Bruenn, 1991; Koonin, 1991). Since no mutagenesis of the conserved residues within this motif have been reported, we have carried out an extensive mutational analysis of EMC virus 3DW' and found that the S293T, G294A, T298S, and N302Q mutants had 0-6% polymerase activity. Does this important region have a structural counterpart in the Klenow structure? Because motif V lies between motifs IV and VI in the RNA-dependent RNA polymerases, we looked for homol- ogies between motif V in EMC virus 3DW1 and the region between Klenow residues D705 and E883, which corresponds to motifs IV-VI (Fig. 5). Using the PALIGN program of Myers and Miller (PCGene), it was surprisingly found that motif V aligns with widely separated domains in Klenow (amino acids 760-771 and 847-859; Fig. 6). S293T corresponds t o Y766 in Klenow, which has a critical role in dNTP binding and interaction with the newly synthesized 3"terminus (Joyce and Steitz, 1987; Polesky et al., 1990). Both the Ser and the Tyr have a hydroxyl group in their side chains which could participate in similar hydrogen bonding interactions. G294 is homologous with G767, the start of a long turn/coil region in Klenow (Figs. 5 and 6 ) , so perhaps the G294A mutation in EMC virus 3Dp"' prevents the formation of a turn. In the distal portion of motif V from EMC virus 3DP"'

EMC3DPo1 I T G G L P S G C A A T S

Klenow 0. 760 F b ; j l s u p 2 :

765 770

3pO 3p5

EMC3DP1 M L I T I M N N I I I R A

Klenow P M Q G T A A D I I K R A , o"0 850 855

FIG. 6. Predicted alignment of EMC virus 3DW' conserved motif V with E. coli DNA polymerase I (Klenow). The sequence from EMC virus 3D"' amino acids 287-312 (Motif V, Fig. 1) was aligned by computer with Klenow amino acids 760-772 (end of 0 helix, Fig. 5 ) and 847-859 (Q helix, Fig. 5 ) . Identical residues are boxed, and positions which correspond to Klenow residues known to be important in catalysis are given in boldface type (see also Table 11).

amino acids 300-312, N302Q aligns with Q849 in Klenow (Fig. 6). Since Asn and Gln are functionally equivalent, and Q849 in Klenow participates in template binding (Polesky et al., 1990), it is worth speculating that N302 has an analogous role in RNA template binding in EMC virus 3DPo1.

Motif VZZ-This less well conserved motif has no counterpart in Klenow, and the reason for the complete loss of polymerase activity resulting from the L380I mutation remains to be determined.

The Putative EMC Virus 3 P ' Catalytic Center-From the above discussion, it is evident that two functionally important motifs (IV and VI) are present in both DNA and RNA polymerases, and we propose that a third (motif V) contains amino acids which have counterparts in E. coli Klenow (Fig. 6). We have placed the positions of the point mutations within these three motifs which have the most drastic effect on the polymerase activity of EMC virus 3DPo1 in the context of the three-dimensional structure of Klenow (Fig. 5). All these positions map within the cleft forming the active site of Klenow, whereas mutations which result in intermediate polymerase activity lie just outside the active site (Fig. 5). In particular, the active site amino acids E710, Y766, Q849, H881, and D882 in Klenow correspond in D240, S293, N302, G332, and D333 in EMC virus 3DW', respectively (Table 11).

Our results provide compelling evidence that a basic structural and functional framework is conserved in the most distantly related classes of nucleic acid polymerases, and will encourage molecular modeling of the active sites of RNA template-dependent RNA and DNA polymerases on the structure of E. coli Klenow.

Acknowledgments-We thank Drs. C. Berry and A. Kush for crit- ically reading the manuscript, Diana Lee for oligonucleotide synthesis, Francis Leong for photography, and M. L. Chua for typing the manuscript. We are also grateful to Drs. S. H. Oon and B. Li for technical advice and stimulating discussions.

REFERENCES Argos, P. (1988) Nuckic Acids Res. 16 , 9909-9916 Auld, D. S., Kawaguchi, H., Livingston, D. M., and Vallee, B. L.

Baron, M. H., and Baltimore, D. (1982) J. Biol. Chem. 257, 12359-

Bernad, A., Lazaro, J. M., Salas, M., and Blanco, L. (1990) Proc.

Blanco, L., Bernad, A., Blasco, M. A., and Salas, M. (1991) Gene

Bruenn, J. A. (1991) Nucleic Acids Res. 19, 217-226

(1975) Biochem. Biophys. Res. Conmun. 62, 296-302

12366

Natl. Acad. Sci. U. S. A. 87, 4610-4614

(Amst.) 100,27-38

10176 Putative Active Site of EMC Virus RNA-Dependent RNA Polymerase

Calguiri, L. A., and Tamm, I. (1968) Virology 35,408-417 Cour, T. F. M., Nyberg, J., Thirup, S., and Clark, B. F. C. (1985)

Dasgupta, A., Baron, M. H., and Baltimore, D. (1979) Proc. Natl.

Dasgupta, A., Zabel, P., and Baltimore, D. (1980) Cell 19,423-429 Delarue, M., Poch, O., Tordo, N., Moras, D., and Argos, P. (1990)

Dmitrieva, T. M., Norkina, K. B., and Agol, V. I. (1991) J. Virol. 65,

Dorsky, D. I., and Crumpacker, C. S. (1990) J. Virol. 6 4 , 1394-1397 Flanegan, J. B., and Baltimore, D. (1977) Proc. Natl. Acad. Sci. U. S.

Guinea, R., and Carrasco, L. (1990) EMBO J. 9, 2011-2016 Hardman, K. D., and Ainsworth, C. F. (1972) Biochemistry 11,4910-

Inokuchi, Y., and Hirashima, A. (1987) J. Virol. 61,3946-3949 Jablonski, S. A., Luo, M., and Morrow, C. D. (1991) J. Virol. 65,

Joung, I., Horwitz, M. S., and Engler, J. A. (1991) Virology 184,235-

EMBO J. 4,2385-2388

Acad. Sci. U. S. A. 76,2679-2683

Protein Eng. 3,461-467

2714-2717

A. 74,3677-3680

4919

4565-4572

Marcy, A. I., Hwang, C. B. C., Ruffner, K. L., and Coen, D. M. (1990) J. Virol. 6 4 , 5883-5890

McCormick, F., Clark, B. F. C., Cour, T. F. M., Kjelgaard, M., Narskov-Lauritsen, L., and Nyberg, J. (1985) Science 230 , 78-82

Misumi, Y., Miki, K., Takatsuki, A., Tamura, G., and Ikehara, Y. (1986) J. Bwl. Chem. 2 6 1 , 11398-11403

Modak, M. J., and Srivastava, A. (1979) J. Biol. Chem. 2 5 4 , 4756- 4759

Morrow, C. D., Warren, B., and Lentz, M. R. (1987) Proc. Natl. Acad. Sci. U. S. A. 84,6050-6054

Nicklin, M. J. H., Toyoda, H., Murray, M. G., and Wimmer, E. (1986) BiolTechnology 4,33-42

Ollis, D. L., Brick, P., Hamlin, R., Xuong, N. G., and Steitz, T. A. (1985) Nature 313, 762-767

Palmenberg, A. C., Kirby, E. M., Janda, M. R., Drake, N. L., Duke, G. M., Potratz, K. F., and Collett, M. S. (1984) Nucleic Acids Res.

Pandey, V. N., Williams, K. R., Stone, K. L., and Modak, M. J. (1987)

Plotch, S. J., Palant, O., and Gluzman, Y. (1989) J. Virol. 6 3 , 216-

Poch, O., Sauvaget, I., Delarue, M., and Tordo, N. (1989) EMBO J.

12,2969-2985

Biochemistry 26,7744-7748

225

Kornberg, A. (1980) DNA Replication, pp. 67-123, W. H. Freeman, Rothstein, M. A., Richards, 0. C., Amin, C., and Ehrenfeld, E. (1988) San Francisco Virology 164,301-308

Kroner, P., Richards, D., Traynor, P., and Ahlquist, P. (1989) J. S a n k s., and Porter, A. G. (1991) J. viral. 66,2993-3000 Virol. 63,5302-5309 Smith, D. B., and Johnson, K. S. (1988) Gene (Amst.) 6 7 , 31-40

Laemmli, U. K. (1970) Nature 227,680-685 Tershak, D. R. (1982) J. Virol. 41,313-318 Larder, B. A., purifoy, D. J. M., powell, K. L., and Darby, G. (1987) Traub, A., Diskin, B., Rosenberg, H., and Kalmar, E. (1976) J. V i d .

Larder, B. A., Kemp, S. D., and Purifoy, D. J. M. (1989) Proc. Natl. chm. 2 5 7 , 4610-4617 Van Dyke, T. A., Rickles, R. J., and Flanegan, J. B. (1982) J. Biol.

Le Grice, s. F. J., Naas, T., Wohlgensinger, B., and Shatz, 0. (1991) p. L., Fothergill, L. A., Perkins, R. E., Conroy, S. C., Dobson, M. Watson, H. C., Walker, N. P. C., Shaw, P. J., Bryant, T. N., Wendell,

Lundblad, V., and Blackburn, E. H. (1990) Cell 60,529-530 EMBO J. 1,1635-1640

Nature 327 , 716-717

Acad. Sci. U. S. A. 86,4803-4807

18,375-382

EMBO J. 10,3905-3911 J. Tuite, M. F., Kingsman, A. J., and Kingsman, S. M. (1982)

Documents

Point Mutations Which Drastically Affect the Polymerization Activity