VIROLOGY

186,725-735

(1992)

Characterization

of the Roles of Conserved Cysteine Residues in Poliovirus 2A Protease

and Histidine

SHUYUARN F. YU AND RICHARD E. LLOYD’ Department

of Microbiology and Immunology, University of Oklahoma Health Sciences Center, P. 0. Box 2690 1, Oklahoma City, Oklahoma 73 190 Received August

14. 199 1; accepted

October 30, 199 1

The primary processing of the poliovirus polyprotein is catalyzed by 2A protease (2Ap’“) which cleaves at the 1 D/2A junction in a very rapid cotranslational reaction. In addition, 2APm also indirectly induces cleavage of the ~220 component of elF-4F, which results in selective inhibition of host protein synthesis. Earlier studies have indicated that 2AP” is related to 3C protease (3CPro) and is structurally similar to trypsin-like serine proteases with the substitution of CyslO9 as the nucleophile. We noticed that 2AP”’ of enteroviruses and rhinoviruses contains a specific motif of Cys55-XaaCys57-Xaa,-Cysll5-Xaa-His117 which is absolutely conserved, but which is not found in viral 3Cp” or known cellular serine proteases. To better understand the specific roles these conserved cysteine and histidine residues played in the structure/function of 2APm, we constructed a series of 2APm mutants by site-specific mutagenesis and analyzed the mutant enzymes with respect to their biochemical properties. Conservative amino acid replacements at Cys55, Cys57, Cysl15, or His1 17 resulted, in each case, in a complete loss of both in cis and in frans activities of 2APro. To determine the function of these residues, we examined the biochemical/structural features of 2AP” expressed in a cell-free rabbit reticulocyte lysate system. Gel mobility shift and chemical modification data suggest that these cysteine residues do not form intra-molecular disulfide linkages as a structural feature of 2A p’o. However, studies with metal chelators did not eliminate the possibility that 2APro contains a metal-binding ligand. Finally, our results suggest that these conserved cysteine and histidine residues, including Cys55, Cys57, Cys115, and His1 17, are critical in maintaining the active conformation of 2Ap” structure and essential in supporting the catalytic activity of 2AP”. o 1992 Academic PESS, I~C.

INTRODUCTION

ity is believed to contribute to the specific inhibition of host cell protein synthesis that occurs rapidly after poliovirus infection of HeLa cells. Since direct involvement of 2APro in the cleavage of ~220 could not be demonstrated (Lloyd eta/., 1986) it was proposed that cleavage of ~220 was catalyzed by a latent, cellular p220-specific proteinase (p220ase) which was activated or induced by 2APfoafter viral infection. However, since p220ase has not been purified and characterized, it is still unclear whether 2APro activates p220ase via a proteolytic event or some other mechanisms. Computer alignments of 2APro amino acid sequences have found that a common consensus sequence around the active center Cys residue of picornavirus 3Cpro is highly conserved among enterovirus and rhinovirus 2AProas well as certain serine proteases (Argos et al., 1984; Blinov et al., 1985; Lloyd et al., 1986). However, in these alignments, the nucleophilic serine residue of serine proteases was replaced with cysteine in the viral 2AP”’ or 3Cpro. When the predicted secondary and tertiary structures of 2APro and 3Cpro were compared with the known structures of chymotrypsin-like proteases, it showed that they were both related to the same class of trypsin-like cellular serine proteases (Bazan et al., 1988; Gorbalenya eta/., 1989). In particular, HisilO, Asp38, and CyslO9, which are all

Poliovirus, a member of the picornavirus family, is a small, positive-stranded RNA virus. The primary translation product of poliovirus is a single large polyprotein which is processed into mature viral proteins by two viral gene products, 2APro and 3Cpro (Hanecak et a/., 1982; Toyoda et al., 1986). 2AProspecifically cleaves at tyrosine-glycine (Y-G) bonds in the polyprotein while 3Cpro or its precursor 3CD carries out cleavage at glutamine-glycine (Q-G) sites (Emini et al., 1982; Larsen et a/., 1982). The initial processing of the viral polyprotein is catalyzed by 2APro which cleaves at the junction of 1D and 2APro in a very rapid cotranslational reaction (Nicklin et a/., 1987; Toyoda et a/., 1986). Most of the other cleavages are then carried out by 3Cpro or one of its precursors (3CD) (Jore et al., 1988; Ypma-Wong et al., 1988). In addition to autocatalytic cleavage activity, 2APro also indirectly induces cleavage of the ~220 component of the eucaryotic translation initiation factor, elF4F (formerly called cap-binding protein complex) (Bernstein et al., 1985; Krausslich et al., 1987; Lloyd et al., 1988). The loss of functional cap binding protein activ’ To whom reprint requests should be addressed. 725

0042-6822192

$3.00

Copynght Q 1992 by Academvz Press, Inc. All rights of repraductlon I” any form reserved.

726

YU AND LLOYD

highly conserved among different picornavirus 2APro, could be superimposed on the equivalent trypsin-like serine protease catalytic triad, His57, Aspl02, and Ser195 (Bazan et al., 1988). The recent results from mutational analyses of 2APro (Hellen et a/., 1991; Yu et a/., 1991) have also suggested that the highly conserved His20, Asp38, and CyslO9 residues may comprise the complete catalytic triad of 2AP” and thus supported the hypothesis that poliovirus 2APro is structurally similar to the tr-ypsin-like family of serine proteases with the substitution of cysteine 109 as the active site nucleophile. When several different protease inhibitors were tested previously in a direct in Vans cleavage assay for 2APro activity (Vu et al., 1991), it was found that iodoacetamide and N-ethylmaleimide could effectively inactivate the in trans activity of 2APro. Both of these compounds are alkylating agents which would modify free cysteine residues exposed on the surface of proteins, thus suggesting that 2AProcontains a critical cysteine residue, most likely in the active site. However, phenylmethanesulfonyl fluoride (PMSF), an inhibitor of serine proteases, unexpectedly inhibited 2AProactivity. Most of the remaining serine orcysteine protease inhibitors tested had marginal or no inhibitory effects on 2APro activity. These inhibitors, including E64 (transbutane), epoxysuccinyl-L-leucyl-amino(4-guanidino) (~-1 chloro-3-(4-tosylamido)-7-amino-2-hepaTLCK tone), and TPCK (~-1 -chloro-3-(4-tosylamido)4-phenyl2-butanone), resemble transition intermediates of specific substrates and may not meet the unique substrate specificity requirements of 2APro. All of these findings as well as the fact that no enzymatic activity was detected after replacement of active site CyslO9 by Ser in 2APro(Yu eta/., 1991) indicated that evolutionary modification of the nucleophilic serine to cysteine must have been accompanied by other specific amino acid changes which preserved activity and which formed a novel active site environment. 2APro from the poliovirus type 1, Mahoney strain, consists of 149 amino acids in which five out of six cysteine residues, Cys55, Cys57, Cys64, CyslO9, and Cysl 15, are highly conserved among different picornaviruses. Among these, only Cys64 is replaced with Ser in the 2APro of rhinovirus type 14 while the other cysteines are absolutely conserved in enteroviruses and rhinoviruses. Interestingly, except for the active site CyslO9, all the other conserved cysteine residues found in 2AProare not found in related 3Cpro or trypsinlike cellular serine proteases. This strongly implies that the different geometries of 2APro and 3CPro structures may contribute to their exclusive substrate specificity and enzymatic activity. Additionally, 2AProcontains only

two conserved histidine residues, His20 and His1 17. We have previously studied His20, which is proposed to comprise part of the catalytic triad (Vu et al., 1991). In this study, we also analyzed the effect of site-directed mutagenesis of His1 17. To understand the specific roles of these conserved cysteine and histidine residues played in the structure and/or function of 2APro, we constructed a series of 2APro mutants in which each conserved Cys or His was replaced by other amino acids and then analyzed these mutant 2APro with respect to their biochemical properties. We have also examined several structural features of 2APro expressed in a cell-free rabbit reticulocyte lysate system. In the study reported here, we provide experimental evidence showing that these highly conserved cysteine and histidine residues, including Cys55, Cys57, Cysl15, and His1 17, are essential to support the catalytic activity of 2APro.

MATERIALS AND METHODS Site-directed mutagenesis and DNA sequencing Eleven point-mutations were individually introduced into the plasmid pEP2A by the oligonucleotide-directed mutagenesis protocol as described previously (Yu et a/., 1991). Oligonucleotide primers specifying singleor double-base changes were synthesized and used in the mutagenesis reactions at each site as follows: 5’-CGlTGCAAlT~CCTTGC-3’ at Cys55; 5’-CTGCGlTGG(T)A(T)AlTGCACC-3’ at Cys57; 5’-CTAGACTCGG(T)A(T)GTAGTACACC-3’ at Cys64; 5’-CGTGGTGAGOA(G)TCTGAGTATGC-3’ at Cysl15; and 5’CACCCCCTGGTGACATCTG-3’ at His 117. Bacteria containing mutant plasmids were first screened by differential colony hybridizations with the appropriate mutagenic oligomers and positive plasmids were subsequently sequenced across the 2A gene region flanking the target site to confirm the presence of the specific mutation and absence of unwanted second site mutations.

Construction of plasmids The plasmid pEPmet2A was derived from the plasmid pEP2A (Vu et a/., 1991) by using the polymerase chain reaction (PCR) technique to delete the entire coding sequence for the leader gene portion of EMCV and the entire 1C’ and 1D genes of poliovirus. After PCR reactions with two primers (5’-CACGATGATTATATGGGATTCGGAC-ACC-3’ and 5’-GGTGTCCGAATCCCATATAATCATCGTG-3’) the 2A gene was placed immediately after the AUG start codon. The predicted translation product of the plasmid pEPmet2A

POLIOVIRUS

thus contained a mature 17-kDa 2APro beginning with the engineered methionine at position -1. To transfer the specific mutant sequence from mutagenized pEP2A derivatives, a 622-bp Pstl fragment containing the new start codon from the wild-type pEPmet2A was inserted into the 3453-bp Pstl fragment of mutagenized pEP2A DNA upstream of the bulk of the 2A sequence, thus replacing the original 9637-bp Pstl fragment. For example, the plasmid pEPmet2A(C55S) was constructed by inserting a 622-bp Pstl fragment from the wild-type pEPmet2A into the 3453-bp Pstl fragment of the mutant plasmid pEP2A(C55S).

In vitro transcription

wild-type pEPmet2A was translated in the absence of radioactive amino acids to produce unlabeled 2APro (enzyme). After reactions were stopped by the addition of RNase A (10 mg/ml), cycloheximide (0.5 mg/ml), and cold L-methionine (1 mn/r), a sample of the unlabeled 2APro was mixed with an equal volume of the reaction mixture containing the labeled, uncleaved precursor substrate derived from the plasmid pEP2A(C57S). Mixtures were then incubated at 30” overnight and analyzed by SDS-polyacr-ylamide gel electrophoresis and autoradiography. Intensities of precursor and processed product bands were quantitated by scanning with a laser densitometer (Molecular Dynamics).

and translation

Plasmid DNAs were purified from cesium chloride gradients and then linearized with Hincll. Transcription reactions were performed with T7 RNA polymerase as specified by the manufacturer (Promega). Reaction mixtures containing RNA products were then extracted twice with phenol/chloroform and precipitated by the addition of ammonium acetate to 2.5 M and 2.5 vol of 95% ethanol. Translation reactions were carried out in rabbit reticulocyte lysates (Promega) as described by the supplier. Twenty milligrams/milliliter RNA was used in each 50ml reaction mixture containing either 0.8 mCi/ml [35S]methionine or 40 mM cold methionine. After 3 hr of incubation at 30”, translation reactions were terminated by addition of cycloheximide (5 mg/ml) and RNase A (10 mg/ml). Translation products were then analyzed on 12.5% SDS-polyacrylamide gels.

~220 cleavage activity

727

2Ap”

and immunoblot

analysis

~220 cleavage activity was assayed in vitro as previously described (Lloyd et al., 1985). The products from in vitro translation reactions were incubated at 37” for 2 hr with the post-mitochondrial cytoplasmic extracts from uninfected HeLa cells. Proteins were separated on a 7% SDS-polyacrylamide gel and subjected to immunoblot analysis using rabbit polyclonal antisera directed against the ~220 degradation products purified from poliovirus-infected cells (Lloyd et al., 1987).

In trans cleavage assay for the 1 D/2A junction A mutant clone, pEP2A(C57S), was first transcribed and translated in vitro in the presence of [35S]methionine to produce radiolabeled, uncleaved 1C’D2A precursor polypeptide (substrate). Parallel incubations were also performed in which RNA from

RESULTS Construction

of Cys and His substitution

mutants

Computer alignments of amino acid sequences from different picornavirus 2APro reveal that five of six cysteine and two of six histidine residues found in poliovirus 2APro are highly conserved among different strains of enteroviruses and rhinoviruses (Fig. 1). Among these conserved residues, His20 and CyslO9 have been predicted to comprise the catalytic active site (Bazan et al., 1988; Gorbalenya et a/., 1989) and we have previously reported that amino acid substitution of His20 or CyslO9 abolished the proteolytic activity of 2APro (Vu et al., 1991). However, little is known about the contribution of the other conserved cysteine and histidine residues to the structure and/or function of 2APro.Therefore, conservative mutations were made individually at each cysteine residue (Cys55, Cys57, Cys64, or Cysl15) as well as His1 17 to test whether these amino acids are essential for the proteolytic function of 2APro. Single amino acid substitutions in 2APro of poliovirus type 1, Mahoney strain, were generated via site-directed mutagenesis using synthetic oligonucleotide primers. The single-stranded DNA templates were prepared from the plasmid pEP2A (Vu et a/., 1991) which contains a T7 RNA polymerase promoter followed by most of the 5’ nontranslated region (ntr) and start codon of EMCV, and a partial 1C gene (93-bp), the entire 1 D and 2A genes, a stop codon in frame, and the beginning of the 2B gene (92-bp) of poliovirus (Fig. 2). The predicted translation product of this plasmid is a fusion protein with 6 amino acids at its N-terminus derived from the leader protein of EMCV and a linker, followed by 31 amino acids of poliovirus VP3 (lc’), the entire VP1 (1 D) and 2APro. Since the authentic 1D/2A junction is preserved in this construct, the active 2APro pro-

728

YU AND LLOYD

PVl CVBl CVB4 CVA9 CVA21 HRV2 HRV14 BEV EV70

PVl CVBl cvB4 CVA9 CVA21 HRV2 HRV14 BEV EV70

20 38 55 51 64 GFGHQNKAVYTAGYKICNYRLATQDDLQNAVNVMWSRDLLVTESRAQGTDSIAR~CNAGVYYCESRRKYYPVS GAFGQQSGAVYVGNYR~RHLATREDWQRCVWEDYNRDLLVSTTTAHGCDIIARCQ~TGVYFCASRNKHYPVS GPYGHQSGAVYVGNYK~RWLATHVDWQNCVWEDYNRDLVSTTTAHGCDTIARCQ~TGVYFCASKSKHYPVS GAFGQQSGAVYVGNYRVINRMATHTDWQNCVWEDYNRDLLVSTTTAHGCDVIARCQ~TGVYFCASKNKHYPVS GFGHQNKAVYVAGYKICNYmATPSDHLNAISVLWDRDLMVVESRAQGTDTIARCSCRCGVYYCESRRKYYLVT GPSDMYVHVGNLIVRNLIILF.NSEMHESILVSYSSDLIIYRTNTVGDDYIPSCDCTQATYYCKHKNRYFPIT GLGPRYGGIYTSNVKIMNY~MTPEDHHNLIAPYPNRDLAIVSTGGHG~TIPH~~SGVYY~TYYRKYYPII PFGQQQGAAYVGSYKILNRRATYADWENEVWQSYQRDLLVTR~AHGCDTIAR~~SGIYY~STAKHYPIV GPGYGGAFVGSYKIINYRLATDEEKERSVYVDWQSDNLVTTVAAHGKHQIARCRCNTHVYYCKHKNRSYPVC * * * 109 115 117 FVGPTFQYMEANNYYPVRYQSHMLIGHGFESPGDCGGILRCHHGVIVIITAGGEGLVAFSDIRDLYAYEEEAMEQ FEGPGLVEVQESEYYPKRYQSHVLLAAGFSEPGDCGGILRCEfiGWGIVTMGGEGWGFADVRDLLWLEDDAMEQ FEGPCLVEVQESEYYPKRYQSHVLLATGFSEPGDCGGILRCEHGVIGLVTMGGEGWGFADVRDLLWLEDDAMEQ FEGPGLVEVQESEYYPKRYQSHVLLAAGFSEPGDCGGILRCEHGVIVIVTMGGEGWGFADVRDLLWLEDDAMEQ FTGPTFRFMEANDYYPARTQSHMLIGCGFAEPGDCGGILR~HGVIGIITAGGEGIVAFADIRDLWVYEEEAMEQ VTSHDWYEIQISEYYPKHIQYNLLIGEGPCEPGDCt;GKLL~~GVIGIVTAGGDNHVAFIDLRHFHC.AEEQ CEKPTNIWIEGNPYYPSRFQAGNMKGVGPAEPGDCGGILRCIHGPIGLLTAGGSGYVCGADIRQLECIAEEQ VTPPSIYKIEANDYYPERMQTHILLGIGFAEPGD~GLLR~~~GILTVGGGDHVGFADV~LLWIEDD~EQ FEGPGIQWINESDYYPARYQTNTLLAMGPCQPGD*CGGLLV~S:SVIGLVTAGGEGIVAFTDIRNLLWLEDDAMEQ

FIG. 1. Alignment of 2APro amino acid sequences. Amino acid sequences of 2APro from representative enteroviruses and rhinoviruses were aligned by hand, which included poliovirus type 1 (PVl) (Kitamura era/., 1981); coxsackievirus type B 1 (CVBl) (lizuka eta/., 1987) type 84 (CVB4) (Jenkins eta/., 1987) type A9 (CVAS) (Chang eta/., 1989) and type A2 1 (CVA2 1) (Hughes ef al., 1989); human rhinovirus type 2 (HRV2) (Skern et a/., 1985) and type 14 (HRV14) (Stanway et al., 1984); bovine enterovirus (BEV) (Earle et a/., 1988); and enterovirus type 70 (Ryan et a/., 1990). Putative active site residues and cysteine and histidine residues mutagenized in this study are shown in bold face. Residues which could not be altered without abolishing enzymatic activity are marked with (*). Residues are numbered according to the amino acid sequence of type 1 poliovirus, Mahoney strain.

cesses this precursor polypeptide into a 38-kDa 1C’D polypeptide and a mature 17-kDa 2APro. Analysis

of autocatalytic

processing

by mutant 2APro

DNA from each mutagenized pEP2A plasmid was transcribed in vitro with T7 RNA polymerase and translated in cell-free rabbit reticulocyte lysates. The expression of 2APro linked to the adjacent upstream viral flanking sequences provides an immediate assay for the autocatalytic cis cleavage activity of 2AP” at the 1 D/2A junction. Results from the expression of wildtype and mutant 2AProare presented in Fig. 3. Translation of RNA derived from the wild-type pEP2A usually resulted in cleavage of greater than 95% of the precursor polypeptides into two major protein products, 1C’D polypeptide (38-kDa) and 2APro(17-kDa) (Fig. 3, lane a). Two single-site mutations at Cys64 (C64S and C64N) did not greatly influence the autocatalytic processing activity of 2AP” (lanes f and g). In contrast, none of the other mutant precursor polypeptides were processed into 1C’D and 2A proteins, indicating that autocatalytic function of these mutant 2APro had been abolished. Thus, we conclude that single amino acid substitutions at Cys55, Cys57, Cysl15, or His1 17 in 2AP”’ completely inhibit autocatalytic cis cleavage activity,

resulting in accumulation of the 55-kDa precursor polypeptides and no detectable production of 1C’D polypeptide or 2APro (Fig. 3, lanes b-e and h-m). A similar result has also been reported by Sommergruber et al. (1989) in which two of the 2APro mutants from human rhinovirus type 2, Cysl 12:Ser and His1 14:Gly, were negative for the autocatalytic cleavage activity when expressed in Escherchia co/i. Several other proteins were observed which we (Vu et a/., 1991) and others (Hellen et a/., 1991) have described previously. The five major protein bands migrating near 50, 44,41, 36, and 34 kDa were immunoprecipitable with anti-2A antisera (data not shown), indicating that these polypeptides were in frame translation products which contained major truncations at the NH, but not COOH terminus. Further, these extra polypeptides were not observed in translation products of RNAs in which the 1D gene had been deleted (Fig. 5A). Together, the data suggest that the background polypeptides detected in translation reactions were not likely derived from alternative cleavage products but rather products from aberrant initiation of translation. Interestingly, expression of wild-type 2APro or 2APro mutants which retain activity tended to suppress the appearance of these alternative polypeptides (compare lanes a, f, g to the other lanes) (Hellen et al.,

POLIOVIRUS

729

2Apro

pEP2A TAA

ATG EMCV 5’ ntr

1C’DZA

(55kDa)

t lCD(38kDa)

+

m

m

ZAP”

(17kDa)

pEPmet2A ATG EMCV 5’ ntr

TAA

I

2A

I 28

met2A pro (17 kDa) FIG. 2. Schematic diagram of the RNA transcribed from plasmids pEP2A and pEPmet2A and their predicted translation products. The EMCV 5 noncoding region is represented by a heavy bar and the protein coding regions are boxed. In the case of the wild-type 2APro, the predicted translation product of pEP2A RNA was a 55-kDa precursor polypeptide (1 C’D2A) which was rapidly processed into a 38-kDa (1 CD) and a 17-kDa (2Ap”) protein by 2AP” autocatalytic activity. The plasmid pEPmet2A was derived from pEP2A by deletion of the gene segment coding for the capsid proteins. Thus, the translation product of pEPmet2A RNA contained only a 17.kDa protein, with an extra methionine at the N-terminus of 2AP”.

1991; Yu et al., 1991). Experiments are currently underway to determine the mechanism responsible for this effect. When translation reactions were performed in the reticulocyte lysates supplemented with uninfected HeLa cell cytoplasmic extracts, most nonspecific translation products from internal initiation were re-

lC’D2A

+

1C’D

+

duced relative to the intended translation products, but were not eliminated (data not shown). However, this addition also reduced the overall efficiency of translation. Since these extra polypeptides did not interfere

abcdefghijklmno 2A

-3 a

bc

def

g

h

i

j

kl

FIG. 3. Expression and autocatalytic cis cleavage activity of mutant 2AP”. RNA derived from the wild-type (lane a) or mutagenized pEP2A (lanes b-l) was translated in rabbit reticulocyte lysates. Radiolabeled products were then separated by 12.5% SDS-polyacrylamide gel electrophoresis and analyzed by autoradiography.

FIG. 4. Induction of ~220 cleavage activity by mutant 2APro. Cytoplasmic extracts from uninfected HeLa cells (Sl 0) were incubated with the translation products of poliovirus RNA (lane c), wild-type (lane d), or mutagenized pEP2A RNAs (lanes e-o). Proteins in the mixtures were then separated by 7% SDS-polyacrylamide gels and analyzed by immunoblot assay with polyclonal antisera specific for ~220. US10 (lane a) and IS10 (lane b) represents the cytoplasmic extracts from mock-infected and poliovirus-infected HeLa cells, separately.

730

YU AND LLOYD

p220 --t P220 ZJ cl. pr.

abed

ef

g

abcde FIG. 5. Expression and analysis of mutant 2APro derived from PEPmet2A. (A) Expression of mutant 2AP” from pEPmet2A. Translation reactions were programmed with the RNA transcripts of wild-type (lane a) or mutagenized pEPmet2A and analyzed by 12.5% SDSPAGE and autoradiography. (B) Induction of ~220 cleavage activity by mutant 2AP” derived from pEPmet2A. HeLa SlO was incubated with the translation products displayed in panel A and analyzed by immunoblot assay with anti-p220 antisera.

(Fig. 4, lane d), replacement of Cys64 with Asn (Fig. 4, lane i) severely impaired the ability of 2AProto activate p220ase in trans. All other 2APro mutants tested also completely lost the ability to induce cleavage of ~220. When these mutants were assayed for direct in trans cleavage activity at the 1D/2A junction (as described under Materials and Methods), a similar result was observed, in which only one mutant, C64S, could efficiently process at the 1 D/2A junction in trans (data not shown). Table 1 summarizes the assay results for the in cis and in trans activities of 2APro measured over the course of several repeat experiments. Overall, we found that any point mutation at Cys55, Cys57, Cysl 15, and His1 17 completely blocked both in cis and in trans activities, while substitution of cysteine at residue 64 with serine only slightly altered activity. Also, the mutant C64N may be important because it can be used to distinguish the in cis and in vans activities of 2APro. Analysis

with observation of 1C’D2A, lC’D, or 2APro, and to maximize sensitivity and ability to detect trace cleavage products, translation reactions were performed in reticulocyte lysates alone. Induction

of ~220 cleavage activity

by mutant 2AP”

We also wanted to determine whether the translation products from wild-type and mutagenized pEP2A RNA were capable of activating p220-specific proteinase (p220ase) in trans and thus inducing cleavage of ~220. Therefore, translation products from the rabbit reticulocyte lysates were mixed with cytoplasmic extracts from uninfected HeLa cells and resultant ~220 cleavage was assayed by immunoblot analysis with p220specific antisera. Figure 4 shows immunoblot analysis of the translation products displayed in Fig. 3. The cytoplasmic extract from uninfected HeLa cells contained some ~220 degradation products which could be easily differentiated from the specific poliovirus-induced ~220 cleavage products due to slightly different compositions (compare Fig. 4, lanes a and b). Differences in bands may reflect the distinct modes of activation of cellular p220ase in uninfected and infected cells, although this will not be known until p220ase is purified and characterized. When wild-type 2AProwas produced, significant ~220 cleavage activity was induced and several specific cleavage products were detected in immunoblot assaywhich co-migrated with cleavage products from infected HeLa cells (Fig. 4, lanes b, c, and d). Interestingly, while substitution of Cys64 with Ser (Fig. 4, lane j) resulted in only a slightly diminished degree of activity from the wild-type 2AP”

of mutant 2APro derived from pEPmet2A

Since p220ase has not been purified and characterized, its mechanism of activation is unknown. To test whether p220ase activation occurs by a catalytic cleavage mechanism or via a conformational change in p220ase induced by a specific binding interaction between 2AProand p220ase, we constructed a new plasmid, pEPmet2A. The plasmid pEPmet2A (Fig. 2) was derived from the parental plasmid pEP2A by deleting the 1C’D region, thereby placing the 2A gene immedi-

TABLE 1 cis AND Vans ACTIVITIESOF MUTANT 2APro 2APro Assay 2A Mutant

Polyprotein cleavage (cis)

p220ase activation (Vans)

wild-type

++++

-

++++

c55s c57s C57N C57T C64N C64S Cl 15P C115Y C115H c115s H117Q

-

-

+++ ++++ -

-

+i++++ -

Note. 2APm activity was graded in comparison to wild-type controls. Activity was determined to be undetectable (-). detectable in some assays (+I-), or positive (+ through ++++).

POLIOVIRUS

ately after the AUG start codon. This manipulation would also separate any potential requirement for autocatalytic activity of 2AProfrom induction of ~220 cleavage activity. For those 2A mutants that lost all measurable in cis and in Vans activities, it was also of interest to know whether the loss of trans activity resulted from inhibition of catalytic activity or merely stearic hindrance or conformational changes imposed by the 1 D polypeptide still attached to the 2AProin these mutants. Therefore, several point mutations in 2AP” sequences (C55S C57S Cl 09S, and Cl 15s) were independently subcloned into pEPmet2A (as detailed under Materials and Methods). The translation products of RNA transcripts from the wild-type or mutant pEPmet2A generated mature 17-kDa 2AProcontaining only an extra methionine residue at the NH, terminus (Fig. 5A). When the wild-type 2APro generated from pEPmet2A RNA was assayed for the ability to induce ~220 cleavage in vans, p220ase activity was easily induced (Fig. 5B, lane c), indicating that the presence of the additional methionine at the native NH, terminus of 2APro did not hinder in vans activity of 2APro. In contrast, none of the four CysASer mutants tested could activate ~220 cleavage (Fig. 5B, lanes d, e, f, and g) even though they could presumably bind p220ase. Furthermore, p220ase was not activated by the enzyme derived from the active site mutant, pEPmet2A(C109S), which should not contain large perturbations in the tertiary structure of 2APro. This result thus implies a catalytic cleavage mechanism for p220ase activation. Our data also suggest that these residues, including Cys55, Cys57, Cysl 15, and His1 17, if not directly involved in the catalytic charge-relay system, are critical in maintaining the active conformation of 2APro structure.

Assay for disulfide bridges in 2APro The proposal for the tertiary structure of 2APro was derived by superimposing 2APro sequence onto the solved p-carbon framework of Strepromyces griseus protease A (SGPA) (Bazan eta/., 1988). Close examination of this proposed structure revealed that two pairs of the residues under study, Cys55Kys57 and Cysl 15/Hisll7, are predicted to reside on two adjacent loops situated near the surface of the molecule. Thus, these residues may be spatially arranged into a very close proximity in the folded conformation of 2APro, despite their distance in the primary amino acid sequence. To explore the specific structural roles played by these conserved cysteine and histidine residues, three possibilities were proposed: (i) 2APro contained a critical disulfide bridge between either Cys55 and

2AP”

731

Cysl15 or Cys57 and Cysl15, (ii) 2APro contained a metal-binding domain stabilized by the tetrahedral coordination of a divalent cation into the motif Cys,,-XaaCys,,-Xaa,-Cys, ,,-Xaa-His,,, , or (iii) a key alternative structural and/or functional role played by these conserved residues which is yet unknown. To examine whether there were intra-molecular disulfide bridges present in 2APro, we first analyzed the electrophoretic mobility of 2AProexpressed in the reticulocyte lysates on denaturing SDS-polyacrylamide gels with or without dithiothreitol (DlT) reduction. Theoretically, if any intra-molecular disulfides are present, the non-reduced form of 2AProwill have an altered, probably greater mobility because restriction of flexibility of the unfolded polypeptide chain decreases the hydrodynamic volume of the molecule (Crelighton eta/., 1990). When RNAase Tl , which contains two intramolecular disulfide bonds, was analyzed on SDS-polyacrylamide gels, its nonreduced form migrated considerately faster than its reduced form (data not shown). But no apparent difference in the electrophoretic mobility of 2APro was detected on the 15% SDS-polyacrylamide gel with or without DlT reduction (Fig. 6A), suggesting an absence of disulfide bridges in 2APro. Nevertheless, we could not exclude the possibility that the mobility change induced by reduction of a disulfide bridge in 2AP” was so subtle that it could not be detected by this type of analysis. Thus, we further probed for the existence of an intramolecular disulfide linkage in 2AProby employing chemical modification of all free thiol groups. Reticulocyte lysates containing radiolabeled 2AProwere treated with excess N-ethylmaleimide (NEM) (MW 125) or iodoacetamide (IAA) (MW 185) to irreversibly bind and sequester any free thiols. This treatment would not break any disulfide bridges present in 2APro. Subsequently, excess blocking reagents were completely removed by exhaustive dialysis and 2APro was reduced by DTT treatment, which would presumably expose a new pair of unmodified free cysteines from each potential disulfide bridge. The sample was dialyzed again to remove the excess DlT and then incubated with iodoacetyl-LC-biotin (n-iodoacetyl-N-biotinylhexylenediamine, MW 510) a sulfhydr-yl-reactive biotinylating reagent. Binding of two iodoacetyl-LC-biotin molecules should impose a retarded electrophoretic mobility on the modified 2APro because of additional size and charge. Lysozyme was used as a positive control for this experiment since it contains eight cysteine residues and forms four intra-molecular disulfide bridges (Creighton era/., 1980). Figure 6B shows that iodoacetyl-LC-biotin modifies lysozyme and causes a large retardation in

732

YU AND LLOYD

C -++-

a

b

-

5ChnMNEM

---

D’lT reduction

-

-

+-

+

-

a

bc

d

e

++ -

abc

+

-

t

de

FIG. 6. Assay for disulfide bridges in 2APro. (A) Electrophoretic mobility of 2AP”. 2AP” was translated in reticulocy-te lysates programmed with the wild-type pEPmet2A RNA and partially purified by HPLC anion exchange chromatography on MA7Q column (BioRad). Proteins were then analyzed by electrophoresis through a 15% SDS-polyacrylamide gel with DlT (lane a) or without DlT (lane b) reduction. Migration of molecular mass markers is indicated in kilodaltons. (B) Chemical modification of lysozyme. Lysozyme (Sigma) was treated as indicated, separated on a 15% SDS-polyacrylamide gel, and visualized by Coomassie blue staining. (C) Chemical modification of 2AP”‘. 2APro was expressed by in vitro translation of wild-type pEPmet2A RNA and treated as indicated. Radiolabeled proteins were separated on a 15% SDS-polyacrylamide gel. Arrows indicate the positions of 2APr0 migration and numbers indicate molecular mass markers (in kilodaltons).

electrophoretic mobility only after the enzyme was reduced with DTT (lanes b and d). Further, pretreatment with NEM did not change the mobility of the non-reduced form of lysozyme (lanes c and e), demonstrating that NEM itself cannot reduce disulfide bridges. These results are entirely consistent with the known structure of lysozyme and confirm that all cysteine residues in lysozyme are sequestered in disulfide bridges. However, when 2APro produced in reticulocyte lysates was analyzed in the same experiment, a different result was obtained (Fig. 6C). If 2APro was first pretreated with NEM (Fig. 6C, lanes d and e) or IAA (data not shown), no difference in binding of iodoacetyl-LC-biotin to 2APro was observed in either DTT-reduced or nonreduced forms, implying again that no intra-molecular disulfide bridges existed in 2APro. The slower mobility of iodoacetyl-LC-biotin-modified 2APro than unmodified 2APro (compare Fig. 6C, lanes b and c with a) suggested that most or all of the cysteine residues in the unmodified form of 2AProwere available for binding with iodoacetylLC-biotin. Additionally, NEM pretreatment blocked subsequent modification with iodoacetyl-LC-biotin (Fig. 6C, lane d), demonstrating that all cysteines which could be modified with iodoacetyl-LC-biotin could also be modified with NEM. Thus, we interpret the results to suggest that the 2APro in Fig. 6C, lanes b and c were modified with 3-5 molecules of iodoacetylLC-biotin, causing a large shift in migration; the 2AProin Fig. 6C, lanes d and e were modified only with 3-5 molecules of NEM, causing a slight shift in migration when compared to control unmodified 2APro (Fig. 6C,

lane a). The data, taken together, suggest that 2APro does not contain an intra-molecular disulfide bridge. Assay for metal-binding

ligands in 2APro

Having found no evidence for disulfide bridges in 2APro, we then tested the hypothesis that these conserved cysteine and histidine residues may coordinate an essential divalent cation. We first monitored the effect of different metal-chelating agents on the proteolytic activity of 2APro using the direct in rrans cleavage assay described under Materials and Methods. Unlabeled wild-type 2AProsynthesized in the rabbit reticulocyte lysates was preincubated with increasing amounts of different metal chelators before addition of [35S]methionine-labeled polypeptide substrates. Among those chelators, I, 10-phenanthroline (I, 1OPTH) has a relatively high specificity for Zn*+ ions while both EDTA and EGTA are commonly used as less specific divalent ion chelators. Our results showed that neither EDTA nor EGTA at concentrations up to 10 m/l/l had apparent effect on 2APro activity (data not shown). However, in several repeat experiments, I, 10-PTH si’gnificantly inhibited the direct in trans cleavage activity of 2APro(Fig. 7). An isomer of I, 10-PTH which does not chelate divalent cations, 1,7-PTH, was included in these experiments as a negative control and had little effect on 2AP” activity. Additionally, DlT (l-l 0 m/V), which has been reported to have a large affinity constant for Zn*+ (Cornell et al., 1972; Miller et al., 1985) did not adversely affect 2APro activity when it was included in the reactions (data not shown).

POLIOVIRUS

13 1,7 phenanthroline H 1,lO phenanthroline

0.5

1 inhibitor

2 concentration

4 (mM)

7

FIG. 7. Inhibition of 2APro activity by 1 ,lO phenanthroline. In tram cleavage assay at the 1 D/2A junction (described under Materials and Methods) was used to test the inhibitory effect of 1,l O-phenanthroline on 2APro. 1.7.phenanthroline was included as a nonchelating control. Unlabeled 2APro (enzyme) derived from wild-type PEPmet2A was preincubated with different amounts of 1 ,lO- or 1,7phenanthroline at 30” for 1 hr before the addition of radiolabeled. uncleaved 1 C’D2A(C57S) polypeptide (substrate) for assaying the in trans cleavage activity of 2AP’“. Proteins were analyzed by SDSPAGE and autoradiography and were quantified by laser densitometry. The result is presented as the percent ratio of the processed product (1 CD) to the precursor protein (1 C’D2A).

DISCUSSION We have previously demonstrated (Vu et al,, 1991) that single amino acid substitutions at the conserved His20 and CyslO9 residues completely abolished the proteolytic activity of 2APro, and thus supported the hypothesized catalytic roles of His20 and CyslO9 in the active site predicted by two independent alignment studies (Bazan et al., 1988; Gorbalenya et al., 1989). However, these alignments excluded the possibility that the other conserved cysteine and histidine residues found in 2A pro, namely, Cys55, Cys57, Cys64, Cysl15, and His1 17, could be part of the catalytic active site. This was probably due to the fact that although these residues are highly conserved among picornavirus 2APro,they are not found in either viral 3Cpro or cellular serine proteases. Nevertheless, the critical nature of these cysteine and histidine residues in normal 2APro structure and/or function, first suggested by their extreme conservation in many different strains of enteroviruses and rhinoviruses, with the exception of C64, are now confirmed by our data. Conservative amino acid replacements at these invariant residues (summarized in Table 1) resulted in each case, in a complete loss of both in cis and in vans activities of 2APro. Thus, each substitution drastically altered the biochemical properties of 2APro, emphasizing that

2AP”

733

these conserved residues are essential for normal enzymatic function. Interestingly, the mutant C64N had differential activities in in cis and in vans assays and is the fourth 2APro mutant that we have generated with this phenotype. However, unlike the other three mutants (Vu eta/., 1991) Cys64 is not proposed to lie near the substrate-binding domain of the 2APromolecule according to the hypothesized structural model (Bazan et a/., 1988). Although reasons for its differential activities are unclear, C64N may be a valuable mutant for assessing the roles of cis and Pans activities of 2APro in the replicative cycle of poliovirus in vivo. Since most of the cysteine and histidine mutants derived from pEP2A lost both in cis and in trans activities of 2APro, we constructed the plasmid pEPmet2A in order to eliminate the need for autocatalytic cis function to test the in vans activity of these mutant 2APro. Interestingly, 2APro expressed from the wild-type PEPmet2A was fully active, capable of cleaving the 1 D/2A junction in trans and activating p220-specific proteinase in trans. A similar result has also been reported by Krausslich et a/. (1987) who have shown that a fusion protein containing 5 amino acids of the L protein of EMCV at its amino terminus, followed by 1 amino acid of Pl , the entire 2AProsequence, and 24 amino acids of 2B of poliovirus type 3 (Sabin) could efficiently induced cleavage of ~220. Taken together, these data demonstrate that 2AProfolds into an active conformation in the total absence of upstream 1 D amino acids. Therefore, 1 D or other capsid polypeptides are not required to serve as a chaperone in order to aid 2APro folding as has been described for other related proteases, such as &tic protease of Lysobacter enzymogenes (Silen et al., 1989). However, the mutant 2APro (C55S, C57S, and Cl 15s) expressed from pEPmet2A could not induce ~220 cleavage activity in Vans even though they could no longer be sterically inhibited by covalently attached 1 D protein (VPl). These data are most consistent with a catalytic mode of activation of p220ase, rather than alteration of p220ase activity through a noncovalent interaction. However, it is possible that these singlesite mutations may have caused gross structural modifications in 2APro molecule which result in sufficient global conformational changes to block both in cis and in Vans activities completely. In general, single-aminoacid substitutions which significantly perturb protein structure are usually found in residues that are not solvent exposed and are immobile in the native structure. In contrast, substitutions of single amino acids on the surface of proteins generally do not drastically perturb tertiary structure. To gain some insight into the structure of 2APro,we have carefully mapped the amino acid

734

YU AND LLOYD

sequences of 2APro in the tertiary structure model proposed by Bazan and Fletterick (1988). This model predicts that while both His20 and CyslO9 lie near the surface of 2AP” structure in the proposed catalytic domain, the other residues (Cys55, Cys57, Cysll5, and His1 17) probably lie close together on the opposite side of the molecule from active site domain, near the surface. Whether these residues are truly solvent exposed (displayed on the surface of 2APro)or not solvent exposed (buried within the core of 2APro) still remains unclear, although several of these cysteines were apparently modified with iodoacetyl-LC-biotin, suggesting surface position. Since these four residues are positioned very closely in the proposed 2APro tertiary structure, several hypotheses were proposed to explain the essential roles of these residues: (i) an intramolecular disulfide bridge between Cys55 and Cysl 15 or Cys57 and Cysl 15, (ii) a metal-binding ligand formed by the motif Cys,,-Xaa-Cys,,-Xaa,-Cys, ,,-XaaHis,,,, or (iii) other alternative key, yet unknown, functions. Analysis of 2APro mobility through reduced and nonreduced gels, together with experimental results with chemical modifications caused us to conclude that 2AProdoes not contain disulfide bridges. Failure to detect a disulfide bridge was not unexpected given the solely cytoplasmic occurrence of 2AP”. In addition, the stoichiomett-y of three cysteine residues in this region of the molecule would still leave an unpaired cysteine even after a bridge was formed. Therefore, the suggestion that Cys55, Cys57, Cysl15, and His1 17 provided a ligand for binding a divalent cation (e.g., Zn2+) became a working hypothesis which might explain the extreme effect of conservative amino acid substitutions on 2APro activity. However, although the spacing of the cysteine and histidine residues (Cys,,-XaaCys,,-Xaa,-Cys,,,-Xaa-His,,,) is conserved among all enterovirus and rhinovirus 2APro, it differs from those in the zinc finger-like DNA- or RNA-binding proteins which typically have two or more amino acid residues separating Cys and His residues on each side of the finger loops (e.g., Cys-Xaa-Xaa-Cys-Xaa,-Cys-XaaXaa-His) (Berg, 1986). Our results with different metal chelators presented here cannot exclude the existence of a metal-binding domain in 2APro expressed in the reticulocyte system. Even though EGTA and EDTA did not inhibit 2APro activity, it is questionable whether these charged chelators could complex cations buried internally or in other hydrophobic domains of the molecule. More definitive experiments will require development of the ability to completely denature, then refold active 2APro. Further studies are undergoing to express and purify a high quantity of active 2APro in order to continue this line of investigation.

It is also possible that the conserved cysteine and histidine residues perform an alternative function which is not known. Recently, several examples of novel functions for cysteine or histidine residues have been described, including inter-molecular protein-protein linkages (such as metal-linked dimers in human growth hormone or Tat protein from human immunodeficiency virus) (Cunningham et a/., 1991; Frankel et a/., 1988), protein-nucleic acid interaction (such as an RNA-binding nucleocapsid protein in Moloney murine leukemia virus) (Gorelick eta/., 1988), and redox regulation of DNA-protein binding (such as DNA binding activity of Fos and Jun proteins) (Abate et a/., 1990). In another example, the roles of the conserved cysteine and histidine residues in the nucleocapsid protein of avian myeloblastosis virus (Jentoft et al., 1988) remained undefined, but were shown not to involve a metal-binding ligand or a disulfide bridge. Finally, although ourexperimental results do not permit us to define the precise roles of these cysteine and histidine residues in supporting the function of 2APro, we have eliminated intermolecular or intra-molecular disulfide linkage as a structural feature of 2APro. Also, our study with metal chelators leaves open the possibility that a metal-binding ligand within 2APro supports its proteolytic activity. Alternatively, these conserved residues may have a novel functional or structural role that is required for the specific enzymatic activity of 2APro.

ACKNOWLEDGMENTS We thank Drs. J. S. Hanas and R. K. Tweten for many helpful suggestions on the biochemical analyses of 2APr0. This work was supported by Public Health Service Grant Al2791 4.

REFERENCES ABATE, C., PATEL, L., RAUSCHER, F. J., III, and CURRAN, T. (1990). Redox regulation of Fos and Jun DNA-binding activity in vitro. Science249, 1157-1161. ARGOS, P., KAMER, G., NICKLIN, M. J. H., and WIMMER, E. (1984). Similarity in gene organization and homology between proteins of animal picornaviruses and a plant comovirus suggest common ancestry of these virus families. Nucleic Acids Res. 12, 725 l-7276. BAZAN. 1. F., and FLEI-~ERICK,R. J. (1988). Viral cysteine proteases: structural and functional implications. Proc. Nat/. Acad. SC;. USA 85,7872-7876. BERG, 1. M. (1986). Potential metal-binding domains in nucleic acid binding proteins. Science 232, 485-487. BERNSTEIN,H. D., SONENBERG.N., and BALTIMORE, D. (1985). Poliovirus mutant that does not selectively inhibit host protein synthesis. Mol. Cell. Biol. 5, 29 13-2923. BLINOV, V. M., DONCHENKO, A. P., and GORBALENYA,A. E. (1985). Internal homology in the primary structure of the poliovirus polyprotein: The possibility of existence of two viral proteinases. Lectures Acad. Sci. USSR 281, 984-987. CHANG, K., AUVINEN. P., HYPPIA, T., and STANWAY, G. (1989). The nucleotide sequence of coxsackievirus A9; implication of receptor

POLIOVIRUS binding and enterovirus classification. /. Gen. Viral. 70, 32693280. CORNELL, N. W., and CRIVARO, K. E. (1972). Stability constant for the zinc-dithiothreitol complex. Anal. Biochem. 47, 203-208. CREIGHTON,T. E. (1980). Counting integral numbers of amino acid residues per polypeptide chain. Nature 284, 487-489. CREIGHTON, T. E. (1990). Disulphide bonds between cysteine residues. In “Protein Structure: A Practical Approach” (T. E. Creighton, Ed.), pp. 155-l 64. IRL Press, Oxford, England. CUNNINGHAM, 6. C., MULKERRIN, M. G., and WELLS, J. A. (1991). Dimerization of human growth hormone by zinc. Science 253, 545548. EARLE,J. A., SKUCE, R. A., HOEY, E. M., and MARTIN, S. J. (1988). The complete nucleotide sequence of a bovine enterovirus. /. Gen. Virol. 69, 253-263. EMINI, E. A., IZINGA. M. E., and WIMMER, E. (1982). Carboxyl-terminal analysis of poliovirus proteins: termination of poliovirus RNA translation and location of unique poliovirus cleavage sites. /. Viral. 42, 194-l 99. FRANKEL,A. D., BREDT, D. S., and PABO, C. 0. (1988). Tat protein from human immunodeficiency virus forms a metal-linked dimer. Science 241, 70-73. GORBALENYA,A. E., DONCHENKO, A. P., BLINOV, V. M.. and KOONIN. E. V. (1989). Cysteine proteases of positive strand RNA viruses and chymotrypsin-like serine proteases. FEBS Lett. 243, 103114. GORELICK,R. J., HENDERSON,L. E., HANSER,J. P., and REIN, A. (1988). Point mutants of Moloney murine leukemia virus that fail to package viral RNA: Evidence for specific RNA recognition by a “zinc finger-like” protein sequence. Proc. Narl. Acad. Sci. USA 85, 8420-8424. HANECAK, R., SEMLER. B. L., ANDERSON, C. W., and WIMMER, E. (1982). Proteolytic processing of poliovirus polypeptides: antibodies to polypeptide P3-37 inhibit cleavage at glutamine-glycine pairs. Proc. Nat/. Acad. Sci. USA 79, 3973-3977. HELLEN, C. U. T., F%KE, M., K~~USSLICH, H. G., LEE, C-K., and WIMMER, E. (1991). Characterization of poliovirus 2A protease by mutational analysis: Residues required for autocatalytic activity are essential for induction of cleavage of eukaryotic initiation factor 4F polypeptide ~220.1. viral. 65, 4226-4231. HUGHES, P. J.. NORTH. C., MINOR, P. D., and STANWAY,G. (1989). The complete nucleotide sequence of coxsackievirus A21. /. Gen. Viral. 70, 2943-2952. IIZUKA, N., KUGE, S., and NOMOTO, A. (1987). Complete nucleotide sequence of the genome of coxsackievirus Bl Virology 156, 6473. JENKINS,0.. BOOTH. J. D., MINOR, P. D., and ALMOND, J. W. (1987). The complete nucleotide sequence of coxsackievirus 84 and its comparison to other members of the picornaviridae. J. Gen. Viral. 68,1835-1848. JENTOFT.J. E., SMITH, L. M., Fu, X., JOHNSON, M., and LEIS, J. (1988). Conserved cysteine and histidine residues of the avian myeloblastosis virus nucleocapsid protein are essential for virus replication but are not “zinc-binding fingers.” Proc. Nat/. Acad. Sci. USA 85, 7094-7098. JORE,J., DEGEUS, B., JACKSON,R. I., POUWELS,P. H., and ENGER-VALK, B. (1988). Poliovirus protein 3CD is the active proteasefor processing of the precursor protein Pl in vitro. 1. Gen. Viral. 69, 16271636. KITAMURA, N., SEMLER, B. L., ROTHBERG,P. G., LARSEN, G. R., ADLER, C. J., DORNER, A. J., EMINI, E. A., HANECAK, R., LEE, J. J., VAN DER WERF, S., ANDERSON,C. W., and WIMMER, E. (1981). Primary struc-

ZAP”

735

ture, gene organization, and polypeptide expression of poliovirus RNA. Nature 291, 547-553. KRAUSSLICH,H. G.. NICKLIN, M. J. H., TOYODA. H., ETCHISON. D., and WIMMER, E. (1987). Poliovirus proteinase 2A induces cleavage of eucaryotic initiation factor 4F polypeptide ~220. /. Viral. 61, 271 l2718. LARSEN, G. R., ANDERSON, C. W., DORNER, A. J., and SEMLER, B. L. (1982). Cleavage sites within the poliovirus capsid protein precursors. 1. Wrol. 41, 340-344. LLOYD, R. E., ETCHISON. D., and EHRENFELD,E. (1985). Poliovirus protease does not mediate cleavage of the 220,000-Da component of the cap binding protein complex. Proc. Nat/. Acad. SC;. USA 82, 2723-2727. LLOYD, R. E., JENSE, H. G., and EHRENFELD,E. (1987). Restriction of translation of capped mRNA in vitro as a model for poliovirus-induced inhibition of host cell protein synthesis: Relationship to ~220 cleavage. J. Viral. 61, 2480-2488. LLOYD, R. E., TOYODA. H., ETCHISON. D., WIMMER. E., and EHRENFELD, E. (1986). Cleavage of the cap binding protein complex polypeptide ~220 is not effected by the second poliovirus protease 2A. Virology 150, 299-303. LLOYD, R. E., GRUBMAN, M. J., and EHRENFELD.E. (1988). Relationship of ~220 cleavage during picornavirus infection to 2A proteinase sequences. J. Viral. 62, 4216-4223. MILLER, J. A., MCLACHLAN, D., and KLUG, A. (1985). Repetitive zincbinding domains in the protein transcription factor IIIA from Xenopus oocytes. EMBO /. 4, 1609-l 614. NICKLIN, M. J. H., KF&USSLICH, H. G., TOYODA. H., DUNN, J. J., and WIMMER. E. (1987). Poliovirus polypeptide precursors: expression in vitro and processing by exogenous 3C and 2A proteinases. Proc. Nat/. Acad. Sci. USA 84, 4002-4006. RYAN, M. D., JENKINS,O., HUGHES, P. J., BROWN, A., KNOWLES, N. J., BOOTH, D., MINOR, P. D., and ALMOND, J. W. (1990). The complete nucleotide sequence of enterovirus type 70: Relationships with other members of the picornaviridae. J. Gen. Viral. 71, 22912299. SILEN, J. L., and AGARD, D. A. (1989). The cu-lytic protease pro-region does not require a physical linkage to activate the protease domain in vivo. Nature 341, 463-464. SKERN, T., SOMMERGRUBER,W., BLAAS, D.. GRUENDLER, P., FRAUENDORFER,F.. PIELER,C., FOGY, I., and KUECHLER,E. (1985). Human rhinovirus 2: Complete nucleotide sequence and proteolytic processing signals in the capsid protein region. Nucleic Acid Res. 13, 2111-2126. SOMMERGRUBER,W., ZORN. M., BLAAS, D., FESSL, F., VOLKMANN, P., MAURER-FOGY. I., PALLAI, P., MERLUZZI, V., MATTEO, M., SKERN. T., and KUECHLER,E. (1989). Polypeptide 2A of human rhinovirus type 2: Identification as a protease and characterization by mutational analysis. Virology 169, 68-77. STANWAY, G., HUGHES, P. J.. MOUNTFORD, R. C.. MINOR, P. D., and ALMOND, J. W. (1984). The complete nucleotide sequence of a common cold virus: Human rhinovirus 14. Nucleic Acid Res. 12, 7859-7875. TOYODA, H., NICKLIN. M. J. H., MURRAY. M. G., ANDERSON, C. W., DUNN, J. J., STUDIER,F. W., and WIMMER, E. (1986). A second virusencoded proteinase involved in proteolytic processing of poliovirus polyprotein. Cell 45, 76 l-770. YPMA-WONG, M. F., DEWALT, P. G.. JOHNSON,V. H., LAMB, J. G., and SEMLER, B. L. (1988). Protein 3CD is the major poliovirus proteinase responsible for cleavage of the Pl capsid precursor. Virology 166,265-270. Yu, S. F., and LLOYD, R. E. (1991). Identification of essential amino acid residues in the functional activity of poliovirus 2A protease. Virology 182, 615-625.