Expression and Purification of Recombinant Rhinovirus 14 3CD Proteinase and Its Comparison to the 3C Proteinase

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 346, No. 1, October 1, pp. 125–130, 1997 Article No. BB970291

Expression and Purification of Recombinant Rhinovirus 14 3CD Proteinase and Its Comparison to the 3C Proteinase Gerard J. Davis, Q. May Wang, Gregory A. Cox, Robert B. Johnson, Mark Wakulchik, Crystal A. Dotson, and Elcira C. Villarreal1 Infectious Diseases Research, Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, Indiana 46285

Received June 6, 1997

Human rhinovirus (HRV) is a positive-stranded RNA virus with an open reading frame that encodes for a single polyprotein of about 3000 amino acids. The HRV polyprotein is proteolytically processed; eight of nine cleavages are catalyzed by the 3C and/or the 3CD proteinases. We have expressed and purified recombinant HRV14 3C and 3CD proteinases and investigated their substrate selectivity and inhibitor sensitivity. Expressed 3CD proteinase had the P1/P1* residues of the 3C/3D cleavage site mutated from Gln/Gly to Ala/Ala in order to prevent autocleavage. The 3CD proteinase activities were measured by utilization of native, chromogenic, and fluorogenic peptide substrates. The 3CD proteinase exhibited £15% activity, compared to 3C, toward peptidyl p-nitroanilide substrates which contain only the p-nitroaniline moiety in the prime side. The 3C and 3CD proteinases exhibited similar activities for both internally quenched fluorogenic and native peptides. These results suggest that the two enzymes have similar but slightly different substrate specificity, especially on their preference for prime side residues. Inhibitor sensitivities toward classical proteinase inhibitors were generally similar for both enzymes. Small peptidyl inhibitors, specifically designed and synthesized for HRV14 3C, also inhibited the 3CD proteinase. Taken together, our data indicate that the 3D domain of 3CD proteinase had some influence on substrate recognition, but did not have dramatic impact on its interaction with inhibitors. q 1997 Academic Press

Key Words: rhinovirus; viral 3C and 3CD protease; protease substrate specificity.

Human rhinovirus (HRV)2 is the etiologic pathogen for greater than 50% of the common colds (1). HRV is in the picornavirus family, which also includes poliovirus, foot-and-mouth disease virus, cardiovirus, and hepatitis A virus (2, 3). Picornavirus genomes consist of small positive strand RNAs of about 10 kb in size. The RNAs encode for single polyproteins of Ç250 kDa, which are proteolytically processed to mature functional proteins. Within the picornaviridae family, general strategies for viral replication and polyprotein processing appear to be very similar (2, 3). Much of our knowledge of picornaviruses comes from investigations of poliovirus, since it has been more extensively investigated than HRV and most other picornaviruses (2, 3). During and after polyprotein production, the polyprotein is cleaved into at least 11 structural and nonstructural proteins which form the viral capsid, function in viral replication, and participate in host cell shutoff (2, 3). Structural capsid proteins are derived from about 40% of the N-terminal region of the polyprotein, while nonstructural proteins involved in replication and host cell shutoff are derived from the C-terminal region. Included in the polyprotein cleavage products are proteinase 3C and its precursor, 3CD. In the case of polioviruses, the 3CD has proteinase (3C) and polymerase (3D) domains. However, only the proteinase domain is catalytically active in the complex. Proteinases 3C and/or 3CD are responsible for 8 of 9 polyprotein proteolytic cleavages. Previous studies on poliovirus 3CD proteinase have suggested that this enzyme is mainly responsible for structural region cleavages of 2

1 To whom correspondence should be addressed at Infectious Diseases Research, Building 98A/3C, Drop Code 0438, Indianapolis, IN 46285. Fax: 317-276-1743. E-mail: [email protected].

Abbreviations used: DMSO, dimethyl sulfoxide; Dabcyl, 4-(4-dimethylaminophenazo) benzoic acid; Edans, 5-[(2-aminoethyl)amino]naphthalene-1-sulfonic acid; E-64, L-trans-epoxysuccinylleucylamide-4-(guanidino)-butane; HRV, human rhinovirus; IPTG, isopropyl b-D-thiogalactopyranoside; pNA, p-nitroaniline or p-nitroanilide; PMSF, phenylmethylsulfonyl fluoride; TFA, trifluoroacetic acid; TLCK, tosyllysine chloromethyl ketone.

125

0003-9861/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.

AID

ABB 0291

/

6b3f$$$$81

09-02-97 11:29:59

arcas

126

DAVIS ET AL.

the polyprotein (4, 5). In addition to their proteolytic functions, these proteins are constituents of viral replication complexes (6, 7). There have been at least three reports describing the ability of HRV serotype 14 (HRV14) 3C proteinase to proteolytically hydrolyze peptide substrates that represent the eight 3C and/or 3CD proteinase polyprotein cleavage sites (8–10). The 3C proteinase cleaves peptides that represent the nonstructural region cleavage sites; however, there was no detectable cleavage of peptides that represented the structural region cleavage sites (8–10). These types of peptide substrate specificity investigations have not been reported for any of the picornavirus 3CD proteinases; neither has there been a report of a systematic comparison of substrate preferences for 3C and 3CD. In this study, we investigated the influence of the polymerase domain on cleavage specificity and inhibitor sensitivity for the HRV14 3CD proteinase domain. This has been accomplished by comparing relative catalytic efficiencies for peptide hydrolysis and effects of protease inhibitors for both 3C and 3CD proteinases under identical assay conditions. In contrast to HRV14 3C, the 3CD protein has not been expressed, purified, and characterized. To determine the relative catalytic efficiencies of HRV14 3C and 3CD to cleave small peptide substrates, here we report the successful expression of HRV14 3CD as soluble active proteinase in Escherichia coli. As was the problem with poliovirus 3CD (11, 12), efforts to express the 3CD were hampered by its autocleavage to 3C and 3D; this problem was addressed by making site-directed mutations at the 3C/3D cleavage site. In addition, these two proteinases were kinetically compared to determine if the 3D domain significantly affects the proteolytic activity of the 3CD. This information is valuable in assessing whether these enzymes should be pursued as potential antirhinoviral targets for drug development. MATERIALS AND METHODS Construction of HRV14 3CD proteinase expression vectors. The coding sequence for HRV14 3CD was derived from a viral cDNA clone, pWR40 (gift from W.-M. Lee and R. Rueckert, University of Wisconsin). The 3CD coding region was amplified by the polymerase chain reaction utilizing the following HRV14 3CD-specific ‘‘outside’’ oligonucleotide primers: forward (5*-CCCCCCCCCATGGGACCAAACACAGAATTTGCA-3*) and reverse (5*-GGGGGCGGCCGCCTATTAAAAGAGGTCCAACCAGCG-3*) primers. Site-directed mutations were generated at the 3C/3D cleavage site by PCR mutagenesis (13). The mutagenesis oligonucleotides were 5*-TATTTTGTAGAGAAAGCAGCCCAAGTAATAGCTAGA-3* (forward) and 5*-TCTAGCTATTACTTGGGCTGCTTTCTCTACAAAATA-3* (reverse). For soluble expression of HRV 3CD in E. coli, the HRV14 3CD cDNA was cloned into the E. coli expression vector pET-30a (Novagen). DNA sequencing and restriction enzyme mapping were utilized to verify accurate construction of the pET-30/3CD expression constructs. Expression and purification of HRV14 3C and 3CD proteinases. HRV14 3C was expressed and purified as previously described (13).

AID

ABB 0291

/

6b3f$$$$82

09-02-97 11:29:59

The 3C enzyme sample was further dialyzed overnight against a buffer containing 25 mM Hepes, pH 7.5, 150 mM NaCl, and 25% glycerol. Significant efforts were made to express soluble 3CD in sufficient quantity, and in a form that was active but could no longer autocleave to 3C and 3D. To achieve this goal, the mutant form of 3CD prepared for these studies had the Gln/Gly at the 3C/3D cleavage junction mutated to Ala/Ala. The expression plasmid was transformed into competent BL21(DE3) E. coli cells. With the pET-30a expression vector, recombinant 3CD would be expressed with two Nterminal affinity tags of (His)6-tag and an epitope S-tag. The transformed E. coli cells were incubated at 257C in 21 TY plus kanamycin (50 mg/ml) until the OD600 reached approximately 0.6, and were then induced by the addition of 0.5 mM isopropyl b-D-thiogalactopyranoside (IPTG). Cells were grown for 6 h after induction and collected by centrifugation at 5000g for 20 min at 47C. Cell pellets were resuspended in column buffer (50 mM sodium phosphate, 300 mM NaCl, 5 mM imidazole, 10% glycerol, pH 8.0) at a concentration of 0.2 g cells/ml column buffer. Cells were then lysed by passage through a French pressure cell press. After lysis, supernatant was harvested by centrifugation at 45,000g for 30 min. Samples were treated with DNase at a concentration of 100 U/ml lysate for Ç12 h and dialyzed for 4 h against a 50-fold excess of column buffer with two changes at 47C. The dialyzed supernatant was filtered through a 0.45-mm cellulose–acetate membrane to remove precipitates. The 3CD proteinase present in the supernatant was purified in two steps by chromatography on HiTrap Ni2/-chelating resin and Superdex 75 column using a Pharmacia FPLC system. The supernatant was applied to a 5-ml prepacked Ni2/-chelating column, precharged with 100 mM NiCl2 , 20 mM sodium acetate, pH 5.0, and then equilibrated with 10 column volumes of column buffer. After sample application and column wash, proteins were eluted with a linear 5–300 mM imidazole gradient. After the affinity purification step, samples containing 3CD were pooled and applied to the Superdex 75 gel filtration column (3 1 70 cm) to separate 3CD from contaminating 3C and other low molecular weight contaminants. Amino terminal sequencing of purified 3CD was performed on an Applied Biosystems 477A protein sequencer (Applied Biosystems). Protein concentrations were determined by the Bradford method with bovine serum albumin (BSA) as a standard. SDS–polyacrylamide gel electrophoresis (SDS–PAGE) and Western analysis were utilized to monitor relative 3CD levels in the pooled fractions. SDS– PAGE was performed using 4–20% Tris–glycine gradient gels. Western blots probed for epitope S-tag were performed according to the manufacturer’s recommendations (Novagen). Western blots probed with the anti-3C proteinase antibodies were detected by the enhanced chemiluminescence method (Amersham). Prior to performing substrate and inhibitor kinetics, proteinases 3C and 3CD levels were normalized utilizing the Bio-Rad GS-700 Imaging Densitometer with BSA as standard, so that kinetics could be performed on equimolar amounts of these enzymes. All enzymes present in the preparations were assumed to be fully active. 3C and 3CD proteinase assays. Proteinase activities for 3C and 3CD were compared by determining their cleavage activities using peptide substrates under different assay conditions. Peptide substrates were solubilized in dimethyl sulfoxide (DMSO) and final DMSO concentrations in all assays were 2.5%. For HPLC method, six different peptides, including VP0/VP3 (RSKSIVPQ/GLPTTTLP), VP3/VP1 (SQTVALTE/GLGDELEE), 2C/3A (DSLETLFQ/GPVYKDLE), 3A/3B (YKLFAQTQ/GPYSGNPP), 3B/3C (RPVVVQ/GPNT since P8/P8* was not soluble under assay conditions), and 3C/3D (KQYFVEKQ/GQVIARHK), were synthesized as described previously (14). Cleavage reactions were performed as previously described (14) with slight modifications. Briefly, reactions with 1 mM peptide substrate and 120 nM enzyme were carried out in column buffer at 307C for 30, 60, and 120 min. Reactions were stopped by addition of 4 vol of a 10% acetonitrile/0.1% TFA (resulting in a pH

arcas

RHINOVIRUS PROTEINASES 3C AND 3CD õ2). Samples (50 ml) were injected into a HPLC (Shimadzu SCL10A) and eluted from a C18 column (4.6 1 250 mm) at 1 ml/min with a linear gradient of 10–40% acetonitrile/0.1% TFA in 20 min. Peaks were monitored at 214 nm. For the assays using chromogenic p-nitroanilide peptides, reactions containing 125 mM pNA peptide substrate and 120 nM enzyme were performed as described previously (15). Free pNA was released upon enzyme cleavage of these peptides; the reactions were monitored continuously at a visible wavelength of 405 nm. Fluorescence assays were performed as previously using peptides that contain fluorescence donor/quencher pairs at the concentration specified (14). Two different fluorogenic peptides, anthranilyl-TLFQ/GPVF(pNO2)K-CO2H (13) and H2N-DSE(Edans)EVLFQ/GPVK(Dabcyl)RD-CO2H were used in these experiments (14). Proteinase inhibitor studies. Classical proteinase inhibitors were purchased from Sigma. Effects of inhibitors on purified 3C and 3CD proteinase were examined in the presence of a single inhibitor concentration using the fluorogenic peptide anthranilyl-TLFQ/GPVF(pNO2)KCO2H as a substrate (13). Percentage inhibition, related to the control containing no inhibitor, was calculated based on a 30-min reaction. All peptide based inhibitors of the HRV14 3C proteinase, including LY338387, LY355455, and LY362270, were synthesized as described previously (16, 17).

RESULTS

While the HRV14 3C proteinase was available using published protocols (13), several major barriers hindered the expression and purification of HRV14 3CD including its low solubility, low level expression, and autoproteolysis. To find a suitable 3CD expression system, trial expressions of 3CD in E. coli utilizing several expression vectors under a variety of different conditions were performed (data not shown). These results indicated that the level of soluble expression was higher with the pET-30a system than with the other vectors tested. To optimize expression conditions, extensive efforts including temperature setting, growth medium choice, induction conditions, expression host, and extraction conditions had been made. Optimized 3CD expression conditions were created as described under Materials and Methods. Dramatic degradation of the native 3CD protein was found when expressed in bacterial cells (not shown), due to the presence of a 3C native cleavage site at the junction of 3C and 3D. In order to hinder self-proteolytic processing of 3CD to 3C and 3D during expression, mutagenesis of the 3C/3D cleavage site was performed. Previous studies with poliovirus 3CD have suggested that the highly conserved amino acids at P1, P1*, and P4 are important for 3C and 3CD cleavage activities (11, 18). In the case of HRV14, the 3C/3D cleavage site has a sequence of VEKQ/GQVI (P4/P4*). Therefore, six different mutations were made at the HRV14 3C/3D cleavage site including insertions between P2 and P3 and substitutions at P4, P1, and P1*. These include (i) Gly to Phe at P1*; (ii) Gln/Gly to Ala/Ala at P1/P1*; (iii) Val to Lys at P4; (iv) double mutation of Gly to Phe at P1* and Val to Lys at P4; (v) Lys insertion between P2

AID

ABB 0291

/

6b3f$$$$82

09-02-97 11:29:59

127

FIG. 1. Analysis of purified recombinant HRV14 3CD protein. Proteins were separated on a 4–20% gradient gel and then stained with Coomassie brilliant blue. Lane 1, molecular weight markers; lane 2, soluble extract of the transformed bacterial cell lysis (2.5 mg); lane 3, Ni2/ column fraction pool (0.25 mg); lane 4, Superdex 75 column fraction pool (0.1 mg). The arrow indicates the expected position of HRV14 3CD protein.

and P3; and (vi) Ser insertion between P2 and P3. The 3CD mutant proteins were then evaluated and ranked for their ability to hinder autocleavage. As determined by Western blot analyses, mutants (ii) and (iv) provided greater hindrance to 3CD breakdown than the others (data not shown). Since the dual substitution of Gln/ Gly at P1/P1* by Ala/Ala was a more conservative change than mutant (iv), the Gln/Gly to Ala/Ala mutant protein of 3CD was then further expressed, purified, and characterized. Purification of recombinant 3CD proteinase was achieved by chromatography on a Ni2/-chelating affinity column utilizing the 61 (His) tag at the amino terminus. Following the affinity purification, a gel filtration step was added to eliminate the contaminating proteins, especially 3C (Fig. 1). As seen in Fig. 1, sample eluted from the gel filtration column had the expected molecular weight of Ç75 kDa and N-terminal sequencing data confirmed the presence of 3CD (not shown). The two-step purification yielded 20–40 mg of purified 3CD protein per liter of transformed cells (data not shown). The purified 3CD could account for approximately 70% of the total protein in the enzyme preparation as determined by densitometric analysis. To investigate the possible effect of the 3D domain on the 3CD proteinase activity, peptide substrate specificity was compared to that of the 3C proteinase. Cleavage efficiencies between the two proteinases were systematically compared by utilization of a series of small synthetic peptides as substrates. There was detectable cleavage of all eight pNA peptides by 3C enzyme as shown in Table I;

arcas

128

DAVIS ET AL. TABLE I

Proteolysis Activity Comparisons between HRV14 3C and 3CD Proteinases

ployed. Since these compounds were synthesized by mimicking the 2C/3A cleavage site of 3C, these data imply that the active site conformation of 3CD was very similar to that of 3C.

(kcat /Km)rel Peptide code

Sequence

HRV14 3C

HRV14 3CD

pNA1 pNA2 pNA3 pNA4 pNA5 pNA6 pNA7 pNA8 pNA9

TLFQ–pNA ETLFQ–pNA EALFQ–pNA EVLFQ–pNA DSLETLFQ–pNA DSLEVLFQ–pNA DSRETLFQ–pNA DRRETLFQ–pNA RKGDIKSY–pNA

1.0 5.4 32.1 30.4 13.6 43.7 4.1 5.1 NC

NC NC 4.8 1.8 NC NC NC NC NC

Note. Amino acid sequences upstream of the hydrolysis bond of the pNA peptides are given in single-letter amino acid code. No side chains or N-terminal groups are blocked. Cleavage efficiency is expressed as (kcat/Km)rel using pNA1 as reference peptide. Data are the average (within {5%) of two independent determinations. ‘‘NC’’ indicates that there was no detectable cleavage. The pNA-9 peptide is derived from the HRV14 2A proteinase cleavage site.

however, an equimolar amount of 3CD proteinase cleaved only pNA3 and pNA4 peptides with the catalytic activities of 6 and 15% relative to 3C, respectively. Interestingly, 3CD could cleave pNA3 and 4, but not pNA2 of the same size (Table I). The only difference of pNA2 from pNA3 and pNA4 is that this peptide contains a polar amino acid at the P4 position, suggesting that 3CD prefers a nonpolar residue at this position as reported for the 3C enzyme (9, 10). In addition, cleavage activities of 3C and 3CD were found to be very similar toward the fluorogenic peptides anthranilyl-TLFQ/GPVF(pNO2)KCO2H and H2N-DSE(Edans)EVLFQ/GPVK(Dabcyl)RDCO2H (Fig. 2). These data indicate that both 3C and 3CD could tolerate the bulky moieties of fluorescent quencher and donor groups on these peptides. A series of unmodified peptides, representing the structural region cleavage sites of 3C, were tested for their ability to be cleaved by these two enzymes. Under the conditions employed, no detectable cleavage of these peptides was observed by either 3C or 3CD (not shown). Sensitivity of these proteinases to classical peptide inhibitors was investigated. Both 3C and 3CD protease activities were not affected by EDTA or EGTA. Consistent with previous studies (8, 14, 15), neither of these proteinases was sensitive to E-64, a specific papain-like cysteine proteinase inhibitor (Table II). Additionally, these proteases had very similar responses to some of the serine proteinase inhibitors as seen in Table II. Moreover, several inhibitors, specifically designed for the HRV14 3C proteinase (16, 17), could also strongly inhibit the 3CD proteinase under the conditions em-

AID

ABB 0291

/

6b3f$$$$83

09-02-97 11:29:59

DISCUSSION

In order to examine the impact of the 3D domain on 3CD proteinase activity, these two enzymes had to be successfully expressed and purified. Refolding and purifying the 3C proteinase had been very successful using the protocol described previously (13). A similar approach had been applied to the overexpressed 3CD since the intact form of 3CD could account for over 10% of granule proteins from bacterial cells. However, 3CD protein refolding was unsuccessful because a variety of different conditions all resulted in the formation of large visible aggregates (data not shown). The molecular weight of 3CD is approximately 75 kDa which may contribute to the refolding difficulties. In addition, compared to 3C, the 3CD contains five extra cysteines, which could form incorrect disulfide bonds and result in aggregate formation. Thus extensive efforts were then made to express 3CD as a soluble protein in E. coli. Even using the optimized conditions, the expression levels of soluble 3CD were very low; a Coomassie blue-stained 3CD band could not be visually resolved in crude lysate by SDS – PAGE when comparing it to a control lane that had no 3CD (data not shown). The 3CD purification protocol included

FIG. 2. Cleavage of fluorogenic peptide substrate by HRV14 3C and 3CD enzymes. Proteinase reactions were carried out with 120 nM of either 3C (n) or 3CD (s) and 35 mM fluorogenic peptide (DSE(Edans)EVLFQ/GPVK(Dabcyl)RD- at 307C for the time indicated under the conditions described in the text. Reaction containing no enzyme was run as control (h). All reactions were monitored continuously by the fluorometer and redrawn using measurements of 20% of the data points.

arcas

129

RHINOVIRUS PROTEINASES 3C AND 3CD TABLE II

Protease Inhibitor Sensitivity Comparison between HRV14 3C and 3CD Percentage inhibition Inhibitor

Inhibitor type

Inhibitor concentration

EDTA EGTA Egg white cystatin E-64 Iodoacetamide N-Ethylmaleimide Pepstatin A Benzamidine Leupeptin PMSF TLCK LY338387 LY355455 LY362270

Metalloproteinases Metalloproteinases Cys proteinase Cys proteinase Cys proteinase Cys proteinase Asp proteinase Ser proteinase Ser/Cys proteinase Ser/Cys proteinase Ser/Cys proteinase 3C Cys proteinasea 3C Cys proteinasea 3C Cys proteinasea

50 mM 50 mM 0.1 mg/ml 100 mM 1 mM 20 mg/ml 20 mM 10 mM 1 mM 4 mM 1 mM 0.22 mM 0.31 mM 10 mM

HRV14 3C

HRV14 3CD

0 0 0 0 53 84 0 36 98 88 52 19 72 43

0 0 0 0 67 86 0 41 50 87 60 27 73 87

a These peptide based inhibitors were developed specifically for HRV14 3C proteinase (16, 17). Their sequences are N-cbz-Phe-Met(SO2) for LY338387, Boc-Val-Leu-Phe-vGln-O-Me for LY355455, and cbz-Phe-vGln-Ome for LY362270. Protease assays were performed using the fluorogenic peptide anthranilyl-TLFQ/GPVF(pNO2) K- as a substrate under the conditions described under Materials and Methods. Control was run under the identical conditions except that only the solvent (no inhibitor) was included in the reaction.

metal chelating affinity and gel filtration chromatoThe availability of 3CD permitted the comparison graphic steps. It was a deliberate strategy to keep studies of 3C and 3CD proteinases. The kinetic experithe purification step simple in order to avoid further ments provided interesting data. Both enzymes exhibreduction of already low yield of the 3CD protein. For ited similar activities toward the fluorogenic peptide this reason, the 3CD was expressed with an amino substrates (Fig. 2). An important feature of these subterminal affinity tag to facilitate purification. After strates is that there are both P and P* residues. Interestthe combination of affinity and gel filtration steps, the ingly, 3C protease exhibited activity to most of the pNA 3CD was purified to approximately 70% with a yield peptides while 3CD had over sevenfold lower activity of 20 – 40 mg/liter of cell culture. In the future, efforts for only two of eight substrates (Table I). It is important to express 3CD in insect or mammalian cells may re- to point out that these substrates have no P* amino acid sult in higher yields of protein. residues. Hence, it is likely that P* residues are more Another reason for the low yield of the 3CD intact important for substrate recognition and proteolytic acprotein was its autocleavage at the junction of 3C and tivity in 3CD than in 3C. This might result from the 3D domains. Mutagenesis was performed in an effort presence of the 3D domain. On the other hand, both to hinder autocatalytic breakdown of 3CD to 3C pro- proteinases displayed a preference for nonpolar Ala or teinase and 3D polymerase. Previous work related to Val residues at the P4 position (Table I). Previous in poliovirus 3CD has shown that insertions between P2 vitro translation studies with poliovirus have indicated and P3 and substitutions at P4 can hinder 3CD au- that 3CD is the active proteinase for proteolytic cleavage tocleavage to 3C and 3D (19). Peptide substrate cleav- at the structural region cleavage sites (4, 5). Neither age selectivity of HRV14 3C has also indicated that the HRV14 3C nor 3CD proteinases cleaved the native pep3C proteinase prefers negatively charged amino acids tide substrates that represent structural region cleavage at P4, Gln at P1, and Gly at P1* (9, 10). Our observation sites on the basis of our experiments. One possible explathat mutations at P1, P1*, and P4 hinder HRV14 3CD nation for this observation is that a cellular cofactor breakdown is consistent with the observations made facilitates efficient cleavage of the structural region sites for poliovirus 3CD (19). Interestingly, in contrast to by the poliovirus 3CD (20). Generally, the inhibitor senpoliovirus 3CD, HRV14 3CD mutants with insertions sitivities were similar for the 3C and 3CD proteinases, between P2 and P3 prevented breakdown to 3C and 3D although the inhibitory levels of certain inhibitors on the least. These efforts not only facilitate expression of the two enzymes were slightly different (Table II). Imintact HRV 3CD but also confirm the notion that P1, portantly, all the HRV14 3C-specific inhibitors tested to P1*, and P4 of HRV14 3CD are important in proteinase date were able to inactivate the 3CD enzyme, indicating recognition. that there is a great potential for inhibitors specially

AID

ABB 0291

/

6b3f$$$$83

09-02-97 11:29:59

arcas

130

DAVIS ET AL.

designed for the 3C to inactivate the 3CD enzyme activity if this enzyme functions in viral replication. To have a better understanding of differences in substrate specificity between 3C and 3CD enzymes, structural investigations of these two proteinases may be essential. Since the X-ray structure of HRV14 3C has been solved (21), availability of the purified 3CD protein as described in this paper will facilitate the structural study of the HRV14 3CD protein. In conclusion, our data on substrate and inhibitor studies suggest that the active site of 3CD has a similar conformation to that of the 3C proteinase. The 3D domain has little effect on peptide and inhibitor recognition, except that it might slightly alter the binding of P1* residues to the enzyme. Thus, the 3C target is likely a good choice for developing anti-HRV compounds because inhibitors designed for 3C proteinase may have great potential to inhibit 3CD as well. ACKNOWLEDGMENTS We thank Mel Johnson for amino acid sequencing and Alex Konstantinidis, Wu Kuang Yeh, and Todd Parsley for helpful comments and suggestions.

REFERENCES 1. Couch, R. B. (1996) in Rhinoviruses (Fields, B. N., Knipe, D. M., and Howley, P. M., Eds.), Virology, 3rd ed., Lippencott–Raven, Philadelphia, PA. 2. Porter, A. G. (1993) J. Virol. 67, 6917–6921. 3. Palmenberg, A. C. (1990) Annu. Rev. Microbiol. 44, 603–623. 4. Ypma-Wong, M. F., Dewalt, P. G., Johnson, V. H., Lamb, J. G., and Semler, B. L. (1988) Virology 166, 265–270.

AID

ABB 0291

/

6b3f$$$$83

09-02-97 11:29:59

5. Jore, J., De Geus, B., Jackson, R. J., Pouwels, P. H., and EngerValk, B. E. (1988) J. Gen. Virol. 69, 1627–1636. 6. Harris, K. S., Xiang, W., Alexander, L., Lane, W. S., Paul, A. V., and Wimmer, E. (1994) J. Biol. Chem. 269, 27004–27014. 7. Xiang, W., Harris, K. S., Alexander, L., and Wimmer, E. (1995) J. Virol. 69, 3658–3667. 8. Orr, D. C., Long, A. C., Kay, J., Dunn, B. M., and Cameron, J. M. (1989) J. Gen. Virol. 70, 2931–2942. 9. Cordingley, M. G., Register, R. B., Callahan, P. L., Garsky, V. M., and Colonno, R. J. (1989) J. Virol. 63, 5037–5045. 10. Cordingley, M. G., Callahan, P. L., Sardanna, V. V., Garsky, V. M., and Colonno, R. J. (1990) J. Biol. Chem. 265, 9062–9065. 11. Harris, K. S., Reddigari, S. R., Nicklin, M. J. H., Hammerle, T., and Wimmer, E. (1992) J. Virol. 66, 7481–7489. 12. Blair, W. S., and Semler, B. L. (1991) J. Virol. 65, 6111–6123. 13. Birch, G. M., Black, T., Malcolm, S. K., Lai, M. T., Zimmerman, R. E., and Jaskunas, S. R. (1995) Protein Express. Purif. 6, 609– 618. 14. Wang, Q. M., Johnson, R. B., Cohen, J. D., Voy, G. T., Richardson, J. M., and Jungheim, L. N. (1997) Antiviral Chem. Chemother. 8, 303–310. 15. Wang, Q. M., Johnson, R. B., Cox, G. A., and Villarreal, E. C. (1997) Anal. Biochem., in press. 16. Venkatraman, S., Kong, J., Furness, K., Aube, J., Hanzlik, R. P., and Wang, Q. M. (1997) in Proceedings, Second Winter Conference on the Bioorganic and Medicinal Chemistry, Steamboat Springs, CO. 17. Shepherd, T. A., Cox, G. A., McKinney, E., Tang, J., Wakulchik, M., Zimmerman, R. E., and Villarreal, E. C. (1996) Bioorg. Med. Chem. Lett. 6, 2893–2896. 18. Roehl, H. H., Parsley, T. B., Ho, T. V., and Semler, B. L. (1997) J. Virol. 71, 578–585. 19. Blair, W. S., Nguyen, J. H., Parsley, T. B., and Semler, B. L. (1996) Virology 218, 1–13. 20. Blair, W. S., Li, X., and Semler, B. L. (1993) J. Virol. 67, 2336– 2343. 21. Mathews, D. A., Smith, W. W., Ferre, R. A., Condon, B., Budahazi, G., Sisson, W., Villafraca, J. E., Janson, C. A., McElroy, H. E., Gribskov, C. L., and Worland, S. (1994) Cell 77, 761–771.

arcas