VIROLOGY

185,779-787

Cloning

(1991)

and in Vitro Characterization R. MARGIS,*ut

*Institut de Siologie Strasbourg, France;

Molhxlaire and tfscola

of the Grapevine M. VIRY,* M. PINCK,*

Fanleaf Virus Proteinase AND

L. PINCK*”

des P/antes du CNRS et Universit& Louis Pasteur, Laboratoire Tknica Federal de Qufmica do Rio de Janeiro, Rua Senador Received

June

12. 133 1; accepted

August

Cistron

de Virologie. 12 rue du G&&al Zimmer, Furtado no 12 7, 20270 Rio de Janeiro-R/,

67084 Brazil

30, 799 7

The region of the genomic RNA-l from grapevine fanleaf virus isolate Fi 3 (GFLV-Fl3), containing the proteinase cistron and flanking sequences (nucleotides 3894 to 4789) of the GFLV polyprotein, was modified by PCR mutagenesis to create a start codon and cloned in a transcription vector. The transcripts from the resulting clone (pVP7) produced, upon translation in rabbit reticulocyte lysate, a 37.8-kDa protein which was subsequently cleaved to a stable 28-kDa product. Autocleavage was maximal at pH 7.0-8.5 and at 30”. Inhibition of the activity was greater than 80% when translation was performed in the wheat germ system. In rabbit reticulocyte lysate, inhibition was also obtained with PMSF, EDTA, E-64, Ca+*, Zn+*, and Co+‘. The pVP7 translation product acts in cis, in the case of its autocleavage, or in trans in the processing of the viral 122-kDa polyprotein from GFLV RNA-2 into a 66-kDa protein and the 56-kDa coat protein. The carboxy extremity of the complete pVP7 translation product, encoded by nucleotides 4633 to 4789 of RNA-l, was not required for the proteinase activity, at least in trans. 0 19% Academic PWSS. IN.

INTRODUCTION

Virus-encoded proteinases from nepo-, coma-, and potyviruses form part of a large group of picornavirus 3C-related proteinases (Wellink and Van Kammen, 1988; Kay and Dunn, 1990). These are cysteine proteinases, as inferred from inhibition studies, characterized by a conserved putative active site containing histidine, aspartic, or glutamic acid and cysteine residues (Bazan and Fletterick, 1988; Gorbalenya et a/., 1989). Despite structural similarities, these proteinases differ in specificity of the cleavage site. For example, tomato black ring virus, the type member of the nepovirus group, grapevine chrome mosaic virus, and GFLV proteinases recognize different cleavage sites involved in production of capsid protein (Brault et a/., 1989; Demangeat et al., 1990; Serghini et a/., 1990). Here we report the cloning of the region of GFLV-F13 RNA-l corresponding to the viral proteinase. Some of the characteristics of the cloned proteinase including optimal temperature and pH and activity in the presence of inhibitors were established. Intramolecular cleavage (&-activity) of the cloned proteinase and intermolecular cleavage @fans-activity), i.e., its ability to process polyprotein P2 of two different GFLV isolates, were also investigated.

Grapevine fanleaf virus (GFLV) is a member of the nepovirus group with a bipartite RNA genome of positive sense. Viral RNAs are polyadenylated at their 3’ ends and covalently attached to a small protein (VPg) at their 5’ extremities (Pinck et a/., 1988). GFLV RNA-l and RNA-2 code, respectively, for polyproteins Pl of 253 kDa (Ritzenthaler et a/., 1991) and P2 of 122 kDa (Serghini et a/., 1990). Morris-Krsinich et a/. (1983) showed that when 35S-labeled polyprotein P2 is synthesized in a rabbit reticulocyte lysate system and mixed with nonradioactive RNA-l translation products, two radioactive proteins of 68 and 58 kDa (coat protein) are produced, demonstrating that a proteolytic activity is associated with the GFLV RNA-l polyprotein. The determination of the nucleotide sequence of GFLV-F13 RNA-2 and of the N-terminal sequence of the coat protein permitted the location of the coat protein cistron in the carboxy-terminal part of polyprotein P2 and implied that mature coat protein was produced by an arginine/ glycine cleavage (Serghini et al., 1990). Recently, the amino acid sequence of the VPg encoded by RNA-l (from serine 12 18 to glycine 1241 polyprotein Pl numbering) has been determined (Pinck et a/., 1991). GFLV-F13 RNA-l has also been sequenced and consensus regions for a nucleotide triphosphate binding protein, the proteinase, and the RNA polymerase were identified in the polyprotein Pl sequence (Ritzenthaler et al., 1991). ’ To whom dressed.

correspondence

and reprint

requests

should

MATERIALS

AND METHODS

Virus, bacteria, and plasmids Grapevine fanleaf strain F13 (Pinck et al., 1988) and strain Tu (Fuchs et a/., 1991) were routinely multiplied in Chenopodium quinoa, a herbaceous host. Escherichia co/i strains NM 522 and JM 109 were trans-

be ad-

779

0042-6822191

$3.00

Copyright 0 1991 by Academic Press. Inc. All rights of reproduction in any form resewed.

780

MARGIS ET At

formed by electroporation (Pfau and Youderian, 1990). All constructions were cloned in the phagemid BlueScribe m 13 plus (BSt) (Stratagene). cDNA synthesis

and cloning

A full-length cDNA from GFLV RNA-l, cloned in the Smal site from BS+ (clone pFLl48), was obtained by primer extension for the first strand and doublestranded cDNA was synthesized as already described (Serghini eta/,, 1990). The double-stranded cDNA was ligated to Smal-cut 5’ dephosphorylated BS+. Aliquots of the ligation medium were used to transform f. co/i DH~cY. Polymerase

chain reaction

(PCR) mutagenesis

Clone pFL148 was used as template to amplify the GFLV proteinase cistron. PCR was performed as follows: 10 ng of double-stranded template DNA, 100 ng of each primer (P2661 and P2667 shown in Fig. l), 8 ~1 of 2.5 mM dNTPs, 10 ~1 10 x PCR buffer (100 m/M Tris-HCI, pH 8.3, 0.1% gelatin, 15 mlLl MgCI,, 500 mN1 KCI) and 5 units of Taq DNA polymerase (Pharmacia) were mixed and the volume was adjusted to 100 ~1 with water. PCR was performed in a programmable thermal cycler (M. J. Research, Inc.). A first denaturation step (94”, 5 min) was followed by 30 to 40 cycles of denaturation (94’, 90 set), hybridization (primer fusion temperature -5”, 90 set), and elongation (72”, 90 set). In vitro transcription

and translation

Uncapped transcripts were synthesized from linearized plasmid DNA using T7 RNA polymerase (Stratagene) by the method of Tabor and Richardson (1987). The size and integrity of transcripts were checked by electrophoresis under denaturing conditions in a lob agarose gel (Gustafson et al., 1982). Transcripts (- 1 pg/pI) were translated either in the wheat germ system prepared as described by Godefroy-Colburn et al, (1985) or in the nuclease-treated rabbit reticulocyte lysate system (NT-RRLS) obtained from Promega (Madison). Translations in the reticulocyte system were performed with an incubation mixture composed of 35 ~1 of NT-RRLS, 1 ~1 of a 1 mlLl amino acid mix (methionine free, Promega) and 60 &i of [35S]methionine (1000 Ci/mmol, Amersham). Four-microliter aliquots of the above translation mixture were mixed with 0.5 to 1 .O ~1 of transcripts, the final volume of each adjusted to 5 PI with water and translation was carried out for different time intervals ranging from 5 to 60 min at 30”. At the end of the incubation period incorporation of radioactive methionine was halted by isotopic dilution by the addition of 0.5 PI nonradioactive 20 mM methio-

nine and the reactions were stopped by the addition of an equivalent volume of electrophoresis loading buffer (1 O”b (w/v) sodium dodecyl sulfate, 25% (v/v) /3-mercaptoethanol, and 160 mM Tris-HCI, pH 6.8) (Demangeat et al., 1990). Translation with nonradioactive amino acids was performed as above but in the presence of 40 @ methionine. Samples were analyzed on polyacrylamide gels (PAG) of appropriate percentage, containing SDS, as described by Laemmli (1970). Marker proteins 170, 97.4, 55.4, 36.5, and 20.1 kDa were from Combithek calibration kit (Boehringer); 45kDa egg albumin and 29-kDa carbonic anhydrase were from Sigma. Determination

of proteinase

autocatalytic

activity

Studies of the proteinase activity were performed as follows: transcripts were translated for 15 min in the NT-RRLS, an incubation time which provided the maximal difference in labeling intensity between the 37.8and the 28-kDa translation products. The synthesis of labeled protein was stopped by adding 0.5 1~~1 of cold 20 mM methionine to the incubation medium. Finally, compounds to be tested such as buffers and inhibitors were added to the translation mix and incubation was continued for 2 to 4 hr at 30”. The extent of VP7 autocleavage activity was quantified by densitometry of the 37.8-kDa nonprocessed proteinase precursor and of the 28-kDa product present on autoradiograms of PAG containing SDS. The 28-kDa/37.8-kDa ratio was used to express the VP7 autocleavage activity in a given condition, taking into account the relative Met content of each component. Densitometry at 560 nm and calculation of the peak areas were performed with a Shimadzu CS9000 densitometer (Shimadzu Corp., Kyoto). To determine the concentration effect on pre-VP7 processing a series of dilutions from 2- to 80-fold was performed on the translation products of VP7-Pvull transcripts 15 min after the translation started. The dilutions were obtained by adding an appropriate volume of either dilution buffer (100 mn/l Tris-acetate, pH 7.6, 5 mM DlT, 5 mlLI magnesium-acetate, 5 mM potassium-acetate) or rabbit reticulocyte lysate to 0.5 ~1 of 35S-labeled translation products. The diluted samples were incubated for 2 hr at 30”, the final volume was adjusted to 40 ~1 with dilution buffer or reticulocyte lysate, and the reaction was stopped by the addition of 40 ~1 of electrophoresis loading buffer. The samples were stored at -80” until analysis in 10%~ PAG containing SDS. The proteinase activity was quantified by densitometry as described above. Inhibitor

and pH studies

The following proteinases inhibitors were used in this study: leupeptine, soybean trypsin inhibitor, egg

GFLV

PROTEINASE

CLONING

AND

CHARACTERIZATION

781

BamHI ,....

P D

P

K P

P A

a.a.1516 + I L

/.....TCCTGATCCTAAGCAGCCCGCCATCCTTAGCGCAGAGG.....

I

3’......ACGAAGATATGGCACAAGTCTCGGAGCGlVLCCTCCTTTCTATA~GG.....

M A

4 nt 3817 *

s

E P

nt 4801 +

R L

E

E . . . . . . . . . . . . . . . t‘

4 a.e.1218

PCR

PstI m d:

I

912 bp

I

0 pBS+ 3204 bp

FIG. 1. Representation of the part of the GFLV RNA-l nucleotide sequence containing the PCR primer hybridization regions (nucleotides 3877 to 4801) used in the construction of VP7. The nine N-terminal residues and the nine C-terminal residues of the open reading frame generated by the PCR mutagenesis are indicated. Mutated amino acids are in bold type and amino acid numbers are those of GFLV polyprotein Pl (Ritzenthaler et a/., 1991). The principal steps of cloning the PCR product are presented. CIP is the calf intestine phosphatase. AUG and UAG indicate the start and stop codon in the pVP7 transcript.

trypsin inhibitor, and lung trypsin inhibitor, from Boehringer-Mannheim. Ethylenediamine tetraacetic acid (EDTA), phenylmethanesulfonyl fluoride (PMSF), and the [/I-(L-3-trans-carboxyoxiran-2-carbonyl)-L-leucyl]-amido(4-guanido)butane (E64) were from Sigma Biochemicals. Effect of pH on the proteinase activity was studied by the addition of HEPES, MOPS, or Tris buffers at 100 m/l/lfinal concentration to 5.5 ~1of translation medium containing cold methionine. RESULTS Mutagenesis and cloning proteinase region

of the GFLV-F13

A full-length cDNA (pFL148) from GFLV-F13 RNA-l was prepared and cloned in the Smal site of BlueScribe plus (BS+). The nucleotide sequence of pFL148 was shown to be identical to that already described (Rizenthaler et a/., 1991) by sequence and by restriction map analysis. PCR mutagenesis was used to isolate the proteinase region using pFL148 full-length clone as template and two primers (2661 and 2667) to introduce specific mutations (Fig. 1). The 912-bp PCR product differs from the pFL148 template by the presence

of three new cloning sites introduced with the primers: BarnHI and Ncol at one extremity of the insert and Pstl at the other. Figure 1 illustrates the cloning steps for the insertion of this fragment into the transcription vector BS+. As the Ncol site contains an ATG start codon a new open reading frame (ORF) is generated (Fig. 1). This new ORF is in phase with the reading frame of the GFLV polyprotein Pl and includes the region of the viral VPg, the proteinase, and the N-terminal part of the RNA polymerase (Ritzenthaler et a/., 1991). In addition to the viral sequence the ORF contains two additional upstream codons (Met and Ala) introduced during the cloning process (Fig. 1). As no in-phase termination codons were present in the cloned fragment the ORF continues for 45 amino acids downstream into the BS+ sequence until an in-phase termination codon is encountered (Fig. 1). Synthesis proteins

and processing

of pVP7 translated

Digestion of pVP7 by Pvull followed by transcription with bacteriophage T7 RNA polymerase produced a transcript, tVP7-Pvull, containing an ORF encoding a

MARGIS

782

ET AL.

100

*..‘.‘.‘..“...“..‘..“,..’ * pre-VW

80 60 pre-VP7

40 20 -

,VP7 Oi...I...l...~...I...l...,... 0 20 40 60 80 100 Incubation Time (min) 1

2

3

4

120

140

5

FIG. 2. Kinetics of synthesis and processing of [35S]methionine translation products from tVP7-Pvull transcript in the NT-RRLS. (a) Samples were analyzed in a 10% PAG containing SDS and autoradiographed 4 hr at room temperature. Position of molecular weight markers (in kDa) are indicated to the right. The major translation product (pre-VP7) and its maturation product (VP7) are indicated. (b) Graphic representation of the processing of pre-VP7 into VP7 and 31 K protein as measured by the relative band intensity of preVP7, VP7, and 31 K after increasing incubation times.

protein with a calculated molecularweight of 37.8 kDa. Translation of tVP7-Pvull in the nuclease-treated rabbit reticulocyte lysate system (NT-RRLS) produced, after 7.5 min of incubation, a major protein (pre-VP-I), with apparent molecular weight of 38 kDa in a 1O”~ PAG containing SDS and a minor species of 31 kDa (Fig. 2a). A kinetic analysis, in which translation times were varied from 7.5 to 120 min showed that both the preVP7 38-kDa protein and the 31-kDa species accumulated in increasing amounts for the first 15 min. At later times both bands progressively decreased in intensity and a major new band (VP7), with an apparent molecular weight of 29 kDa, appeared and became abundant (Fig. 2a). The 31-kDa protein could correspond to a product of an internal initiation, due to the relatively low level of fidelity of the reticulocyte lysate translation system (Dasso and Jackson, 1989) or to an intermediate in the cleavage pathway between pre-VP7 and VP7. The VP7 band progressively accumulated as the 38-kDa protein (pre-VP7) disappeared (Fig. 2b), suggesting that the 38-kDa protein was cleaved to produce the 29-kDa VP7 protein. As the pre-VP7 protein contains the VPg, the proteinase and a part of the polymerase cistron, autocleavage of pre-VP7 by its own proteinase region may be occurring. Serghini et a/. (1990) proposed that viral polyprotein P2 is cleaved at an arginine/glycine site to produce the capsid protein. An Arg/

Gly site (residues 1460-l 461 of polyprotein Pl) is also present in Pl near the point predicted, upon the basis of the apparent molecular weight of the product, to mark the boundary between the proteinase and the polymerase cistrons. It should be noted, however, that a second potential cleavage site (Gly/Glu; residues 146 l-l 462) is separated from the Arg/Gly site by only one residue in the Pl polyprotein sequence. Proteinase cleavage at this site would recognize the same sequence as for the C-terminus of VPg (Pinck et al., 1991). If one of these sites is recognized in pre-VP7 and cleaved by the proteinase, a protein with a calculated molecular weight of about 27.8 kDa containing the VPg and proteinase would be generated. The other predicted product of the reaction (calculated molecular weight = 10 kDa) is too small to detect in the PAGE system used. To test the above hypothesis concerning location of the cleavage site, two transcripts (tVP7-Mel and tVP7Accl) shorter than tVP7-Pvull were synthesized (Fig. 3a) by linearizing the template DNA with other restriction enzymes. The Mel site was chosen for its proximity to the predicted proteinase-polymerase border; its translation product should be a peptide only three or four amino acids longer than the putative 27.8-kDa VPg-proteinase product derived from pre-VP7 cleavage (Fig. 3a). The Accl site, on the other hand, is lo-

GFLV

PROTEINASE

CLONING

AND

CHARACTERIZATION

tvF7

(a) Q GFLV-F13

?

N&I

783

tvP7 AccI

tvp7 PVUII

W

CS GE RG

RNA-1

t

Barn 4 I

PstI

-

pre-VP7

-VP7 MASFE

GGmv

I PBS+ LQJ NTP binding = vpg u protease tia polymerase 0 sToPcodon

123456789

FIG. 3. Location of the pre-VP7 cleavage site deduced from the nature of the maturation products of pVP7 partial transcripts. (a) Genomic organization of GFLV-F13 RNA-l. The different functional domains are represented and the known C/S and G/E and the postulated RIG cleavage sites are indicated in the upper part. Arrows underneath indicate the limits of the cloned region from the original pVP7. Endonuclease restriction sites (Pvull, Mel, and Accl) used to generate the different pVP7-derived transcripts are indicated on pVP7. The polypeptides corresponding to each transcript are represented as a dark line on which the five amino acids at their amino and carboxy-terminus and the calculated molecular weight of each polypeptide are specified. (b) Autoradiography from a 12.5% PAG containing SDS, after 12 hr exposure. The three different transcripts tVP7-Mel, tVP7-Accl, and tVP7-Pvull were translated for 15 min in the NT-RRLS in the presence of [35S]methionine and then either frozen at -80” or incubated for 4 or 24 more hr at 30” after addition of cold methionine.

cated within the proteinase cistron and the resulting translation product should lack the C-terminal portion of the proteinase and, in particular, cysteine-1420 which may form part of the proteinase catalytic triad (Fig. 3a). A cysteine residue is conserved in an analogous C-terminal position of many related viral proteinases (Bazan and Fletterick, 1988). The translation products produced from the truncated transcripts at different times of incubation are shown in Fig. 3b. The tVP7-Noel transcript, with a coding capacity of 28.2 kDa, is translated after 15 min of incubation to yield a 48-kDa protein whose origin will be discussed below and a 29-kDa protein which is gradually processed into a protein with slightly faster mobility, comigrating with the 27.8-kDa VP7 protein produced by tVP7-Pvull (Fig. 3b, lanes l-3). The size of this protein is as expected from the removal of the 3 or 4 C-terminal residues from the initial 28.2-kDa translation product after cleavage at the aforesaid argininejglycine or glycine/glutamic site (Fig. 3a). Translation of tVP7-Accl gave a major stable band at 21 .l kDa (Fig. 3b, lanes 4-6) and a protein of 42 kDa after a 15 min incubation. The bands of 48 kDa (lane 1) and of 42 kDa (lane 4) diminished after longer incubation times. As no proteins were synthesized in a translation assay without added tran-

scripts (not shown), the origin of the 48- and 42-kDa proteins could be related to the presence of transcripts longer than expected that were detected as minor bands in denaturing agarose gels. Effect of temperature

and pH

The proteinase activity of VP7 was initially detected by addition of pVP7 translation products to the translation mix itself, at 30°, the optimal temperature for translation in NT-RRLS (Promega technical bulletin). In order to obtain maximal activity and to gain insight into the characteristics of the proteinase, autocleavage of pre-VP7 was studied at different pHs and temperatures (data not shown). Activity was maximal at pH 8.0, but extensive activity was obtained over the neutral-alkaline range of pH 7.0-9.0. A sharp decrease of activity was observed when pH changed from 7.0 to 6.5. PreVP7 presented maximal autocatalytic activity at 30”. The activity decreased at high temperature: 40% at 37” and 90% at 45”. The rest of the experiments reported in this paper were performed under the optimal pH and temperature conditions. Inhibitors of proteinase activity To obtain additional information concerning the nature of the VP7 proteinase activity, we investigated the

784

MARGIS TABLE

1

EFFECTS OF INHIBITORS IN VP7 MATURATION Compound Leupeptln PMSF Egg trypsin inhibitor Lung trypsrn Inhibitor Soybean trypsin inhibitor E-64 E-64 E-64 Wheat germ Wheat germ Wheat germ Dn EDTA Zn sulfate Ca chloride Co chloride Mg acetate Mg acetate K chloride K acetate

Concentration 1 mM 2mM 100 pg/ml 100 pg/ml 100 @g/ml 0.1 mM 0.5 mM 2.0 mM 0.5 @I 1 .o jll 2.0 fil 1 .O mM 1 .O mM 1 .O mM 1 .O mM 1 .O mM 1.0 m/l/l 10.0 mM 1 .O mM 1 .O mM

% Inhibition 26 32 45 38 27 0 5 93 53 74 85 24 88 89 94 96 0 12 0 0

effects of various proteinases inhibitors and possible activators on the processing of pre-VP7 into VP7 (Table 1). Specific trypsin inhibitor from chicken egg, bovine lung, and soybean presented a moderate degree of inhibition at high concentration (100 pg/ml). Leupeptin, a peptidyl aldehyde that inactivates serine proteinases as well as cysteine proteinases (Rich, 1986) has no significant inhibitory effect. PMSF, a general serine proteinase inhibitor (Powers and Harper, 1986) produced only mild inhibitory effects. The epoxide E-64, a very specific and irreversible inhibitor of cysteine proteinase (Rich, 1986), was inhibitory at a concentration greater than 0.5 mM (Table 1). A proteinase inhibitor able to prevent maturation of cowpea mosaic virus (CPMV) polyprotein was previously reported to exist in wheat germ extracts (Shih et al., 1987). Inhibition of the maturation of pre-VP7 ranging from 53 to 89% was obtained when increasing quantities of wheat germ extract were added to a tVP7Pvull NT-RRLS incubation mixture 15 min after the start of translation (Table 1). If translation was performed in the wheat germ system, an almost complete inhibition of maturation was observed even after doubling the amount of tVP7-Pvull translated (Fig. 4a). Dithiothreitol, previously reported as an activator of cysteine proteinases such as papain and cathepsin (Barret, 1986) produced no activation; on the contrary, a mild inhibitory action was observed (Table 1). The action of different metal ions on VP7 proteinase activity was also investigated (Fig. 4b, Table 1). Four

ET AL.

divalent cations, zinc, cobalt, calcium, and magnesium, were tested. Zinc, cobalt, and calcium ions inhibited pre-VP7 processing with high efficiency. Zinc inhibition of proteinase activity was previously described for papain (Barret, 1986) for the 49-kDa proteinase from tobacco etch virus (TEV) (Dougherty et a/., 1989) and for CPMV proteinase (Peng and Shih, 1984). Calcium, on the other hand, is normally associated with activation of enzyme activities (Barret, 1986). Magnesium presented no inhibitory effect even at 10 m/v!. EDTA, a chelating agent which is an inhibitor of metallo-proteinases, decreased processing of pre-VP7 to VP7 by 88% (Table 1). As magnesium acetate was present in the NT-RRLS at 0.7 mM final concentration, we decided to verify if the observed EDTA inhibition of pre-VP7 processing was due to its ability to sequester the added Mgf2 ions. Figure 4b shows the pre-VP7 processing under conditions in which the EDTA concentration was kept constant (2 mn/l) and Mg+2 increased from 0.5 to 4 mlVI. The observed reversal of EDTA inhibition by increasing magnesium acetate concentration indicates the importance of Mgf2 to maintain enzyme activity. Chloride ions have been reported to inhibit the poliovirus proteinase-2A even in trace amounts (Konig and Rosenwirth, 1988), but VP7 proteinase activity is unaffected as demonstrated by incubation with KCI (Table 1). Potassium acetate was used as a control to demonstrate that K+ had no action on pre-VP7 processing (Table 1).

0)

a-2mMEDTA b - EDTA + 0.5 mM MgAc c - EDTA + 2.0 mM MgAc d - EDTA + 4.0 mM MgAc FIG. 4. Effects of different Inhibitors on pre-VP7 processing. (a) Autoradiography of samples corresponding to the translation products obtained from 2 ~1 (lane 1) and 4 ~1 (lane 2) of tVP7-Pvull in wheat germ system, analyzed in a 12.5% PAG containing SDS and autoradiographed 16 hr at room temperature. Position of molecular weight markers (in kDa) are indicated to the left. The major translatron product (pre-VP7) and its maturation product (VP7) are indicated. (b) The processing of pre-VP7 under conditions in which EDTA concentration was always constant (2 mM) and Mg+* increased from 0.5 to 4 mM.

GFLV

VP7 proteinase

PROTEINASE

CLONING

AND

785

CHARACTERIZATION

acts in cis and in tram

The proteolytic cleavage produced by VP7 could occur by two different mechanisms, i.e., intermolecular cleavage (in rrans) or intramolecular cleavage (which acts in cis). Palmenberg and Rueckert (1982) have demonstrated a cis cleavage activity for protein C of encephalomyocarditis virus by a serial dilution experiment which revealed that the proteolytic activity of protein C is independent of the molar concentration of the enzyme and the substrate. To determine whether the proteolytic cleavage of pre-VP7 into VP7 is sensitive to concentration effects, viral protein synthesized by in vitro translation was allowed to undergo processing over a 2-to 80-fold range of dilutions. The starting sample was a VP7 transcript translated for 15 min. Two sets of dilution tests were performed. In the first, the 15 min translation products were diluted with Tris-acetate buffer; in the second test, dilution was with reticulocyte lysate as described under Materials and Methods. In both experiments no significant decrease of the pre-VP7 processing was produced by the successive dilutions (data not shown). These results support the hypothesis of an intramolecular self cleavage, or cleavage in cis, of the pre-VP7 to produce VP7. Since the GFLV proteinase is assumed to be responsible for the maturation of both polyprotein Pl and polyprotein P2, the ability of VP7 proteinase to process polyprotein P2, i.e., to act in trans, was investigated. A tram activity has already been demonstrated by incubation of GFLV RNA-l and RNA-2 translation products (Morris-Krsinich eta/., 1983). These authors reported a proteinase activity, associated with RNA-l, able to process GFLV polyprotein P2 to form the capsid protein and a 68-kDa protein equivalent to the 66-kDa protein reported here. We incubated the translation products of GFLV total RNA with proteins translated from tVP7Pvull to assess Vans activity of VP7. GFLV RNAs and tVP7-Pvull, translated separately during 45 min, were then mixed and incubation was continued for 2 hr at 30”. Two different strains of GFLV, GFLV-F13 and GFLV-Tu, were used. Bands corresponding to the capsid protein and the 66-kDa protein were obtained with both virus strains (Fig. 5). The 66-kDa protein corresponds to the N-terminal part of the 122-kDa protein translated from GFLV RNA-2, as established from translation of RNA-2 transcripts (data not shown). Figure 5 (lanes 1 and 3) demonstrates that when total GFLV RNAs were translated without the addition of VP7 translation products, no polyprotein processing is visible. This observation suggests that VP7 proteinase is synthesized in much smaller amounts from GFLV polyprotein Pl than from the VP7 transcript so that its activity is below the level of detection in these experi-

170 97.4 I

-F13 -Tu

36.5

1

2

3456

FIG. 5. Proteinase activity in trans of the translation products of VP7 transcripts. Autoradiography of a 10% PAG containing SDS exposed 16 hr at room temperature. RNAs from GFLV strains Tu (lanes 1 and 2) and F13 (lanes 3 to 6) were translated in the NT-RRLS in the presence of [35S]methionine during 45 min. Samples were mixed with an equal volume of the nonradioactive translation products of tVP7-Pvull from a 45 min incubation (lanes 2 and 4) tVP7-Mel (lane 5), tVP7-Accl (lane 6) or with water (lanes 1 and 3). Positions of molecular weight markers (in kDa) are indicated to the left. The major translation product from GFLV RNA-2, the polyprotein P2 and its maturation products, the 66K protein and the capsid protein (CP), are indicated. The migration position of capsid protein from purified virions of GFLV strains F13 and Tu are indicated to the right (CP-F13 and CP-Tu).

ments. The tVP7-Noel translation product, corresponding to the VPg-proteinase peptide with only four supplementary amino acids at its carboxy end, is also able to process GFLV-F13 polyprotein P2 (Fig. 5, lane 5). The tVP7-Accl translation product, however, in which the proteinase carboxy extremity has been deleted, was unable to process viral polyprotein P2 (Fig. 5, lane 6).

DISCUSSION The characteristics of the VP7 proteinase from GFLV-F13 reported above do not permit this proteinase to be classified among the four conventional proteinase groups: serine, cysteine, aspartic, and metallo

786

MARGIS

proteinases (Barrett, 1980). A similar situation also exists for the proteinases of TEV (Dougherty et al., 1989) and the CPMV proteinases (Verver et al., 1987). The pH profile of VP7 proteinase activity and the inhibitory effect of zinc and cobalt ions on pre-VP7 processing are distinct from the aspartic or metallo-proteinases, since aspartic proteinases function at acidic pHs and zinc and cobalt ions are indispensable cofactors for metallo-proteinase activity. The strong inhibition obtained with E64, an epoxysuccinylpeptide that attaches covalently to the cysteine residues of the active site of cysteine proteinase, indicates that VP7 proteinase contains a cysteine at its active site. Bazan and Fletterick (1988) have compared the viral cysteine proteinases to the trypsin-like family of serine proteinases and proposed a model in which the nucleophilic active site of the catalytic triad, Ser-195, is changed to a Cys residue in the viral proteinases. The amino acid sequence of the part of the polyprotein Pl corresponding to the VP7 proteinase is consistent with this model (Ritzenthaler et a/., 1991). VP7 proteinase from GFLV has an unusual characteristic since it requires the presence of Mg Q ions but not Caf2, as do some of the cysteine and serine proteinases. The role of magnesium could be that of an activator as are calcium ions for clostripain, calpains, or factor X (Barret, 1986). Purified VP7 proteinase would be necessary to determine if magnesium acts as a cofactor or if it changes the enzyme conformation at the active site. The cleavage of the P2 polyprotein to yield the 66kDa protein and the coat protein indicates that the VP7 proteinase recognized this R/G site in a specific way among the various R/G linkages present in the P2 polyprotein. The apparent molecular weight of VP7 (27.8 kDa) resulting from the pre-VP7 processing suggests that the G/E link between the VPg and the proteinase, established from VPg sequencing (Pinck et al., 1991), is not cleaved by the VP7 proteinase under our conditions. We assume therefore that the cleavage between the proteinase and the polymerase in pre-VP7 to yield VP7 occurs at the R/G site. However, since no amino acid data for VP7-processed products are available cleavage at the G/E site cannot be strictly ruled out. Demangeat et al. (1991) demonstrated that TBRV RNA-l translation products are able to cleave polyproteins encoded by TBRV RNA-2 and GCMV RNA-2, suggesting that the cleavage sites in these two polyproteins are similar but probably distinct from those of the GFLV-F13 polyprotein P2, which was not cleaved. The cloning of the GFLV-F13 proteinase will permit us to focus on the maturation process of viral polyproteins Pl and P2, the identification of the peptides produced and their specific cleavage sites, an objective which

ET AL.

was difficult up to now because of the low proteinase activity associated with intact GFLV polyprotein Pl .

ACKNOWLEDGMENTS The authors are grateful to Dr. K. Richards for Improving the manuscript. Rogerio Margis has a grant from National Counceil of Research and Technology (CNPq, Brazil). The EMBL Databank accession number for grapevine fanleaf virus RNA1 is: DO091 5:GFVRNAl.

REFERENCES BARRET, A. J. (1986). An lntroductlon to the proteinases. In “Proteinases inhibitors” (A. J. Barrett and G. Salvesen, Eds.). pp. 3-22. Elsevier. New York. BAZAN, 1. F., and FLE~ERICK, R. 1. (1988). Viral cysteme proteases are homologous to the trypsin-like family of serine proteases: structural and functional implications. Proc. Nat/. Acad. Sci. USA 85, 7872-7876. BRAULT, V., HIBRAND, L., CANDRESSE, T., LE GALL, 0.. and DUNEZ. J. (1989). Nucleotide sequence and genetic organization of Hungarian grapevine chrome mosaic nepovirus RNA2. Nucieic Acids Res. 17,7809-78 19. DASSO. M. C., and JACKSON, R. J. (1989). On the fidelity of mRNA translation in the nuclease-treated rabbit retlculocyte lysate system. NucleicAcids Res. 17, 3129-3144. DEMANGEAT, G.. GREIF, C., HEMMER, O., and FRITSCH, C. (1990). Analysis of the In vitro cleavage products of the tomato black ring virus RNA-1 -encoded 250K polyprotein. I Gen. Viral. 71, 1649-l 654. DEMANGEAT, G., HEMMER, O., FRITSCH, C.. LE GALL, O., and CANDRESSE, T. (1991). In vitro processing of the RNA-2 encoded polyprotein of 2 nepoviruses: tomato black ring virus and grapevine chrome mosaic virus. i. Gen. Viroi. 72, 247-252. DOUGHERTY, W. G., PARKS, T. D., CARY. S. M., BA~AN. J. F., and FLETTERICK, R. 1. (1989). Characterization of the catalytic residues of the tobacco etch virus 4%kDa proteinase. virology 172, 302-310. FUCHS, M., PINCK, M., SERGHINI, M. A., RAVELONANDRO, M., WALTER, B., and PINCK, L. (1989). The nucleotlde sequence of satellite RNA in grapevine fanleaf virus, strain F13. J. Gen. Viral. 70, 955-962. GODEFROY-COLBURN, T., THIVENT, C.. and PINCK, L. (1985). Translational discnmlnation between the four RNAs of alfalfa mosaic virus. A quantitative evaluation. fur J. Biochem. 147, 541-548. GORBALENYA, A E., DONCHENKO, A. P., BUNOV. V. M., and KOONIN, E. V. (1989). Cysteine proteases of positive strand RNA viruses and chymotrypsln-like serine proteases. FEBS Letf. 243, 103114. GORBALENYA. A. E., and KOONIN, E. V. (1989). Viral proteins containing the purine NTP-binding sequence pattern. Nucleic Acids Res. 17,8413-8440. GUSTAFSON, G. D., MILNER, J. L., MCFARLAND, J. E., PEDERSEN. K., LARKINS, B. A., and JACKSON, A. 0. (1982). Investigation of the complexity of barley stripe mosaic virus RNAs with recombinant DNA clones. Virology 120, 182-l 93. KAY, J., and DUNN, B. M. (1990). Viral proteinases: weakness in strength. Bochim. Biophys. Acta 1048, l-1 8. KBNIG, H., and ROSENWIRTH. B. (1988). Purification and partial characterization of poliovirus protease 2A by means of a functional assay. J. Virol. 62, 1243-l 250. L~EMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227,680-685. MORRIS-KRSINICH, B. A. M., FORSTER, R. L. S., and MOSSOP, D. W. (1983). The synthesis and processing of the nepovirus grapevine

GFLV

PROTEINASE

CLONING

fanleaf virus proteins in rabbit reticulocyte lysate. virology 130, 523-526. PALMENBERG, A., and RUECKEFIT, R. R. (1982). Evidencefor intramolecular self-cleavage of picornaviral replicase precursors. J. Viral. 41, 244-249. PENG, X. X., and SHIH, D. S. (1984). Proteolytic processing of the proteins translated from the bottom component RNA of cowpea mosaic virus. J. Ho/. Chem. 259, 3197-3201. PFAU, J., and YOUDERIAN, P. (1990). Transferring plasmid DNA between different bacterial species with electroporation. Nucleic Acids Res. 18, 6165. PINCK, L., FUCHS, M., PINCK, M., RAVELONANDRO, M., and WALTER, B. (1988). A satellite RNA in grapevine fanleaf virus strain F13. J. Gen. Virol. 69, 233-239. PINCK, M., REINBOLT, J., LOUDES, A. M., LE RET, M., and PINCK, L. (1991). Primary structure and location of the genome-linked protein (VPg) of grapevine fanleaf nepovirus. FEBS Lett 284, 117119. POWERS, J. C., and HARPER, J. W. (1986). Inhibitors of serine proteinases. In “Proteinases inhibitors” (A. J. Barrett and G. Salvesen, Eds.), pp. 55-l 52. Elsevier, New York.

AND

CHARACTERIZATION

787

RICH, D. H. (1986). Inhibitors of cysteine proteinases. ln “Proteinases inhibitors” (A. J. Barrett and G. Salvesen, Eds.), pp. 153-l 78. Elsevier, New York. RITZENTHALER, C., VIRY, M., PINCK, M., MARGIS, R., FUCHS, M., and PINCK, L. (1991). Complete nucleotide sequence and genetic organization of grapevine fanleaf nepovirus. J. Gen. Viral., in press. SERGHINI, M. A., FUCHS, M., PINCK, M., REINBOLT, J., WALTER, B., and PINCK, L. (1990). RNA2 of grapevine fanleaf virus: sequence analysis and coat protein cistron location. J. Gen. Viral. 71, 1433-l 44 1. SHIH, D. S., Bu, M., PRICE, M. A., and SHIH, C. Y. T. (1987). Inhibition of cleavage of a plant viral polyprotein by an inhibitor activity present in wheat germ and cowpea embryos. J. Virol. 61,912-915. TABOR, S., and RICHARDSON, C. C. (1987). DNA sequence analysis with a modified bacteriophage T7 DNA polymerase. Proc. Nat/. Acad. Sci. USA 84, 4767-4771. VERVER, J., GOLDBACH, R., GARCIA, J. A., and Vos, P. (1987). In vitro expression of a full-length DNA copy of cowpea mosaic virus B RNA: Identification of the B RNA encoded 24.kd protein as a viral protease. fMB0 J. 6, 549-554. WELLINK, J., and VAN KAMMEN, A. (1988). Proteases involved in the processing of viral polyproteins. Arch. Viral. 98, l-26.