- Email: [email protected]
Comparison of the Nucleic Acid- and NTP-Binding Properties of the Movement Protein of Cucumber Mosaic Cucumovirus and Tobacco Mosaic Tobamovirus
VIROLOGY
216, 71–79 (1996) 0035
ARTICLE NO.
Comparison of the Nucleic Acid- and NTP-Binding Properties of the Movement Protein of Cucumber Mosaic Cucumovirus and Tobacco Mosaic Tobamovirus QIUBO LI and PETER PALUKAITIS1 Department of Plant Pathology, Cornell University, Ithaca, New York 14853 Received September 11, 1995; accepted November 17, 1995 The cucumber mosaic virus (CMV) 3a movement protein (MP) was compared directly to the well-characterized tobacco mosaic virus (TMV) 30K MP by cloning the genes encoding these proteins into Escherichia coli, isolating the E. coli-expressed MPs, and characterizing them with regard to RNA- and NTP-binding activities. The two MPs were shown to bind singlestranded RNA and DNA cooperatively, but with no sequence specificity. However, discrete lengths of CMV RNA 3 could be protected against RNase digestion by the CMV 3a protein, indicating that the RNA was not uniformly covered by the MP after cooperative binding. The TMV 30K:RNA complex was more stable in NaCl than the CMV 3a:RNA complex; about 50% of the corresponding complexes were stable in 0.6 and 0.4 M NaCl, respectively. Both MPs could bind GTP strongly and UTP weakly, but not ATP or CTP. The CMV 3a protein expressed either in E. coli or in planta from RNA 3 of CMV was tagged at its C-terminus with six histidine residues, which facilitated its purification by affinity chromatography on a matrix containing Ni2/-nitrilotriacetate. The soluble, His-tagged 3a proteins, affinity-purified from E. coli and zucchini squash, both were able bind CMV RNA 3 in vitro. q 1996 Academic Press, Inc.
case, the MP was found to bind ATP and CTP and have ATPase activity (Rouleau et al., 1994). Cucumber mosaic virus (CMV) is an isometric plant virus with a tripartite RNA genome (review in Palukaitis et al., 1992). RNAs 1 and 2 are required for replication while RNA 3 encodes two proteins, the 30K 3a (movement) protein and 24.5K 3b (capsid) protein, both of which are involved in virus movement. The 3a protein is generally regarded as the MP of CMV because: (i) it is a nonstructural protein not required for replication, but required for infection of plants (Suzuki et al., 1991); (ii) its function can be complemented in transgenic plants expressing the 3a gene (Kaplan et al., 1995); (iii) it can traffic itself as well as RNA from cell to cell (Ding et al., 1995); and (iv) it contains sequence similarities to putative MPs of other viruses (Davies and Symons, 1988; Melcher, 1990; Mushegian and Koonin, 1993). Although the CMV 3a protein has the above characteristics shared by the tobacco mosaic virus (TMV) 30K protein and other virus MPs, the CMV 3a MP is related only remotely to the well-characterized TMV 30K MP, based on sequence alignment analysis. However, the CMV 3a protein also shows some clear differences from other virus MPs. For example, unlike most other virus MPs, which are usually localized in the cell wall of the infected plant, the CMV 3a protein was detected predominantly in the soluble cytoplasmic fraction of the infected plant (Kaplan et al., 1995). Furthermore, CMV requires the coat protein along with the 3a protein for viral cell-to-cell movement, in contrast to TMV, where only the 30K protein is needed for viral cell-to-cell movement. On the other hand, in contrast
INTRODUCTION The movement proteins (MPs) encoded by plant viruses play a pivotal role in plant virus movement and therefore have received considerable attention in recent studies (reviewed by Hull, 1989; Maule, 1991; McLean et al., 1993). The characterization of the MPs of several plant viruses in different genera has been carried out, although to different extents. In all cases the proteins examined were found to be essential for virus movement but not for replication in single cells (protoplasts) (Nishiguchi et al., 1978; Meshi et al., 1987; Osman et al., 1991; Suzuki et al., 1991; Thomas et al., 1993). Complementation has been observed in three cases where the virus carrying a dysfunctional MP gene could systemically infect a transgenic plant expressing the homologous MP (Deom et al., 1987; Giesman-Cookmeyer et al., 1995; Kaplan et al., 1995). Some of these MPs have been expressed in vitro (e.g., in Escherichia coli and in Saccharomyces cerevisiae) and the isolated MPs showed the ability to bind single-stranded (ss) nucleic acids, although with no sequence specificity (Citovsky et al., 1990, 1991; Osman et al., 1992; Schoumacher et al., 1992; Pascal et al., 1994; Thomas and Maule, 1995). Several of the E. coli-expressed MPs were also found to be able to traffic themselves as well as viral RNAs from one cell to another (Fujiwara et al., 1993; Noueiry et al., 1994; Waigmann et al., 1994; Lucas and Gilbertson, 1994). Finally, in one
1
To whom reprint requests should be addressed.
0042-6822/96 $12.00
71
/ 6a0f$$7678
01-12-96 08:20:51
vira
Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AP-Virology
72
LI AND PALUKAITIS
to some other viruses that require the coat protein for movement (van Lent et al., 1990; Perbal et al., 1993), CMV does not induce tubular structures protruding from the cell membrane during infection (Ding et al., 1995). Thus, some aspects of the mode of cell-to-cell movement of CMV may be different from that of TMV, cowpea mosaic virus, or cauliflower mosaic virus (CaMV). To further characterize the CMV 3a MP and to compare it to the TMV 30K MP, both MPs were expressed in E. coli and the isolated proteins were compared for their characteristics to bind nucleic acid and NTPs. In addition, the CMV 3a protein was isolated from infected plant tissue by affinity chromatography and characterized. MATERIALS AND METHODS Expression and purification of the CMV 3a protein and the TMV 30K protein in E. coli The nucleotide sequence encoding the CMV 3a protein (Fny strain) and the TMV 30K protein (U1 strain) was amplified from cDNA clones via a polymerase chain reaction (PCR) protocol (Sambrook et al., 1989) that introduced an NdeI site and a BamHI site flanking the 5* end of the initiation codon and the 3* end of the stop codon, respectively. The cDNA clones used as templates for the PCR reactions were a TMV cDNA clone containing sequences from nucleotide 4903 to the 3* end in pUC129, which was generously provided by Dr. W. O. Dawson (University of Florida, Lake Alfred, FL), and pFny309, a CMV RNA 3 cDNA clone, described in Rizzo and Palukaitis (1990). After digestion with the above enzymes, the fragments containing the genes encoding the MPs were ligated into the expression vector pET11a (Novagen, Madison, WI) previously digested with the same two enzymes. The ligated DNAs were transformed into E. coli strain BL21(DE3) and overexpressed as indicated by the supplier (Novagen). The expressed proteins were purified as described by Citovsky et al. (1990) except that sonication (on ice; 2 min) was used to facilitate lysis of the cells and breakage of the bacterial chromosomal DNA, the inclusions were washed with lysis buffer containing 2% Triton X-100, and 8 M urea was used to solubilize the protein without incubation at 567. For refolding, the solubilized MPs were dialyzed sequentially against buffer L of Citovsky et al. (1990) containing 4 M urea, 1 M urea, and no urea, and the proteins were stored in buffer L at 47. In vitro transcription, gel electrophoresis retardation assay, and UV crosslinking Single-stranded TMV RNA and double-stranded (ds) CMV RNAs were isolated (Palukaitis et al., 1983), and total ss RNAs were purified from virus particles (Kaplan et al., 1995), both as previously described. 32P-labeled full-length Fny CMV RNA 3 was generated by in vitro
/ 6a0f$$7678
01-12-96 08:20:51
vira
transcription of pFny309 (Rizzo and Palukaitis, 1990) with T7 polymerase in the presence of 5 mCi [32P]UTP (400 Ci/mmol; Amersham) as described in the Promega protocols. The labeled transcript was mixed with purified MPs in 15 ml of binding buffer A [50 mM Tris–Cl, 1 mM EDTA (pH 7.0), 50 mM NaCl, 1 mM DTT, 1 mg/ml bovine serum albumin (BSA), and 10% glycerol]. After incubation on ice for 30 min, the mixture was either subjected to electrophoresis in 1% agarose gels in TAE buffer or irradiated with UV light (twice at 9000 mJ) in a Stratalinker (Stratagene), treated with RNase A, and analyzed by electrophoresis on 10 or 12% polyacrylamide gels containing SDS (SDS–PAGE). The gels were dried and autoradiographed. In nucleotide binding assays, 0.05 mCi of each a-32P nucleoside triphosphate (400 Ci/mmol; Amersham) was incubated with 0.25 mg of the expressed protein in 10 ml of binding buffer B (50 mM Tris–Cl, pH 8.0, 100 mM NaCl, 1 mM MgCl2 , 1 mM DTT, and 10% glycerol) at room temperature for 1.5 hr. The mixture was then irradiated under UV light (twice at 9000 mJ) and analyzed by SDS– PAGE. The gels were dried and autoradiographed. Nitrocellulose membrane filter binding assay Nitrocellulose membrane filter binding assays were done as described by Wong and Lohman (1993) except that the binding buffer was changed to binding buffer A (above) containing different concentrations of NaCl. After incubation of the 3a protein and labeled transcript in 15 ml of the buffer, the mixture was filtered through a 45mm nitrocellulose membrane using the (Schleicher & Schuell) Minifold aspiration device. The membrane was then washed three times with 1 ml buffer A, dried, autoradiographed, and counted for radioactivity retained using a liquid scintillation counter (Beckman). Tagging of the CMV 3a protein with six histidines and protein purification Six histidine codons were inserted into the CMV RNA 3 cDNA clone, pFny309, immediately before the stop codon of the 3a open reading frame (ORF) by site-directed mutagenesis, using a synthesized primer (5*-CTAATACGCACCAAAGTACTAGTGGTGATGGTGATGGTGAAGACCGTTAACCACCTG-3*; the bold A was a silent change to create an SpeI site for screening and the underlined nucleotides were introduced to generate the insertion of six histidines), resulting in the modified cDNA clone, pFny309-His. Infectious transcripts were obtained by in vitro transcription of pFny309-His, combined with transcripts of RNA 1 and RNA 2 and inoculated to tobacco or zucchini squash plants (Shintaku et al., 1992). Fresh tissues were harvested 8 days postinoculation and ground in binding buffer C (50 mM sodium phosphate, pH 8.0, 300 mM NaCl, 10 mM 2-mercaptoethanol, 5 mM imidazole; 1 ml buffer/g, tissue). The extract was centri-
AP-Virology
CMV AND TMV MOVEMENT PROTEINS
fuged at 12,000 g for 20 min and the supernatant was applied to a 1-ml column of Ni2/-nitrilotriacetic acid (Ni2/NTA; QIAGEN) preequilibrated with binding buffer C. After washing the column with 10 ml binding buffer C and 10 ml elution buffer D (50 mM sodium phosphate, pH 6.0, 10 mM imidazole, and 10% glycerol), the Histagged 3a protein was eluted from the column with elution buffer D containing various concentrations of imidazole up to 250 mM; 3 1 1-ml fractions were collected for each imidazole concentration. The His-tagged 3a gene from pFny309-His was subcloned into the E. coli expression vector pET11a as above, and the His-tagged 3a protein was expressed and purified from bacteria according to the protocol provided by the supplier of the Ni2/-NTA matrix (QIAGEN). Analysis of the His-tagged 3a proteins was by SDS–PAGE and either silver staining using a kit (Bio-Rad) or electroblotting the proteins in the gels to nitrocellulose membranes and immunoprobing the blots with rabbit antiserum to the CMV 3a protein (Kaplan et al., 1995), alkaline phosphatase-conjugated goat anti-rabbit (secondary) antibody, and chromogenic reagents for development, as described in Gal-On et al. (1994).
RESULTS Overexpression and purification of the CMV 3a and TMV 30K proteins in E. coli The genes encoding the CMV 3a protein and the TMV 30K protein were subcloned into the E. coli expression vector pET11a, linked to the strong bacteriophage T7 promoter, producing pETMP3a and pETMP30K, respectively. A low-protease E. coli strain, BL21(DE3) containing a lac promoter-controlled T7 polymerase gene was used as the host for protein expression. Both MPs were expressed to a high level after induction of T7 polymerase expression in the host bacterial cells by adding IPTG (Fig. 1A, lanes 2 and 7). These high levels of expression resulted in the formation of insoluble aggregates of the proteins (inclusion bodies; Fig. 1A, lanes 4 and 9), which were separated readily from the bulk of the soluble bacterial proteins (Fig. 1A, lanes 3 and 8) by centrifugation. Soluble bacterial proteins were further removed by washing the inclusion bodies with lysis buffer containing 2% Triton X-100. The MPs were solubilized in a buffer containing 8 M urea and most of the preparation remained soluble after removal of the urea by dialysis. The final purified protein preparations contained very few detectable E. coli proteins (Fig. 1A, lanes 5 and 10). The expressed 3a and 30K proteins showed the expected sizes on SDS–polyacrylamide gels. Under the same expression conditions, bacteria carrying only the vector pET11a did not express any exogenous protein with a size similar to that of either the 3a or the 30K protein (not shown).
/ 6a0f$$7678
01-12-96 08:20:51
vira
73
Nucleic acid binding characteristics of the CMV 3a and TMV 30K MPs The TMV 30K protein has been shown to bind ss nucleic acids (Citovsky et al., 1990). To examine whether the CMV 3a protein also has the ability to bind ss RNA, 32 P-labeled full-length RNA 3 of Fny-CMV derived from in vitro transcription of the full-length cDNA clone was incubated with the purified 3a protein and analyzed by electrophoresis on nondenaturing agarose/TAE gels. As shown in Fig. 1B, at low protein/RNA (w/w) ratios, most of the labeled RNA migrated as free RNA. However, at higher protein/RNA ratios (ú50), the mobility of the RNA was drastically retarded. The RNA retardation pattern of the CMV 3a protein was similar to that of the TMV 30K protein (Fig. 1B). The rapid mobility transition from free RNA to completely retarded RNA as the protein concentration increases indicates that either the binding of both proteins to the RNA is cooperative or the binding occurred between the RNA and aggregated protein. To further analyze the RNA binding properties, the 3a protein was preincubated with proteinase K (Fig. 2A, lane 1), heated at 1007 for 10 min (lane 2), or treated with 0.1% SDS (lane 3) prior to incubation with labeled RNA transcript (lane 5). All three treatments either abolished or greatly reduced the ability of the 3a protein to bind ss RNA when compared to the nontreated 3a protein (Fig. 2A, lane 4). The remaining RNA binding after proteinase K treatment probably was due to incomplete digestion of the 3a protein. Digesting the 3a protein:RNA complexes with RNase A gave rise to a protein:RNA core (Fig. 2A, lanes 7–9) which migrated faster than the free RNA transcript (lane 5) and showed only a slight difference in mobility when digested with different amounts of RNase A, indicating that the 3a protein binds to discrete lengths of the RNA and not as an aggregate of protein to some of the RNA. No such RNase-resistant cores were observed in the absence of the 3a protein (Fig. 2A, lane 6). These data suggest that cooperative binding occurs between the 3a protein and the viral RNA. To determine whether the 3a protein has any sequence specificity or preference in binding RNA, competition binding assays were performed by incubating unlabeled competitor nucleic acids of various types and 32Plabeled CMV RNA 3 together with the CMV 3a protein, followed by analysis of the incubation mixture on a nondenaturing agarose gel. The results shown in Fig. 2B demonstrated that the heterologous ss viral RNA (lanes 1 and 2, TMV RNA), homologous ss viral RNA (lanes 3 and 4, CMV RNAs), and heterologous ss viral DNA (lanes 7 and 8, M13 phage DNA) were all able to compete effectively with the labeled transcript in binding to the 3a protein. At the higher competitor concentration (0.1 mg vs 0.025 mg/15 ml), more labeled RNA showed a shift from the position of completely retarded RNA to the position of free RNA (Fig. 2B, lanes 1 vs 2 and 7 vs 8). However,
AP-Virology
74
LI AND PALUKAITIS
FIG. 1. Expression, purification, and analysis of the CMV and TMV movement proteins. (A) SDS–PAGE of the expressed CMV 3a (lanes 1–5) and TMV 30K (lanes 6–10) proteins extracted from E. coli. The total bacterial lysate was analyzed before (lanes 1 and 6) and after (lanes 2 and 7) induction of expression. The induced lysate was centrifuged to remove the overexpressed protein, present in the pellet as an inclusion (lanes 4 and 9) from the soluble bacterial proteins (lanes 3 and 8). The inclusions were solubilized in 8 M urea, centrifuged to remove insoluble materials, and dialyzed to remove the urea (lanes 5 and 10). The gel was stained with Coomassie blue. The positions of the molecular weight markers are indicated. The arrow indicates the position of the movement proteins. (B) Retardation gel electrophoresis assay for RNA binding by the movement proteins purified from E. coli. Increasing amounts of either the TMV 30K protein (lanes 1–6) or the CMV 3a protein (lanes 7–11) were incubated in 15 ml of binding buffer A with 1 ng of 32P-labeled CMV RNA 3 transcript, and the mixture was electrophoresed in a 1% nondenaturing agarose gel. The amount of movement proteins used in the assays is indicated.
neither ds CMV RNAs (Fig. 2B, lanes 9 and 10) nor sheared yeast RNA fragments (õ50 nt; Fig. 2B, lanes 5 and 6) were effective competitors, although the ds RNAs did show a weak competition. Thus, the CMV 3a protein binds specifically to ss nucleic acid without obvious sequence preference. Verification of the ability of the CMV 3a protein and not just contaminating higher molecular weight E. coli proteins to bind ss RNA (Thomas and Maule, 1995) was demonstrated by UV light crosslinking experiments. The 32 P-labeled RNA was incubated with the 3a protein and
the incubation mixture was irradiated with UV light. This treatment can covalently crosslink protein:RNA complexes that are in close molecular association. Unbound RNA was removed by RNase A digestion, and the mixture was analyzed by SDS–PAGE and autoradiography. As shown in Fig. 3 (lanes 1–4), a strong band could be seen in the position expected for the 3a protein (arrow). Samples containing less RNase showed a band with a broader and slower mobility (Fig. 3, lane 4 vs lane 1). When no RNase was added, this band disappeared into a smeared background (Fig. 3, lane 5), with most of the
FIG. 2. Determination of the in vitro binding parameters of the CMV 3a movement protein. (A) RNA binding characteristics of the CMV 3a protein analyzed by retardation gel electrophoresis. The CMV 3a protein (250 ng in 15 ml) was pretreated with proteinase K (1 mg at 377 for 30 min; lane 1), 0.1% SDS (lane 2), or heat (1007, 10 min; lane 3) before incubation with 5 ng of 32P-labeled CMV RNA 3 transcript. The 3a protein (250 ng in 15 ml) was incubated with 32P-labeled RNA (5 ng) and either not treated further (lane 4) or incubated with different amounts of RNase A (room temperature, 15 min): 1 mg (lane 7), 250 ng (lane 8), or 60 ng (lane 9). The above samples and protein-free 32P-labeled RNA (5 ng), either incubated (lane 6) or not incubated (lane 5) with 60 ng of RNase A, were resolved on a 1% nondenaturing agarose gel. (B) Competition binding assay with the CMV 3a protein. The CMV 3a protein (250 ng in 15 ml) was incubated with 5 ng of labeled CMV RNA 3 transcript either in the absence of (lane 11) or in the presence of nonlabeled competitors (lanes 1–10): TMV RNA (lanes 1 and 2); CMV RNAs (lanes 3 and 4); sheared, yeast RNA (lanes 5 and 6); M13 ss DNA (lanes 7 and 8); or ds CMV RNAs (lanes 9 and 10). Two levels of competitor were used: 1 mg (lanes 1, 3, 5, 7, and 9) or 250 ng (lanes 2, 4, 6, 8, and 10). The labeled RNA was also incubated with buffer alone (lane 12).
/ 6a0f$$7678
01-12-96 08:20:51
vira
AP-Virology
CMV AND TMV MOVEMENT PROTEINS
FIG. 3. UV crosslinking of the movement proteins with labeled RNA. The CMV 3a protein was incubated with labeled RNA (10 ng) for 30 min on ice. After UV irradiation, the mixture was either untreated (lane 5) or incubated with different amounts of RNase A (lanes 1–4, 6–9): 2 mg (lane 1), 500 ng (lane 2), 130 ng (lane 3), or 30 ng (lanes 4, 6–9). Incubations were also carried out in the presence of either excess nonlabeled competitor ss CMV RNAs (1 mg; lane 7) or ds CMV RNAs (1 mg; lane 8), or either BSA (lane 6) or the TMV 30K protein (10 ng; lane 9) instead of the CMV 3a protein. The samples were analyzed by SDS–PAGE and autoradiography. The arrow indicates the position of the 3a protein.
labeled RNA at the top of the gel. When the incubation was carried out in the presence of excess unlabeled CMV RNA (Fig. 3, lane 7), when the UV irradiation was omitted (not shown), or when BSA was substituted for the 3a protein (lane 6), no band was observed. Moreover, ds RNA could not compete with the binding of labeled RNA to the 3a protein (Fig. 3, lane 8). The TMV 30K protein also was UV-crosslinked by the labeled RNA and gave a band of the expected size (Fig. 3, lane 9). Stability of MP:RNA complexes at different salt concentrations Stability of protein:RNA complexes with respect to salt concentration is often a criterion used to evaluate the strength of a protein:RNA association. The complexes formed between the 32P-labeled RNA transcript and the 3a or 30K proteins were incubated in buffers with different concentrations of NaCl, and the incubation mixtures were filtered through a nitrocellulose membrane and washed to remove unbound RNA, leaving only protein and protein:RNA complexes on the membrane. This method is effective in detecting small amounts of protein:RNA complexes which otherwise would not been observed by gel retardation. As shown in Fig. 4, both MPs bind RNA maximally between 50 and 200 mM NaCl, and at 400 mM NaCl both MPs showed a significant decrease in RNA binding. At 600 mM NaCl, the 3a protein could no longer bind RNA, while the 30K protein still retained about half of its RNA binding capability (vs at 200 mM NaCl) (Fig. 4). At higher salt concentrations, the 30K protein was also dissociated from the RNA. Nucleotide binding specificity of the CMV and TMV MPs Sequence analyses of the tobamovirus MPs indicated that they contain sequence motifs of possible NTP bind-
/ 6a0f$$7678
01-12-96 08:20:51
vira
75
FIG. 4. Salt stabilities of RNA:movement protein complexes. The TMV 30K and the CMV 3a proteins (250 ng) were incubated on ice with labeled RNA (1 ng) in the presence of the indicated concentrations of NaCl. After incubation, the mixtures were analyzed by nitrocellulose membrane filter binding. RNA binding was quantified by determining the radioactivity on the membrane by liquid scintillation counting.
ing domains (Saito et al., 1988). Thus, both the TMV 30K protein and the CMV 3a protein were tested here for their ability to bind to any of the four NTPs. Both MPs were incubated with each of the four a-32P-labeled NTPs, followed by UV irradiation to crosslink covalently any bound NTP, and were analyzed by SDS–PAGE. As shown in Fig. 5, both MPs were clearly labeled by GTP and weakly labeled by UTP, but were not labeled by either ATP or CTP. In some but not all experiments in which BSA was substituted for the MPs, a very weak binding to GTP was observed (not shown), similar in intensity to that observed between the TMV 30K protein and UTP (Fig. 5, lane 4). When an excess of unlabeled GTP (25 mM) was incubated along with 32P-labeled GTP (12.5 nM) and the 3a protein prior to UV irradiation, very little 3a protein became labeled (Fig. 5, lane 5), while the presence of
FIG. 5. Nucleotide binding by the movement proteins. The TMV 30K and the CMV 3a proteins (250 ng in 10 ml) were incubated with each of the four [a-32P]NTPs: ATP (lane 1), CTP (lane 2), GTP (lanes 3 and 7), and UTP (lane 4), at 12.5 nM. The incubation mixtures were irradiated under UV light and analyzed by SDS–PAGE. Incubation and irradiation of the CMV 3a protein was also carried out with [a-32P]GTP in the presence of the nonlabeled competitors 25 mM GTP (lane 5) and 25 mM ATP (lane 6). The samples were analyzed by SDS–PAGE and autoradiography. The arrow indicates the position of the 3a protein.
AP-Virology
76
LI AND PALUKAITIS
unlabeled ATP (25 mM) in the incubation mixture with the 3a protein and labeled GTP only slightly reduced the binding of 3a protein to GTP (Fig. 5, lane 6). The residual labeling seen in Fig. 5, lane 5, could be due to either a high Km for GTP concentration or the high concentration of protein (0.8 mM) in the assay. A similar competition assay was not done with labeled UTP or using the TMV 30K protein. These data indicate that the 3a and 30K proteins are GTP (and possibly UTP)-binding proteins. The higher molecular weight bands present in Fig. 5, lane 3, may be crosslinked aggregates of the respective MPs or contaminating E. coli proteins. Bands of similar mobility were also observed when RNA was bound to the MPs by UV crosslinking (Fig. 3). Tagging of the 3a protein with six histidines An in-frame insertion of six histidine (His) codons was introduced into the 3* terminus of the 3a ORF of pFny309 by site-directed mutagenesis, resulting in addition of six His residues to the C-terminus of the 3a protein. The presence of the introduced six His codons in pFny309-His, the modified full-length Fny-CMV RNA 3 cDNA clone, was verified by sequencing (data not shown). Transcript of RNA 3 obtained by in vitro transcription of pFny309-His, along with transcripts of Fny-CMV RNAs 1 and 2, was inoculated onto tobacco plants. Five days postinoculation, the tobacco plants showed green mosaic symptoms indistinguishable from those induced by wild-type Fny-CMV. Virus was extracted from the infected plants 12 days postinoculation and the yield was equivalent to wild-type Fny-CMV. The six-His-codon tag was retained in the viral RNA during two passages in tobacco plants and no other changes were detected by RNA sequencing (data not shown). The Fny-CMV encoding the His-tagged 3a protein was also able to infect zucchini squash with infectivity kinetics similar to those of the unmodified virus. Isolation of His-tagged 3a protein from both E. coli and plants The His codon-tagged 3a gene was ligated into the vector pET11a and cloned into and overexpressed in E. coli, and the His-tagged 3a protein was purified by Ni2/NTA-affinity chromatography. The protein purified from the E. coli cells transformed with pETMP3a-His contained a single band of the expected size less contaminated by bacterial protein than the conventionally purified nontagged 3a protein (Fig. 6A, lane 5, and data not shown). Under the same conditions, no protein was obtained from E. coli cells transformed with pETMP3a, in which the 3a gene was not tagged (data not shown). The isolated protein samples from infected plants were far less homogeneous. As shown in Fig. 6A, a band with the molecular weight expected for the 3a protein (arrow) appeared in fractions eluting at 50 mM (lane 2) and higher molarity (lanes 3 and 4) imidazole. Several
/ 6a0f$$7678
01-12-96 08:20:51
vira
other bands with mobilities less than and greater than that of the 3a protein also appeared in fractions eluting with 50 and 100 mM imidazole (Fig. 6A, lanes 2 and 3). In the fraction eluting in 250 mM imidazole (Fig. 6A, lane 4), only one band with a mobility similar to that of the E. coli-expressed 3a protein (Fig. 6A, lane 5) was observed. This band and several higher molecular weight bands reacted very strongly with the 3a antiserum in an immunoblot assay (Fig. 6B, lanes 1–5), indicating their relationship to the 3a protein. The higher molecular weight forms of the 3a protein may be the aggregates previously described in gels showing RNAs (Fig. 3) or NTPs (Fig. 5) crosslinked to the CMV 3a protein expressed in E. coli. None of the bands described above were detected when extracts of noninfected plants were applied to the Ni2/NTA column and eluted with imidazole (data not shown). RNA-binding activity of the His-tagged 3a protein Using Ni2/-NTA chromatography, the His-tagged 3a protein expressed in E. coli was purified to homogeneity as demonstrated by silver staining (Fig. 6A, lane 5). The His-tagged 3a protein was then examined for RNA binding ability by gel retardation (not shown) and UV-crosslinking assays (Fig. 6C). The His-tagged 3a protein was able to bind ss RNA cooperatively (not shown) with a pattern similar to that of the nontagged 3a protein (Fig. 1B). In UV-crosslinking assays, the His-tagged 3a protein purified from E. coli was crosslinked to the labeled RNA (Fig. 6C, lane 5) and was of the same size as the crosslinked nontagged 3a protein (Fig. 6C, lane 3a). An additional band with about twice the molecular weight of the 3a protein was also crosslinked to the labeled RNA. Since the His-tagged 3a sample purified from E. coli was devoid of any detectable bacterial contaminant proteins (Fig. 6A, lane 5), this additional band is probably a dimer of the 3a protein. The intense bands migrating faster than the 3a protein (Fig. 6C, lane 5) may be degradation products of the His-tagged 3a protein. The His-tagged 3a protein sample isolated from infected squash and eluting from the Ni2/-NTA column in 250 mM imidazole also was analyzed in a similar UV-crosslinking assay for the ability to bind RNA. As shown in Fig. 6C, a weak band with a slightly faster migration than the E. coliexpressed 3a protein was crosslinked to the RNA (lanes 4, arrow). The data shown in Fig. 6 were obtained using extracts of CMV-infected squash plants. These contained a higher concentration of MP than extracts of CMV-infected tobacco (data not shown). Nevertheless, the Histagged 3a protein isolated from squash was not tested for either cooperative RNA binding or NTP binding, because of the low levels of protein present in these assays compared to those present in previous ones (Figs. 1–5). DISCUSSION The results presented here show that the CMV (3a) MP binds ss RNA in a cooperative manner similar to that
AP-Virology
CMV AND TMV MOVEMENT PROTEINS
77
FIG. 6. Isolation and characterization of the His-tagged 3a protein from both E. coli and CMV-infected squash leaves. The squash leaf extract was subjected to affinity chromatography on a Ni2/-NTA column. After extensive washing with buffer containing 10 mM imidazole, the proteins were eluted from the column sequentially in buffer containing imidazole at 25 mM (lane 1), 50 mM (lane 2), 100 mM (lane 3), and 250 mM (lane 4). The last wash fraction (lane 0), the elution fractions (lanes 1–4), and the His-tagged 3a protein isolated from E. coli and purified by Ni2/-NTA-affinity chromatography (lane 5) were analyzed by SDS–PAGE and either (A) stained with silver reagents or (B) blotted to a nitrocellulose membrane and immunoprobed with an antiserum to the 3a protein. In C, samples corresponding to those seen in lanes 4 and 5 of A and B as well as nontagged 3a protein expressed in and purified from E. coli by the conventional method were incubated with 32P-CMV RNA 3, UV-irradiated, and incubated with RNase A (1 mg) as in Fig. 3 prior to analysis by SDS–PAGE and autoradiography. The arrows indicate the position of the 3a protein.
of the TMV (30K) MP. No sequence specificity was found in the RNA binding by the 3a protein. The 3a protein cannot bind ds RNA or very small fragments of ss RNA (e.g., sheared yeast RNA fragments), indicating that the binding requires some minimum length of nucleic acid. In these respects, the CMV MP is similar to the other viral MPs studied previously. The observation that discrete lengths of CMV RNA 3, but not the entire RNA molecule, were protected against digestion with RNase indicated that the entire RNA 3 is not coated with MP as a result of the cooperative binding. There are no extensive sequence similarities between the MPs of different virus families. Results from experiments to identify the sequences in the MPs of TMV (Citovsky et al., 1992), alfalfa mosaic virus (AlMV; Schoumacher et al., 1992), CaMV (Thomas and Maule, 1995), and red clover necrotic mosaic virus (RCNMV; Osman et al., 1993; Giesman-Cookmeyer and Lommel, 1993) responsible for RNA binding showed that no consensus RNA-binding domain or motif exists, although most MPs are found to be rich in basic amino acids. It may be questioned whether the 3-D structure itself of the MP is important for its RNA-binding activity or whether the
/ 6a0f$$7678
01-12-96 08:20:51
vira
binding is just a result of electrostatic attractions between the positively charged sidechains of the MP and the negatively charged backbone of the RNA. When the 3a protein was heat denatured, which presumably would cause more charged amino acid sidechains to become exposed, it lost the ability to bind ss RNA. We also observed that complete refolding of the protein by extensive dialysis was crucial for it to achieve full capacity of RNA binding (data not shown). These data suggest that the binding of the 3a protein is structure-dependent. Although the CMV 3a protein showed an RNA binding pattern similar to that of the TMV 30K protein, the protein:RNA complex formed by the CMV 3a protein is dissociated more readily by NaCl than the complex formed by the TMV 30K protein. Different virus MPs have been reported to have different degrees of salt stability in binding to ss RNA. The AlMV 3a protein (Schoumacher et al., 1992) and the CaMV P1 (Citovsky et al., 1991) have been shown to have significantly decreased RNA binding activity at 0.2 M NaCl. The TMV 30K protein was reported to be able to withstand 0.6 M NaCl (Citovsky et al., 1990), similar to the result obtained here. Binding of the MPs of RCNMV (Osman et al., 1992) and foxtail mosaic potex-
AP-Virology
78
LI AND PALUKAITIS
virus (FoMV) (Rouleau et al., 1994) to RNA is stable at 0.4 M NaCl, similar to what is shown here for the CMV 3a protein. Both the TMV 30K MP and the CMV 3a MP showed the ability to bind GTP. There is a predicted NTP-binding sequence in both proteins, although this same region containing a stretch of hydrophobic residues preceding a conserved Asp was proposed in another sequence alignment study to be important for membrane association of the MPs (Mushegian and Koonin, 1993). However, one of the triple gene block proteins (26K) of potexviruses contains a predicted sequence motif conserved in a wide range of NTP-binding and hydrolysis proteins (Gorbalenya and Koonin, 1989). Recently, the 26K protein of FoMV was expressed in E. coli and was found to bind both ATP and CTP and to have apparent ATPase activity (Rouleau et al., 1994). It was speculated that ATP hydrolysis by the 26K protein may provide the energy needed for transport of the virus from cell to cell. However, the TMV 30K or CMV 3a protein showed a specificity in nucleotide binding different from the FoMV 26K protein. It is very tempting in this case to speculate that these MPs may participate in different plant cell signaling pathways which lead to down-regulation of the gating capacity of plasmodesmata. With regard to the various biochemical properties examined, the CMV MP is most similar to the TMV MP. Both viruses also require particular sequences near, but not at, the C-terminus of the respective MPs for movement (Kaplan et al., 1995). However, the difference in subcellular distribution of the two MPs and the inability of the CMV 3a protein expressed transgenically to complement the movement of movement-defective TMV (Kaplan et al., 1995), while transgenically expressed TMV MP will complement the local movement of movementdefective CMV (I. B. Kaplan and P. Palukaitis, unpublished), indicate that there are similar aspects to the movement processes of these two viruses, but the corresponding MP functions are not identical. This is also consistent with the observation that other sequences in both viral genomes can influence viral movement (Nelson et al., 1993; Carr et al., 1994; Gal-On et al., 1994). The reported RNA binding activity of other viral MPs were all conducted with the proteins expressed in an in vitro system and usually the expressed proteins appeared to be insoluble. In most cases, the expressed proteins were solubilized in a buffer containing a denaturant (e.g., urea or SDS) and then dialyzed to remove the denaturant. In this study, we employed the powerful His-tagging system to facilitate purification of the 3a MP from E. coli as well as from CMV-infected plants. Tagging the 3a protein with six histidines did not affect RNA binding of the E. coli-expressed protein. Using Ni2/-NTA-affinity chromatography, the in planta-generated 3a protein was purified extensively. The His-tagged 3a protein samples isolated from the infected plants were also able to
/ 6a0f$$7678
01-12-96 08:20:51
vira
bind ss RNA. This is the first time that a virus movement protein has been purified from infected plants and shown to contain RNA binding activity. This system should allow purification of larger quantities of the native MP, as well as allow us to discern whether any of the proteins copurifying with the 3a protein (Fig. 6A) are involved in the movement process. ACKNOWLEDGMENT This work was supported by Grant DE-FG02-86ER13505 from the U.S. Department of Energy.
REFERENCES Bush, G. L., Tassin, A., Friden, H., and Meyer, D. I. (1991). Purification of a translocation-competent secretory protein precursor using nickel ion affinity chromatography. J. Biol. Chem. 266, 13811–13814. Carr, J. P., Gal-On, A., Palukaitis, P., and Zaitlin, M. (1994). Replicasemediated resistance to cucumber mosaic virus in transgenic plants involves suppression of both virus replication in the inoculated leaves and long-distance movement. Virology 199, 439–447. Citovsky, V., Knorr, D., Schuster, G., and Zambryski, P. (1990). The p30 movement protein of tobacco mosaic virus is a single-strand nucleic acid binding protein. Cell 60, 637–648. Citovsky, V., Knorr, D., Schuster, G., and Zambryski, P. (1991). Gene I, a potential cell-to-cell movement locus of cauliflower mosaic virus, encodes an RNA-binding protein. Proc. Natl. Acad. Sci. USA 88, 2476–2480. Citovsky, V., Wong, M. L., Shaw, A. L., Prasad, B. V. V., and Zambryski, P. (1992). Visualization and characterization of tobacco mosaic virus movement protein binding to single-stranded nucleic acids. Plant Cell 4, 397–411. Davies, C., and Symons, R. H. (1988). Further implications for the evolutionary relationships between tripartite plant viruses based on cucumber mosaic virus RNA 3. Virology 165, 216–244. Deom, C. M., Oliver, M. J., and Beachy, R. N. (1987). The 30-kilodalton gene product of tobacco mosaic virus potentiates virus movement. Science 237, 389–394. Ding, B., Li, Q., Nguyen, L., Palukaitis, P., and Lucas, W. J. (1995). Cucumber mosaic virus 3a protein potentiates cell-to-cell trafficking of CMV RNA in tobacco plants. Virology 207, 345–353. Fujiwara, T., Giesman-Cookmeyer, D., Ding, B., Lommel, S. A., and Lucas, W. J. (1993). Cell-to-cell trafficking of macromolecules through plasmodesmata potentiated by the red clover necrotic mosaic virus movement protein. Plant Cell 5, 1783–1794. Gal-On, A., Kaplan, I., Roossinck, M. J., and Palukaitis, P. (1994). The kinetics of infection of zucchini squash by cucumber mosaic virus indicate a role for RNA 1 in virus movement. Virology 205, 280–289. Giesman-Cookmeyer, D., and Lommel, S. A. (1993). Alanine scanning mutagenesis of a plant virus movement protein identifies three functional domains. Plant Cell 5, 973–982. Giesman-Cookmeyer, D., Silver, S., Vaewhongs, A. A., Lommel, S. A., and Deom, C. M. (1995). Tobamovirus and dianthovirus movement proteins are functionally homologous. Virology 213, 38–45. Gorbalenya, A. E., and Koonin, E. V. (1989). Viral proteins containing the purine NTP-binding sequence pattern. Nucleic Acids Res. 17, 8413–8440. Hull, R. (1989). The movement of viruses in plants. Annu. Rev. Phytopathol. 27, 213–340. Kaplan, I. B., Shintaku, M. H., Li, Q., Zhang, L., Marsh, L. E., and Palukaitis, P. (1995). Complementation of movement defective mutants in transgenic tobacco expressing the cucumber mosaic virus movement gene. Virology 209, 188–199. Lucas, W. J., and Gilbertson, R. L. (1994). Plasmodesmata in relation
AP-Virology
CMV AND TMV MOVEMENT PROTEINS to viral movement within leaf tissues. Annu. Rev. Phytopathol. 32, 387–411. Maule, A. J. (1991). Virus movement in infected plants. Crit. Rev. Plant Sci. 9, 457–474. McLean, B. G., Waigmann, E., Citovsky, V., and Zambryski, P. (1993). Cell-to-cell movement of plant viruses. Trends Microbiol. 1, 105–109. Melcher, U. (1990). Similarities between putative transport proteins of plant viruses. J. Gen. Virol. 71, 1009–1018. Meshi, T., Watanabe, Y., Saito, T., Sugimoto, A., Maeda, T., and Okada, Y. (1987). Function of the 30kd protein of tobacco mosaic virus: Involvement in cell-to-cell movement and dispensability for replication. EMBO J. 6, 2557–2563. Mushegian, A. R., and Koonin, E. V. (1993). Cell-to-cell movement of plant viruses: Insights from amino acid sequence comparisons of movement proteins and from analogies with cellular transport systems. Arch. Virol. 133, 239–257. Nelson, R. S., Guoxuan, L., Hodgson, R. A. J., Beachy, R. N., and Shintaku, M. H. (1993). Impeded phloem-dependent accumulation of the masked strain of tobacco mosaic virus. Mol. Plant–Microbe Interact. 6, 45–54. Nishiguchi, M., Motoyoshi, F., and Oshima, N. (1978). Behavior of a temperature sensitive strain of tobacco mosaic virus in tomato leaves and protoplasts. J. Gen. Virol. 39, 53–61. Noueiry, A. O., Lucas, W. J., and Gilbertson, R. L. (1994). Two proteins of a plant DNA virus coordinate nuclear and plasmodesmal transport. Cell 76, 925–932. Osman, T. A. M., Ingles, P. J., Miller, S. J., and Buck, K. W. (1991). A spontaneous red clover necrotic mosaic virus mutant with a truncated movement protein. J. Gen. Virol. 72, 1793–1800. Osman, T. A. M., Hayes, R. J., and Buck, K. W. (1992). Cooperative binding of the red clover necrotic mosaic virus movement protein to single-stranded nucleic acids. J. Gen. Virol. 73, 223–227. Osman, T. A. M., Thommes, P., and Buck, K. W. (1993). Localization of a single-stranded RNA-binding domain in the movement protein of red clover necrotic mosaic dianthovirus. J. Gen. Virol. 74, 2453–2457. Palukaitis, P., Garcia-Arenal, F., Sulzinski, M. A., and Zaitlin, M. (1983). Replication of tobacco mosaic virus. VII. Further characterization of single- and double-stranded virus-related RNAs from TMV-infected tobacco plants. Virology 131, 533–545. Palukaitis, P., Roossinck, M. J., Dietzgen, R. G., and Francki, R. I. B. (1992). Cucumber mosaic virus. Adv. Virus Res. 41, 281–348. Pascal, E., Sanderfoot, A. A., Ward, B. M., Medville, R., Turgeon, R., and Lazarowitz, S. G. (1994). The geminivirus BR1 movement protein
/ 6a0f$$7678
01-12-96 08:20:51
vira
79
binds single-stranded DNA and localizes to the cell nucleus. Plant Cell 6, 995–1006. Perbal, M. C., Thomas, C. L., and Maule, A. J. (1993). Cauliflower mosaic virus gene I product P1 forms tubular structures which extend from the surface of infected protoplasts. Virology 195, 281–285. Rizzo, T. M., and Palukaitis, P. (1990). Construction of full-length cDNA clones of cucumber mosaic virus RNAs 1, 2, and 3: Generation of infectious RNA transcripts. Mol. Gen. Genet. 222, 249–256. Rouleau, M., Smith, R. J., Bancroft, J. B., and Mackie, G. A. (1994). Purification, properties, and subcellular localization of foxtail mosaic potexvirus 26-kDa protein. Virology 204, 254–265. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). ‘‘Molecular Cloning: A Laboratory Manual,’’ 2nd ed. Cold Spring Harbor Laboratories, Cold Spring Harbor, NY. Saito, T., Imai, Y., Meshi, T., and Okada, Y. (1988). Interviral homologies of the 30K proteins of tobamoviruses. Virology 176, 653–656. Schoumacher, F., Erny, C., Berna, A., Godefroy-Colburn, T., and StussiGaraud, C. (1992). Nucleic acid-binding properties of the alfalfa mosaic virus movement protein produced in yeast. Virology 188, 896– 899. Shintaku, M. H., Zhang, L., and Palukaitis, P. (1992). A single amino acid substitution in the coat protein of cucumber mosaic virus induces chlorosis in tobacco. Plant Cell 4, 751–757. Suzuki, M., Kuwata, S., Kataoka, J., Masuta, C., Nitta, N., and Takanami, Y. (1991). Functional analysis of deletion mutants of cucumber mosaic virus RNA3 using an in vitro transcription system. Virology 183, 106–113. Thomas, C. L., and Maule, A. J. (1995). Identification of the cauliflower mosaic virus movement protein RNA-binding domain. Virology 206, 1145–1149. Thomas, C. L., Perbal, C., and Maule, A. J. (1993). A mutation of cauliflower mosaic virus gene I interferes with virus movement but not virus replication. Virology 192, 415–421. van Lent, J., Wellink, J., and Goldbach, R. (1990). Evidence for the involvement of the 58K and 48K proteins in the intercellular movement of cowpea mosaic virus. J. Gen. Virol. 71, 219–224. Waigmann, E., Lucas, W. J., Citovsky, V., and Zambryski, P. (1994). Direct functional assay for tobacco mosaic virus cell-to-cell movement protein and identification of a domain involved in increasing plasmodesmal permeability. Proc. Natl. Acad. Sci. USA 91, 1433–1437. Wong, I., and Lohman, T. M. (1993). A double-filter method for nitrocellulose-filter binding: Application to protein–nucleic acid interactions. Proc. Natl. Acad. Sci. USA 90, 5428–5432.
AP-Virology
216, 71–79 (1996) 0035
ARTICLE NO.
Comparison of the Nucleic Acid- and NTP-Binding Properties of the Movement Protein of Cucumber Mosaic Cucumovirus and Tobacco Mosaic Tobamovirus QIUBO LI and PETER PALUKAITIS1 Department of Plant Pathology, Cornell University, Ithaca, New York 14853 Received September 11, 1995; accepted November 17, 1995 The cucumber mosaic virus (CMV) 3a movement protein (MP) was compared directly to the well-characterized tobacco mosaic virus (TMV) 30K MP by cloning the genes encoding these proteins into Escherichia coli, isolating the E. coli-expressed MPs, and characterizing them with regard to RNA- and NTP-binding activities. The two MPs were shown to bind singlestranded RNA and DNA cooperatively, but with no sequence specificity. However, discrete lengths of CMV RNA 3 could be protected against RNase digestion by the CMV 3a protein, indicating that the RNA was not uniformly covered by the MP after cooperative binding. The TMV 30K:RNA complex was more stable in NaCl than the CMV 3a:RNA complex; about 50% of the corresponding complexes were stable in 0.6 and 0.4 M NaCl, respectively. Both MPs could bind GTP strongly and UTP weakly, but not ATP or CTP. The CMV 3a protein expressed either in E. coli or in planta from RNA 3 of CMV was tagged at its C-terminus with six histidine residues, which facilitated its purification by affinity chromatography on a matrix containing Ni2/-nitrilotriacetate. The soluble, His-tagged 3a proteins, affinity-purified from E. coli and zucchini squash, both were able bind CMV RNA 3 in vitro. q 1996 Academic Press, Inc.
case, the MP was found to bind ATP and CTP and have ATPase activity (Rouleau et al., 1994). Cucumber mosaic virus (CMV) is an isometric plant virus with a tripartite RNA genome (review in Palukaitis et al., 1992). RNAs 1 and 2 are required for replication while RNA 3 encodes two proteins, the 30K 3a (movement) protein and 24.5K 3b (capsid) protein, both of which are involved in virus movement. The 3a protein is generally regarded as the MP of CMV because: (i) it is a nonstructural protein not required for replication, but required for infection of plants (Suzuki et al., 1991); (ii) its function can be complemented in transgenic plants expressing the 3a gene (Kaplan et al., 1995); (iii) it can traffic itself as well as RNA from cell to cell (Ding et al., 1995); and (iv) it contains sequence similarities to putative MPs of other viruses (Davies and Symons, 1988; Melcher, 1990; Mushegian and Koonin, 1993). Although the CMV 3a protein has the above characteristics shared by the tobacco mosaic virus (TMV) 30K protein and other virus MPs, the CMV 3a MP is related only remotely to the well-characterized TMV 30K MP, based on sequence alignment analysis. However, the CMV 3a protein also shows some clear differences from other virus MPs. For example, unlike most other virus MPs, which are usually localized in the cell wall of the infected plant, the CMV 3a protein was detected predominantly in the soluble cytoplasmic fraction of the infected plant (Kaplan et al., 1995). Furthermore, CMV requires the coat protein along with the 3a protein for viral cell-to-cell movement, in contrast to TMV, where only the 30K protein is needed for viral cell-to-cell movement. On the other hand, in contrast
INTRODUCTION The movement proteins (MPs) encoded by plant viruses play a pivotal role in plant virus movement and therefore have received considerable attention in recent studies (reviewed by Hull, 1989; Maule, 1991; McLean et al., 1993). The characterization of the MPs of several plant viruses in different genera has been carried out, although to different extents. In all cases the proteins examined were found to be essential for virus movement but not for replication in single cells (protoplasts) (Nishiguchi et al., 1978; Meshi et al., 1987; Osman et al., 1991; Suzuki et al., 1991; Thomas et al., 1993). Complementation has been observed in three cases where the virus carrying a dysfunctional MP gene could systemically infect a transgenic plant expressing the homologous MP (Deom et al., 1987; Giesman-Cookmeyer et al., 1995; Kaplan et al., 1995). Some of these MPs have been expressed in vitro (e.g., in Escherichia coli and in Saccharomyces cerevisiae) and the isolated MPs showed the ability to bind single-stranded (ss) nucleic acids, although with no sequence specificity (Citovsky et al., 1990, 1991; Osman et al., 1992; Schoumacher et al., 1992; Pascal et al., 1994; Thomas and Maule, 1995). Several of the E. coli-expressed MPs were also found to be able to traffic themselves as well as viral RNAs from one cell to another (Fujiwara et al., 1993; Noueiry et al., 1994; Waigmann et al., 1994; Lucas and Gilbertson, 1994). Finally, in one
1
To whom reprint requests should be addressed.
0042-6822/96 $12.00
71
/ 6a0f$$7678
01-12-96 08:20:51
vira
Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AP-Virology
72
LI AND PALUKAITIS
to some other viruses that require the coat protein for movement (van Lent et al., 1990; Perbal et al., 1993), CMV does not induce tubular structures protruding from the cell membrane during infection (Ding et al., 1995). Thus, some aspects of the mode of cell-to-cell movement of CMV may be different from that of TMV, cowpea mosaic virus, or cauliflower mosaic virus (CaMV). To further characterize the CMV 3a MP and to compare it to the TMV 30K MP, both MPs were expressed in E. coli and the isolated proteins were compared for their characteristics to bind nucleic acid and NTPs. In addition, the CMV 3a protein was isolated from infected plant tissue by affinity chromatography and characterized. MATERIALS AND METHODS Expression and purification of the CMV 3a protein and the TMV 30K protein in E. coli The nucleotide sequence encoding the CMV 3a protein (Fny strain) and the TMV 30K protein (U1 strain) was amplified from cDNA clones via a polymerase chain reaction (PCR) protocol (Sambrook et al., 1989) that introduced an NdeI site and a BamHI site flanking the 5* end of the initiation codon and the 3* end of the stop codon, respectively. The cDNA clones used as templates for the PCR reactions were a TMV cDNA clone containing sequences from nucleotide 4903 to the 3* end in pUC129, which was generously provided by Dr. W. O. Dawson (University of Florida, Lake Alfred, FL), and pFny309, a CMV RNA 3 cDNA clone, described in Rizzo and Palukaitis (1990). After digestion with the above enzymes, the fragments containing the genes encoding the MPs were ligated into the expression vector pET11a (Novagen, Madison, WI) previously digested with the same two enzymes. The ligated DNAs were transformed into E. coli strain BL21(DE3) and overexpressed as indicated by the supplier (Novagen). The expressed proteins were purified as described by Citovsky et al. (1990) except that sonication (on ice; 2 min) was used to facilitate lysis of the cells and breakage of the bacterial chromosomal DNA, the inclusions were washed with lysis buffer containing 2% Triton X-100, and 8 M urea was used to solubilize the protein without incubation at 567. For refolding, the solubilized MPs were dialyzed sequentially against buffer L of Citovsky et al. (1990) containing 4 M urea, 1 M urea, and no urea, and the proteins were stored in buffer L at 47. In vitro transcription, gel electrophoresis retardation assay, and UV crosslinking Single-stranded TMV RNA and double-stranded (ds) CMV RNAs were isolated (Palukaitis et al., 1983), and total ss RNAs were purified from virus particles (Kaplan et al., 1995), both as previously described. 32P-labeled full-length Fny CMV RNA 3 was generated by in vitro
/ 6a0f$$7678
01-12-96 08:20:51
vira
transcription of pFny309 (Rizzo and Palukaitis, 1990) with T7 polymerase in the presence of 5 mCi [32P]UTP (400 Ci/mmol; Amersham) as described in the Promega protocols. The labeled transcript was mixed with purified MPs in 15 ml of binding buffer A [50 mM Tris–Cl, 1 mM EDTA (pH 7.0), 50 mM NaCl, 1 mM DTT, 1 mg/ml bovine serum albumin (BSA), and 10% glycerol]. After incubation on ice for 30 min, the mixture was either subjected to electrophoresis in 1% agarose gels in TAE buffer or irradiated with UV light (twice at 9000 mJ) in a Stratalinker (Stratagene), treated with RNase A, and analyzed by electrophoresis on 10 or 12% polyacrylamide gels containing SDS (SDS–PAGE). The gels were dried and autoradiographed. In nucleotide binding assays, 0.05 mCi of each a-32P nucleoside triphosphate (400 Ci/mmol; Amersham) was incubated with 0.25 mg of the expressed protein in 10 ml of binding buffer B (50 mM Tris–Cl, pH 8.0, 100 mM NaCl, 1 mM MgCl2 , 1 mM DTT, and 10% glycerol) at room temperature for 1.5 hr. The mixture was then irradiated under UV light (twice at 9000 mJ) and analyzed by SDS– PAGE. The gels were dried and autoradiographed. Nitrocellulose membrane filter binding assay Nitrocellulose membrane filter binding assays were done as described by Wong and Lohman (1993) except that the binding buffer was changed to binding buffer A (above) containing different concentrations of NaCl. After incubation of the 3a protein and labeled transcript in 15 ml of the buffer, the mixture was filtered through a 45mm nitrocellulose membrane using the (Schleicher & Schuell) Minifold aspiration device. The membrane was then washed three times with 1 ml buffer A, dried, autoradiographed, and counted for radioactivity retained using a liquid scintillation counter (Beckman). Tagging of the CMV 3a protein with six histidines and protein purification Six histidine codons were inserted into the CMV RNA 3 cDNA clone, pFny309, immediately before the stop codon of the 3a open reading frame (ORF) by site-directed mutagenesis, using a synthesized primer (5*-CTAATACGCACCAAAGTACTAGTGGTGATGGTGATGGTGAAGACCGTTAACCACCTG-3*; the bold A was a silent change to create an SpeI site for screening and the underlined nucleotides were introduced to generate the insertion of six histidines), resulting in the modified cDNA clone, pFny309-His. Infectious transcripts were obtained by in vitro transcription of pFny309-His, combined with transcripts of RNA 1 and RNA 2 and inoculated to tobacco or zucchini squash plants (Shintaku et al., 1992). Fresh tissues were harvested 8 days postinoculation and ground in binding buffer C (50 mM sodium phosphate, pH 8.0, 300 mM NaCl, 10 mM 2-mercaptoethanol, 5 mM imidazole; 1 ml buffer/g, tissue). The extract was centri-
AP-Virology
CMV AND TMV MOVEMENT PROTEINS
fuged at 12,000 g for 20 min and the supernatant was applied to a 1-ml column of Ni2/-nitrilotriacetic acid (Ni2/NTA; QIAGEN) preequilibrated with binding buffer C. After washing the column with 10 ml binding buffer C and 10 ml elution buffer D (50 mM sodium phosphate, pH 6.0, 10 mM imidazole, and 10% glycerol), the Histagged 3a protein was eluted from the column with elution buffer D containing various concentrations of imidazole up to 250 mM; 3 1 1-ml fractions were collected for each imidazole concentration. The His-tagged 3a gene from pFny309-His was subcloned into the E. coli expression vector pET11a as above, and the His-tagged 3a protein was expressed and purified from bacteria according to the protocol provided by the supplier of the Ni2/-NTA matrix (QIAGEN). Analysis of the His-tagged 3a proteins was by SDS–PAGE and either silver staining using a kit (Bio-Rad) or electroblotting the proteins in the gels to nitrocellulose membranes and immunoprobing the blots with rabbit antiserum to the CMV 3a protein (Kaplan et al., 1995), alkaline phosphatase-conjugated goat anti-rabbit (secondary) antibody, and chromogenic reagents for development, as described in Gal-On et al. (1994).
RESULTS Overexpression and purification of the CMV 3a and TMV 30K proteins in E. coli The genes encoding the CMV 3a protein and the TMV 30K protein were subcloned into the E. coli expression vector pET11a, linked to the strong bacteriophage T7 promoter, producing pETMP3a and pETMP30K, respectively. A low-protease E. coli strain, BL21(DE3) containing a lac promoter-controlled T7 polymerase gene was used as the host for protein expression. Both MPs were expressed to a high level after induction of T7 polymerase expression in the host bacterial cells by adding IPTG (Fig. 1A, lanes 2 and 7). These high levels of expression resulted in the formation of insoluble aggregates of the proteins (inclusion bodies; Fig. 1A, lanes 4 and 9), which were separated readily from the bulk of the soluble bacterial proteins (Fig. 1A, lanes 3 and 8) by centrifugation. Soluble bacterial proteins were further removed by washing the inclusion bodies with lysis buffer containing 2% Triton X-100. The MPs were solubilized in a buffer containing 8 M urea and most of the preparation remained soluble after removal of the urea by dialysis. The final purified protein preparations contained very few detectable E. coli proteins (Fig. 1A, lanes 5 and 10). The expressed 3a and 30K proteins showed the expected sizes on SDS–polyacrylamide gels. Under the same expression conditions, bacteria carrying only the vector pET11a did not express any exogenous protein with a size similar to that of either the 3a or the 30K protein (not shown).
/ 6a0f$$7678
01-12-96 08:20:51
vira
73
Nucleic acid binding characteristics of the CMV 3a and TMV 30K MPs The TMV 30K protein has been shown to bind ss nucleic acids (Citovsky et al., 1990). To examine whether the CMV 3a protein also has the ability to bind ss RNA, 32 P-labeled full-length RNA 3 of Fny-CMV derived from in vitro transcription of the full-length cDNA clone was incubated with the purified 3a protein and analyzed by electrophoresis on nondenaturing agarose/TAE gels. As shown in Fig. 1B, at low protein/RNA (w/w) ratios, most of the labeled RNA migrated as free RNA. However, at higher protein/RNA ratios (ú50), the mobility of the RNA was drastically retarded. The RNA retardation pattern of the CMV 3a protein was similar to that of the TMV 30K protein (Fig. 1B). The rapid mobility transition from free RNA to completely retarded RNA as the protein concentration increases indicates that either the binding of both proteins to the RNA is cooperative or the binding occurred between the RNA and aggregated protein. To further analyze the RNA binding properties, the 3a protein was preincubated with proteinase K (Fig. 2A, lane 1), heated at 1007 for 10 min (lane 2), or treated with 0.1% SDS (lane 3) prior to incubation with labeled RNA transcript (lane 5). All three treatments either abolished or greatly reduced the ability of the 3a protein to bind ss RNA when compared to the nontreated 3a protein (Fig. 2A, lane 4). The remaining RNA binding after proteinase K treatment probably was due to incomplete digestion of the 3a protein. Digesting the 3a protein:RNA complexes with RNase A gave rise to a protein:RNA core (Fig. 2A, lanes 7–9) which migrated faster than the free RNA transcript (lane 5) and showed only a slight difference in mobility when digested with different amounts of RNase A, indicating that the 3a protein binds to discrete lengths of the RNA and not as an aggregate of protein to some of the RNA. No such RNase-resistant cores were observed in the absence of the 3a protein (Fig. 2A, lane 6). These data suggest that cooperative binding occurs between the 3a protein and the viral RNA. To determine whether the 3a protein has any sequence specificity or preference in binding RNA, competition binding assays were performed by incubating unlabeled competitor nucleic acids of various types and 32Plabeled CMV RNA 3 together with the CMV 3a protein, followed by analysis of the incubation mixture on a nondenaturing agarose gel. The results shown in Fig. 2B demonstrated that the heterologous ss viral RNA (lanes 1 and 2, TMV RNA), homologous ss viral RNA (lanes 3 and 4, CMV RNAs), and heterologous ss viral DNA (lanes 7 and 8, M13 phage DNA) were all able to compete effectively with the labeled transcript in binding to the 3a protein. At the higher competitor concentration (0.1 mg vs 0.025 mg/15 ml), more labeled RNA showed a shift from the position of completely retarded RNA to the position of free RNA (Fig. 2B, lanes 1 vs 2 and 7 vs 8). However,
AP-Virology
74
LI AND PALUKAITIS
FIG. 1. Expression, purification, and analysis of the CMV and TMV movement proteins. (A) SDS–PAGE of the expressed CMV 3a (lanes 1–5) and TMV 30K (lanes 6–10) proteins extracted from E. coli. The total bacterial lysate was analyzed before (lanes 1 and 6) and after (lanes 2 and 7) induction of expression. The induced lysate was centrifuged to remove the overexpressed protein, present in the pellet as an inclusion (lanes 4 and 9) from the soluble bacterial proteins (lanes 3 and 8). The inclusions were solubilized in 8 M urea, centrifuged to remove insoluble materials, and dialyzed to remove the urea (lanes 5 and 10). The gel was stained with Coomassie blue. The positions of the molecular weight markers are indicated. The arrow indicates the position of the movement proteins. (B) Retardation gel electrophoresis assay for RNA binding by the movement proteins purified from E. coli. Increasing amounts of either the TMV 30K protein (lanes 1–6) or the CMV 3a protein (lanes 7–11) were incubated in 15 ml of binding buffer A with 1 ng of 32P-labeled CMV RNA 3 transcript, and the mixture was electrophoresed in a 1% nondenaturing agarose gel. The amount of movement proteins used in the assays is indicated.
neither ds CMV RNAs (Fig. 2B, lanes 9 and 10) nor sheared yeast RNA fragments (õ50 nt; Fig. 2B, lanes 5 and 6) were effective competitors, although the ds RNAs did show a weak competition. Thus, the CMV 3a protein binds specifically to ss nucleic acid without obvious sequence preference. Verification of the ability of the CMV 3a protein and not just contaminating higher molecular weight E. coli proteins to bind ss RNA (Thomas and Maule, 1995) was demonstrated by UV light crosslinking experiments. The 32 P-labeled RNA was incubated with the 3a protein and
the incubation mixture was irradiated with UV light. This treatment can covalently crosslink protein:RNA complexes that are in close molecular association. Unbound RNA was removed by RNase A digestion, and the mixture was analyzed by SDS–PAGE and autoradiography. As shown in Fig. 3 (lanes 1–4), a strong band could be seen in the position expected for the 3a protein (arrow). Samples containing less RNase showed a band with a broader and slower mobility (Fig. 3, lane 4 vs lane 1). When no RNase was added, this band disappeared into a smeared background (Fig. 3, lane 5), with most of the
FIG. 2. Determination of the in vitro binding parameters of the CMV 3a movement protein. (A) RNA binding characteristics of the CMV 3a protein analyzed by retardation gel electrophoresis. The CMV 3a protein (250 ng in 15 ml) was pretreated with proteinase K (1 mg at 377 for 30 min; lane 1), 0.1% SDS (lane 2), or heat (1007, 10 min; lane 3) before incubation with 5 ng of 32P-labeled CMV RNA 3 transcript. The 3a protein (250 ng in 15 ml) was incubated with 32P-labeled RNA (5 ng) and either not treated further (lane 4) or incubated with different amounts of RNase A (room temperature, 15 min): 1 mg (lane 7), 250 ng (lane 8), or 60 ng (lane 9). The above samples and protein-free 32P-labeled RNA (5 ng), either incubated (lane 6) or not incubated (lane 5) with 60 ng of RNase A, were resolved on a 1% nondenaturing agarose gel. (B) Competition binding assay with the CMV 3a protein. The CMV 3a protein (250 ng in 15 ml) was incubated with 5 ng of labeled CMV RNA 3 transcript either in the absence of (lane 11) or in the presence of nonlabeled competitors (lanes 1–10): TMV RNA (lanes 1 and 2); CMV RNAs (lanes 3 and 4); sheared, yeast RNA (lanes 5 and 6); M13 ss DNA (lanes 7 and 8); or ds CMV RNAs (lanes 9 and 10). Two levels of competitor were used: 1 mg (lanes 1, 3, 5, 7, and 9) or 250 ng (lanes 2, 4, 6, 8, and 10). The labeled RNA was also incubated with buffer alone (lane 12).
/ 6a0f$$7678
01-12-96 08:20:51
vira
AP-Virology
CMV AND TMV MOVEMENT PROTEINS
FIG. 3. UV crosslinking of the movement proteins with labeled RNA. The CMV 3a protein was incubated with labeled RNA (10 ng) for 30 min on ice. After UV irradiation, the mixture was either untreated (lane 5) or incubated with different amounts of RNase A (lanes 1–4, 6–9): 2 mg (lane 1), 500 ng (lane 2), 130 ng (lane 3), or 30 ng (lanes 4, 6–9). Incubations were also carried out in the presence of either excess nonlabeled competitor ss CMV RNAs (1 mg; lane 7) or ds CMV RNAs (1 mg; lane 8), or either BSA (lane 6) or the TMV 30K protein (10 ng; lane 9) instead of the CMV 3a protein. The samples were analyzed by SDS–PAGE and autoradiography. The arrow indicates the position of the 3a protein.
labeled RNA at the top of the gel. When the incubation was carried out in the presence of excess unlabeled CMV RNA (Fig. 3, lane 7), when the UV irradiation was omitted (not shown), or when BSA was substituted for the 3a protein (lane 6), no band was observed. Moreover, ds RNA could not compete with the binding of labeled RNA to the 3a protein (Fig. 3, lane 8). The TMV 30K protein also was UV-crosslinked by the labeled RNA and gave a band of the expected size (Fig. 3, lane 9). Stability of MP:RNA complexes at different salt concentrations Stability of protein:RNA complexes with respect to salt concentration is often a criterion used to evaluate the strength of a protein:RNA association. The complexes formed between the 32P-labeled RNA transcript and the 3a or 30K proteins were incubated in buffers with different concentrations of NaCl, and the incubation mixtures were filtered through a nitrocellulose membrane and washed to remove unbound RNA, leaving only protein and protein:RNA complexes on the membrane. This method is effective in detecting small amounts of protein:RNA complexes which otherwise would not been observed by gel retardation. As shown in Fig. 4, both MPs bind RNA maximally between 50 and 200 mM NaCl, and at 400 mM NaCl both MPs showed a significant decrease in RNA binding. At 600 mM NaCl, the 3a protein could no longer bind RNA, while the 30K protein still retained about half of its RNA binding capability (vs at 200 mM NaCl) (Fig. 4). At higher salt concentrations, the 30K protein was also dissociated from the RNA. Nucleotide binding specificity of the CMV and TMV MPs Sequence analyses of the tobamovirus MPs indicated that they contain sequence motifs of possible NTP bind-
/ 6a0f$$7678
01-12-96 08:20:51
vira
75
FIG. 4. Salt stabilities of RNA:movement protein complexes. The TMV 30K and the CMV 3a proteins (250 ng) were incubated on ice with labeled RNA (1 ng) in the presence of the indicated concentrations of NaCl. After incubation, the mixtures were analyzed by nitrocellulose membrane filter binding. RNA binding was quantified by determining the radioactivity on the membrane by liquid scintillation counting.
ing domains (Saito et al., 1988). Thus, both the TMV 30K protein and the CMV 3a protein were tested here for their ability to bind to any of the four NTPs. Both MPs were incubated with each of the four a-32P-labeled NTPs, followed by UV irradiation to crosslink covalently any bound NTP, and were analyzed by SDS–PAGE. As shown in Fig. 5, both MPs were clearly labeled by GTP and weakly labeled by UTP, but were not labeled by either ATP or CTP. In some but not all experiments in which BSA was substituted for the MPs, a very weak binding to GTP was observed (not shown), similar in intensity to that observed between the TMV 30K protein and UTP (Fig. 5, lane 4). When an excess of unlabeled GTP (25 mM) was incubated along with 32P-labeled GTP (12.5 nM) and the 3a protein prior to UV irradiation, very little 3a protein became labeled (Fig. 5, lane 5), while the presence of
FIG. 5. Nucleotide binding by the movement proteins. The TMV 30K and the CMV 3a proteins (250 ng in 10 ml) were incubated with each of the four [a-32P]NTPs: ATP (lane 1), CTP (lane 2), GTP (lanes 3 and 7), and UTP (lane 4), at 12.5 nM. The incubation mixtures were irradiated under UV light and analyzed by SDS–PAGE. Incubation and irradiation of the CMV 3a protein was also carried out with [a-32P]GTP in the presence of the nonlabeled competitors 25 mM GTP (lane 5) and 25 mM ATP (lane 6). The samples were analyzed by SDS–PAGE and autoradiography. The arrow indicates the position of the 3a protein.
AP-Virology
76
LI AND PALUKAITIS
unlabeled ATP (25 mM) in the incubation mixture with the 3a protein and labeled GTP only slightly reduced the binding of 3a protein to GTP (Fig. 5, lane 6). The residual labeling seen in Fig. 5, lane 5, could be due to either a high Km for GTP concentration or the high concentration of protein (0.8 mM) in the assay. A similar competition assay was not done with labeled UTP or using the TMV 30K protein. These data indicate that the 3a and 30K proteins are GTP (and possibly UTP)-binding proteins. The higher molecular weight bands present in Fig. 5, lane 3, may be crosslinked aggregates of the respective MPs or contaminating E. coli proteins. Bands of similar mobility were also observed when RNA was bound to the MPs by UV crosslinking (Fig. 3). Tagging of the 3a protein with six histidines An in-frame insertion of six histidine (His) codons was introduced into the 3* terminus of the 3a ORF of pFny309 by site-directed mutagenesis, resulting in addition of six His residues to the C-terminus of the 3a protein. The presence of the introduced six His codons in pFny309-His, the modified full-length Fny-CMV RNA 3 cDNA clone, was verified by sequencing (data not shown). Transcript of RNA 3 obtained by in vitro transcription of pFny309-His, along with transcripts of Fny-CMV RNAs 1 and 2, was inoculated onto tobacco plants. Five days postinoculation, the tobacco plants showed green mosaic symptoms indistinguishable from those induced by wild-type Fny-CMV. Virus was extracted from the infected plants 12 days postinoculation and the yield was equivalent to wild-type Fny-CMV. The six-His-codon tag was retained in the viral RNA during two passages in tobacco plants and no other changes were detected by RNA sequencing (data not shown). The Fny-CMV encoding the His-tagged 3a protein was also able to infect zucchini squash with infectivity kinetics similar to those of the unmodified virus. Isolation of His-tagged 3a protein from both E. coli and plants The His codon-tagged 3a gene was ligated into the vector pET11a and cloned into and overexpressed in E. coli, and the His-tagged 3a protein was purified by Ni2/NTA-affinity chromatography. The protein purified from the E. coli cells transformed with pETMP3a-His contained a single band of the expected size less contaminated by bacterial protein than the conventionally purified nontagged 3a protein (Fig. 6A, lane 5, and data not shown). Under the same conditions, no protein was obtained from E. coli cells transformed with pETMP3a, in which the 3a gene was not tagged (data not shown). The isolated protein samples from infected plants were far less homogeneous. As shown in Fig. 6A, a band with the molecular weight expected for the 3a protein (arrow) appeared in fractions eluting at 50 mM (lane 2) and higher molarity (lanes 3 and 4) imidazole. Several
/ 6a0f$$7678
01-12-96 08:20:51
vira
other bands with mobilities less than and greater than that of the 3a protein also appeared in fractions eluting with 50 and 100 mM imidazole (Fig. 6A, lanes 2 and 3). In the fraction eluting in 250 mM imidazole (Fig. 6A, lane 4), only one band with a mobility similar to that of the E. coli-expressed 3a protein (Fig. 6A, lane 5) was observed. This band and several higher molecular weight bands reacted very strongly with the 3a antiserum in an immunoblot assay (Fig. 6B, lanes 1–5), indicating their relationship to the 3a protein. The higher molecular weight forms of the 3a protein may be the aggregates previously described in gels showing RNAs (Fig. 3) or NTPs (Fig. 5) crosslinked to the CMV 3a protein expressed in E. coli. None of the bands described above were detected when extracts of noninfected plants were applied to the Ni2/NTA column and eluted with imidazole (data not shown). RNA-binding activity of the His-tagged 3a protein Using Ni2/-NTA chromatography, the His-tagged 3a protein expressed in E. coli was purified to homogeneity as demonstrated by silver staining (Fig. 6A, lane 5). The His-tagged 3a protein was then examined for RNA binding ability by gel retardation (not shown) and UV-crosslinking assays (Fig. 6C). The His-tagged 3a protein was able to bind ss RNA cooperatively (not shown) with a pattern similar to that of the nontagged 3a protein (Fig. 1B). In UV-crosslinking assays, the His-tagged 3a protein purified from E. coli was crosslinked to the labeled RNA (Fig. 6C, lane 5) and was of the same size as the crosslinked nontagged 3a protein (Fig. 6C, lane 3a). An additional band with about twice the molecular weight of the 3a protein was also crosslinked to the labeled RNA. Since the His-tagged 3a sample purified from E. coli was devoid of any detectable bacterial contaminant proteins (Fig. 6A, lane 5), this additional band is probably a dimer of the 3a protein. The intense bands migrating faster than the 3a protein (Fig. 6C, lane 5) may be degradation products of the His-tagged 3a protein. The His-tagged 3a protein sample isolated from infected squash and eluting from the Ni2/-NTA column in 250 mM imidazole also was analyzed in a similar UV-crosslinking assay for the ability to bind RNA. As shown in Fig. 6C, a weak band with a slightly faster migration than the E. coliexpressed 3a protein was crosslinked to the RNA (lanes 4, arrow). The data shown in Fig. 6 were obtained using extracts of CMV-infected squash plants. These contained a higher concentration of MP than extracts of CMV-infected tobacco (data not shown). Nevertheless, the Histagged 3a protein isolated from squash was not tested for either cooperative RNA binding or NTP binding, because of the low levels of protein present in these assays compared to those present in previous ones (Figs. 1–5). DISCUSSION The results presented here show that the CMV (3a) MP binds ss RNA in a cooperative manner similar to that
AP-Virology
CMV AND TMV MOVEMENT PROTEINS
77
FIG. 6. Isolation and characterization of the His-tagged 3a protein from both E. coli and CMV-infected squash leaves. The squash leaf extract was subjected to affinity chromatography on a Ni2/-NTA column. After extensive washing with buffer containing 10 mM imidazole, the proteins were eluted from the column sequentially in buffer containing imidazole at 25 mM (lane 1), 50 mM (lane 2), 100 mM (lane 3), and 250 mM (lane 4). The last wash fraction (lane 0), the elution fractions (lanes 1–4), and the His-tagged 3a protein isolated from E. coli and purified by Ni2/-NTA-affinity chromatography (lane 5) were analyzed by SDS–PAGE and either (A) stained with silver reagents or (B) blotted to a nitrocellulose membrane and immunoprobed with an antiserum to the 3a protein. In C, samples corresponding to those seen in lanes 4 and 5 of A and B as well as nontagged 3a protein expressed in and purified from E. coli by the conventional method were incubated with 32P-CMV RNA 3, UV-irradiated, and incubated with RNase A (1 mg) as in Fig. 3 prior to analysis by SDS–PAGE and autoradiography. The arrows indicate the position of the 3a protein.
of the TMV (30K) MP. No sequence specificity was found in the RNA binding by the 3a protein. The 3a protein cannot bind ds RNA or very small fragments of ss RNA (e.g., sheared yeast RNA fragments), indicating that the binding requires some minimum length of nucleic acid. In these respects, the CMV MP is similar to the other viral MPs studied previously. The observation that discrete lengths of CMV RNA 3, but not the entire RNA molecule, were protected against digestion with RNase indicated that the entire RNA 3 is not coated with MP as a result of the cooperative binding. There are no extensive sequence similarities between the MPs of different virus families. Results from experiments to identify the sequences in the MPs of TMV (Citovsky et al., 1992), alfalfa mosaic virus (AlMV; Schoumacher et al., 1992), CaMV (Thomas and Maule, 1995), and red clover necrotic mosaic virus (RCNMV; Osman et al., 1993; Giesman-Cookmeyer and Lommel, 1993) responsible for RNA binding showed that no consensus RNA-binding domain or motif exists, although most MPs are found to be rich in basic amino acids. It may be questioned whether the 3-D structure itself of the MP is important for its RNA-binding activity or whether the
/ 6a0f$$7678
01-12-96 08:20:51
vira
binding is just a result of electrostatic attractions between the positively charged sidechains of the MP and the negatively charged backbone of the RNA. When the 3a protein was heat denatured, which presumably would cause more charged amino acid sidechains to become exposed, it lost the ability to bind ss RNA. We also observed that complete refolding of the protein by extensive dialysis was crucial for it to achieve full capacity of RNA binding (data not shown). These data suggest that the binding of the 3a protein is structure-dependent. Although the CMV 3a protein showed an RNA binding pattern similar to that of the TMV 30K protein, the protein:RNA complex formed by the CMV 3a protein is dissociated more readily by NaCl than the complex formed by the TMV 30K protein. Different virus MPs have been reported to have different degrees of salt stability in binding to ss RNA. The AlMV 3a protein (Schoumacher et al., 1992) and the CaMV P1 (Citovsky et al., 1991) have been shown to have significantly decreased RNA binding activity at 0.2 M NaCl. The TMV 30K protein was reported to be able to withstand 0.6 M NaCl (Citovsky et al., 1990), similar to the result obtained here. Binding of the MPs of RCNMV (Osman et al., 1992) and foxtail mosaic potex-
AP-Virology
78
LI AND PALUKAITIS
virus (FoMV) (Rouleau et al., 1994) to RNA is stable at 0.4 M NaCl, similar to what is shown here for the CMV 3a protein. Both the TMV 30K MP and the CMV 3a MP showed the ability to bind GTP. There is a predicted NTP-binding sequence in both proteins, although this same region containing a stretch of hydrophobic residues preceding a conserved Asp was proposed in another sequence alignment study to be important for membrane association of the MPs (Mushegian and Koonin, 1993). However, one of the triple gene block proteins (26K) of potexviruses contains a predicted sequence motif conserved in a wide range of NTP-binding and hydrolysis proteins (Gorbalenya and Koonin, 1989). Recently, the 26K protein of FoMV was expressed in E. coli and was found to bind both ATP and CTP and to have apparent ATPase activity (Rouleau et al., 1994). It was speculated that ATP hydrolysis by the 26K protein may provide the energy needed for transport of the virus from cell to cell. However, the TMV 30K or CMV 3a protein showed a specificity in nucleotide binding different from the FoMV 26K protein. It is very tempting in this case to speculate that these MPs may participate in different plant cell signaling pathways which lead to down-regulation of the gating capacity of plasmodesmata. With regard to the various biochemical properties examined, the CMV MP is most similar to the TMV MP. Both viruses also require particular sequences near, but not at, the C-terminus of the respective MPs for movement (Kaplan et al., 1995). However, the difference in subcellular distribution of the two MPs and the inability of the CMV 3a protein expressed transgenically to complement the movement of movement-defective TMV (Kaplan et al., 1995), while transgenically expressed TMV MP will complement the local movement of movementdefective CMV (I. B. Kaplan and P. Palukaitis, unpublished), indicate that there are similar aspects to the movement processes of these two viruses, but the corresponding MP functions are not identical. This is also consistent with the observation that other sequences in both viral genomes can influence viral movement (Nelson et al., 1993; Carr et al., 1994; Gal-On et al., 1994). The reported RNA binding activity of other viral MPs were all conducted with the proteins expressed in an in vitro system and usually the expressed proteins appeared to be insoluble. In most cases, the expressed proteins were solubilized in a buffer containing a denaturant (e.g., urea or SDS) and then dialyzed to remove the denaturant. In this study, we employed the powerful His-tagging system to facilitate purification of the 3a MP from E. coli as well as from CMV-infected plants. Tagging the 3a protein with six histidines did not affect RNA binding of the E. coli-expressed protein. Using Ni2/-NTA-affinity chromatography, the in planta-generated 3a protein was purified extensively. The His-tagged 3a protein samples isolated from the infected plants were also able to
/ 6a0f$$7678
01-12-96 08:20:51
vira
bind ss RNA. This is the first time that a virus movement protein has been purified from infected plants and shown to contain RNA binding activity. This system should allow purification of larger quantities of the native MP, as well as allow us to discern whether any of the proteins copurifying with the 3a protein (Fig. 6A) are involved in the movement process. ACKNOWLEDGMENT This work was supported by Grant DE-FG02-86ER13505 from the U.S. Department of Energy.
REFERENCES Bush, G. L., Tassin, A., Friden, H., and Meyer, D. I. (1991). Purification of a translocation-competent secretory protein precursor using nickel ion affinity chromatography. J. Biol. Chem. 266, 13811–13814. Carr, J. P., Gal-On, A., Palukaitis, P., and Zaitlin, M. (1994). Replicasemediated resistance to cucumber mosaic virus in transgenic plants involves suppression of both virus replication in the inoculated leaves and long-distance movement. Virology 199, 439–447. Citovsky, V., Knorr, D., Schuster, G., and Zambryski, P. (1990). The p30 movement protein of tobacco mosaic virus is a single-strand nucleic acid binding protein. Cell 60, 637–648. Citovsky, V., Knorr, D., Schuster, G., and Zambryski, P. (1991). Gene I, a potential cell-to-cell movement locus of cauliflower mosaic virus, encodes an RNA-binding protein. Proc. Natl. Acad. Sci. USA 88, 2476–2480. Citovsky, V., Wong, M. L., Shaw, A. L., Prasad, B. V. V., and Zambryski, P. (1992). Visualization and characterization of tobacco mosaic virus movement protein binding to single-stranded nucleic acids. Plant Cell 4, 397–411. Davies, C., and Symons, R. H. (1988). Further implications for the evolutionary relationships between tripartite plant viruses based on cucumber mosaic virus RNA 3. Virology 165, 216–244. Deom, C. M., Oliver, M. J., and Beachy, R. N. (1987). The 30-kilodalton gene product of tobacco mosaic virus potentiates virus movement. Science 237, 389–394. Ding, B., Li, Q., Nguyen, L., Palukaitis, P., and Lucas, W. J. (1995). Cucumber mosaic virus 3a protein potentiates cell-to-cell trafficking of CMV RNA in tobacco plants. Virology 207, 345–353. Fujiwara, T., Giesman-Cookmeyer, D., Ding, B., Lommel, S. A., and Lucas, W. J. (1993). Cell-to-cell trafficking of macromolecules through plasmodesmata potentiated by the red clover necrotic mosaic virus movement protein. Plant Cell 5, 1783–1794. Gal-On, A., Kaplan, I., Roossinck, M. J., and Palukaitis, P. (1994). The kinetics of infection of zucchini squash by cucumber mosaic virus indicate a role for RNA 1 in virus movement. Virology 205, 280–289. Giesman-Cookmeyer, D., and Lommel, S. A. (1993). Alanine scanning mutagenesis of a plant virus movement protein identifies three functional domains. Plant Cell 5, 973–982. Giesman-Cookmeyer, D., Silver, S., Vaewhongs, A. A., Lommel, S. A., and Deom, C. M. (1995). Tobamovirus and dianthovirus movement proteins are functionally homologous. Virology 213, 38–45. Gorbalenya, A. E., and Koonin, E. V. (1989). Viral proteins containing the purine NTP-binding sequence pattern. Nucleic Acids Res. 17, 8413–8440. Hull, R. (1989). The movement of viruses in plants. Annu. Rev. Phytopathol. 27, 213–340. Kaplan, I. B., Shintaku, M. H., Li, Q., Zhang, L., Marsh, L. E., and Palukaitis, P. (1995). Complementation of movement defective mutants in transgenic tobacco expressing the cucumber mosaic virus movement gene. Virology 209, 188–199. Lucas, W. J., and Gilbertson, R. L. (1994). Plasmodesmata in relation
AP-Virology
CMV AND TMV MOVEMENT PROTEINS to viral movement within leaf tissues. Annu. Rev. Phytopathol. 32, 387–411. Maule, A. J. (1991). Virus movement in infected plants. Crit. Rev. Plant Sci. 9, 457–474. McLean, B. G., Waigmann, E., Citovsky, V., and Zambryski, P. (1993). Cell-to-cell movement of plant viruses. Trends Microbiol. 1, 105–109. Melcher, U. (1990). Similarities between putative transport proteins of plant viruses. J. Gen. Virol. 71, 1009–1018. Meshi, T., Watanabe, Y., Saito, T., Sugimoto, A., Maeda, T., and Okada, Y. (1987). Function of the 30kd protein of tobacco mosaic virus: Involvement in cell-to-cell movement and dispensability for replication. EMBO J. 6, 2557–2563. Mushegian, A. R., and Koonin, E. V. (1993). Cell-to-cell movement of plant viruses: Insights from amino acid sequence comparisons of movement proteins and from analogies with cellular transport systems. Arch. Virol. 133, 239–257. Nelson, R. S., Guoxuan, L., Hodgson, R. A. J., Beachy, R. N., and Shintaku, M. H. (1993). Impeded phloem-dependent accumulation of the masked strain of tobacco mosaic virus. Mol. Plant–Microbe Interact. 6, 45–54. Nishiguchi, M., Motoyoshi, F., and Oshima, N. (1978). Behavior of a temperature sensitive strain of tobacco mosaic virus in tomato leaves and protoplasts. J. Gen. Virol. 39, 53–61. Noueiry, A. O., Lucas, W. J., and Gilbertson, R. L. (1994). Two proteins of a plant DNA virus coordinate nuclear and plasmodesmal transport. Cell 76, 925–932. Osman, T. A. M., Ingles, P. J., Miller, S. J., and Buck, K. W. (1991). A spontaneous red clover necrotic mosaic virus mutant with a truncated movement protein. J. Gen. Virol. 72, 1793–1800. Osman, T. A. M., Hayes, R. J., and Buck, K. W. (1992). Cooperative binding of the red clover necrotic mosaic virus movement protein to single-stranded nucleic acids. J. Gen. Virol. 73, 223–227. Osman, T. A. M., Thommes, P., and Buck, K. W. (1993). Localization of a single-stranded RNA-binding domain in the movement protein of red clover necrotic mosaic dianthovirus. J. Gen. Virol. 74, 2453–2457. Palukaitis, P., Garcia-Arenal, F., Sulzinski, M. A., and Zaitlin, M. (1983). Replication of tobacco mosaic virus. VII. Further characterization of single- and double-stranded virus-related RNAs from TMV-infected tobacco plants. Virology 131, 533–545. Palukaitis, P., Roossinck, M. J., Dietzgen, R. G., and Francki, R. I. B. (1992). Cucumber mosaic virus. Adv. Virus Res. 41, 281–348. Pascal, E., Sanderfoot, A. A., Ward, B. M., Medville, R., Turgeon, R., and Lazarowitz, S. G. (1994). The geminivirus BR1 movement protein
/ 6a0f$$7678
01-12-96 08:20:51
vira
79
binds single-stranded DNA and localizes to the cell nucleus. Plant Cell 6, 995–1006. Perbal, M. C., Thomas, C. L., and Maule, A. J. (1993). Cauliflower mosaic virus gene I product P1 forms tubular structures which extend from the surface of infected protoplasts. Virology 195, 281–285. Rizzo, T. M., and Palukaitis, P. (1990). Construction of full-length cDNA clones of cucumber mosaic virus RNAs 1, 2, and 3: Generation of infectious RNA transcripts. Mol. Gen. Genet. 222, 249–256. Rouleau, M., Smith, R. J., Bancroft, J. B., and Mackie, G. A. (1994). Purification, properties, and subcellular localization of foxtail mosaic potexvirus 26-kDa protein. Virology 204, 254–265. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). ‘‘Molecular Cloning: A Laboratory Manual,’’ 2nd ed. Cold Spring Harbor Laboratories, Cold Spring Harbor, NY. Saito, T., Imai, Y., Meshi, T., and Okada, Y. (1988). Interviral homologies of the 30K proteins of tobamoviruses. Virology 176, 653–656. Schoumacher, F., Erny, C., Berna, A., Godefroy-Colburn, T., and StussiGaraud, C. (1992). Nucleic acid-binding properties of the alfalfa mosaic virus movement protein produced in yeast. Virology 188, 896– 899. Shintaku, M. H., Zhang, L., and Palukaitis, P. (1992). A single amino acid substitution in the coat protein of cucumber mosaic virus induces chlorosis in tobacco. Plant Cell 4, 751–757. Suzuki, M., Kuwata, S., Kataoka, J., Masuta, C., Nitta, N., and Takanami, Y. (1991). Functional analysis of deletion mutants of cucumber mosaic virus RNA3 using an in vitro transcription system. Virology 183, 106–113. Thomas, C. L., and Maule, A. J. (1995). Identification of the cauliflower mosaic virus movement protein RNA-binding domain. Virology 206, 1145–1149. Thomas, C. L., Perbal, C., and Maule, A. J. (1993). A mutation of cauliflower mosaic virus gene I interferes with virus movement but not virus replication. Virology 192, 415–421. van Lent, J., Wellink, J., and Goldbach, R. (1990). Evidence for the involvement of the 58K and 48K proteins in the intercellular movement of cowpea mosaic virus. J. Gen. Virol. 71, 219–224. Waigmann, E., Lucas, W. J., Citovsky, V., and Zambryski, P. (1994). Direct functional assay for tobacco mosaic virus cell-to-cell movement protein and identification of a domain involved in increasing plasmodesmal permeability. Proc. Natl. Acad. Sci. USA 91, 1433–1437. Wong, I., and Lohman, T. M. (1993). A double-filter method for nitrocellulose-filter binding: Application to protein–nucleic acid interactions. Proc. Natl. Acad. Sci. USA 90, 5428–5432.
AP-Virology