Molecular cloning of DNA complementary to tobacco vein mottling virus RNA

128, 210-220

VIROLOGY

(1983)

Molecular GARY

Cloning of DNA Complementary Vein Mottling Virus RNA

M. HELLMANN,

MUHAMMAD ROBERT

AND Departments

of Biochemistry

and *Plant Received

Pathology,

January

to Tobacco

SHAHABUDDIN, E. RHOADS’

JOHN

University

L..exington,

26, 198.5’; accepted

of Kentucky, March

G. SHAW,* Kentucky

40536

8, 1983

RNA isolated from tobacco vein mottling virus (TVMV) was used as a template for avian myeloblastosis virus RNA-dependent DNA polymerase, primed with oligo(dT). The largest single-stranded cDNA synthesized was 10 kb, the same as the viral RNA. This material was converted to double-stranded cDNA with Esherichia coli DNA polymerase I and digested with restriction endonuclease Hi?zdIII. The cDNA fragments were ligated to HindIII-digested plasmid pBR322 and the product used to transform E. coli strain DG-75. Clones containing recombinant plasmids were selected by ampicillin resistance, and those containing TVMV RNA sequences were selected by colony hybridization with a single-stranded cDNA probe. Four different sizes of recombinant plasmid were reproducibly observed. The inserted DNA portion could be excised in each case with HindIII. The lengths of inserted DNA were 3.0, 1.85, 1.1, and 0.72 kh. A similar procedure was used with PstI-digested cDNA and pBR322. A single type of recombinant plasmid, containing a DNA insertion of 1.85 kb, was reproducibly observed. Hybridization with TVMV RNA confirmed that the five inserted DNA segments were derived from the viral RNA. Hybridization of each recombinant plasmid with the others established that each of the cloned Hind111 fragments was unique and that one of them overlapped the cloned PstI fragment. The cloned cDNA fragments were ordered by establishing a restriction map of the cDNA. Together the cloned cDNA fragments account for over 80% of the viral genome.

netic map and to locate the termini of cistrons more exactly, we have undertaken the molecular cloning of DNA complementary to the RNA of one member of the potyvirus group, tobacco vein mottling virus (TVMV).’

INTRODUCTION

The potyvirus group is perhaps the largest, most widely distributed, and most destructive of the groups of plant viruses (Edwardson, 1974). In spite of this, relatively little is known of the manner in which these viruses replicate and express their genetic information. Previous work (Dougherty and Hiebert, 1980a, b; Hellmann et al, 1980,1983) has established the existence of five polypeptides encoded by potyviral RNA. Based on translational studies, a genetic map locating the cistrons corresponding to four of these proteins on the viral genome has been proposed by Dougherty and Hiebert (198Oc). To provide further evidence for this ge’ Author addressed.

to whom

0042-6822/83 Copyright All rights

requests

for

reprints

$3.00

0 1983 by Academic Press, Inc. of reproduction in any form reserved.

should

MATERIALS

AND

METHODS

Materials. Avian myeloblastosis virus (AMV) reverse transcriptase was provided by Dr. J. W. Beard, Life Sciences, Inc. All restriction endonucleases as well as DNA polymerase I were from New England Biolabs. T4 DNA ligase and polynucleotide kinase were from Bethesda Research Lab* Abbreviations used: TVMV, tobacco vein mottling virus; AMV, avian myeloblastosis virus; cDNA, complementary DNA; SSC, 0.15 M NaCI, 0.015 M sodium citrate; SDS, sodium dodecyl sulfate; BSA, bovine serum albumin.

be

210

CLONING

OF

oratories. RNasin (placental ribonuclease inhibitor) was obtained from Biotec, Inc. Actinomycin D was a gift from Merck, Sharp, and Dohme Research Laboratories. S, nuclease was from Miles. Nitrocellulose filters were from Schleicher and Schuell. TVMV RNA was purified as previously described (Hellmann et al, 1980). Each batch of RNA was analyzed for purity and integrity on both polyacrylamide/urea gels (Maniatis et al, 1975) and, after glyoxylation, on agarose gels (McMaster and Carmichael, 1977). Agarose gel electrophwesis. Alkaline gel electrophoresis of DNA was performed in 30 mM NaOH, 2 mM EDTA by the procedure of McDonnell et al. (1977). Neutral gel electrophoresis was performed in 50 mM Tris-HCl, pH 7.9, 4 mM sodium acetate, 1 mMEDTA by the method of Wigler et al. (1977). In all cases, X DNA digested with Hind111 was used as molecular weight markers. DNA bands were visualized by staining with ethidium bromide. Synthesis of cDNA. DNA complementary to TVMV RNA was prepared in three different ways. Low specific radioactivity, double-stranded cDNA for cloning was synthesized as follows. Reaction mixtures, typically 100 ~1, containing 50 mM TrisHCl, pH 8.3, 8 mM MgClz, 4 mM dithiothreitol, 40 pg/ml actinomycin D, 20 pg/ ml oligo(dT)lz-ls (P-L Biochemicals), 800 PM each of dATP, dCTP, dGTP, dTTP, 50 mM KCl, 1000 U/ml RNasin, 180 U/ml reverse transcriptase, and 50 pg/ml TVMV RNA, were incubated for 1 hr at 41” in the presence of 0.5 &i/p1 of [a-32P]dCTP (Amersham). Incorporation of radioactivity into trichloroacetic acid-precipitable form was determined by spotting aliquots on glass-fiber squares (Rhoads and Hellmann, 1978). Size determination of the single-stranded cDNA was on alkaline 0.6% agarose gels. The cDNA was prepared for synthesis of the second DNA strand as described by Goodman and MacDonald (1979). Synthesis of the second strand was performed by the method of Bothwell et al. (1981). Reaction volumes of 100 ~1 were typically used, and incorporation of [a“2P]dCTP (5 &i) into trichloroacetic acidprecipitable form was monitored as de-

TVMV

RNA

211

scribed above. Double-stranded cDNA was extracted with phenol:chloroform and chromatographed on G-75 Sephadex (Goodman and MacDonald, 1979). cDNA samples were assayed for conversion from single-stranded to double-stranded form by digestion with S1 nuclease as previously described (Rhoads and Hellmann, 1978). High specific radioactivity, singlestranded cDNA for use as probe was synthesized in reactions containing 50 mM Tris-HCl, pH 8.3, 10 mM MgC12, 1 mM dithiothreitol, 80 mM KCl, 10 pg/ml oligo(dT)12-18, 500 pMeach of dATP, dGTP, and dTTP, 50 PM dCTP, 100 U/ml RNasin, 800 U/ml AMV reverse transcriptase, 100 wg/ml TVMV RNA, and 1 &i/p1 [W ‘“P]dCTP (Amersham). The reaction mixture was incubated for 1 hr at 41”C, made 40 mM in EDTA, and extracted with phenol:chloroform. The organic phase was back-extracted and the aqueous phases pooled and made 130 mM in NaOH. The mixture was heated at 70” for 15 min, passed over a Sephadex G-75 column, and the first peak of radioactivity collected. These preparations had a somewhat lower proportion of full-length material than those prepared by the above procedure. For some experiments, a single-stranded 3’-terminal probe was synthesized as outlined above, but instead of incubation for 1 hr, reactions were terminated after 5 min. The size of the resulting cDNA was analyzed on alkaline agarose gels and found to be approximately 3000 nucleotides, representing the 3’-terminal one-third of the RNA. Restriction endonuclease digestion oj’ DNA. Aliquots of double-stranded cDNA or plasmid DNA were digested with var-, ious restriction endonucleases. The reaction mixtures for PvuII, BamHI, KpnI, and Sal1 contained 150 mM NaCl, 6 mM TrisHCl, pH 7.4, 6 mM MgC12, 6 mM Z-mercaptoethanol, and 100 vg/ml BSA. For HindIII, EcoRI, AvaI, and PstI, the reaction mixtures were the same except 50 mM NaCl was used. For XbaI, the reaction mixture was the same except 6 mM NaCl was used. In all cases, DNA (typically 1 pg) was incubated for 60 min at 37” in the presence of 2 units of enzyme.

212

HELLMANN

Construction of recombinant plasmids. pBR322 was digested with either PstI or Hind111 and treated with alkaline phosphatase as described by Goodman and MacDonald (1979). The linearized plasmid was purified by electrophoresis and eluted from the gel (Zain and Roberts, 1979). Concentration and purity of the material were checked by again subjected an aliquot to electrophoresis. Ligation was performed under the conditions described by Sippel et al. (1978) by mixing 1 pg of linearized pBR322 with 30 ng of cDNA which had been digested with the same enzyme (either Hind111 or Pstl), extracted with phenol:chloroform and precipitated with ethanol. Transformation of E. coli and screening of plasmids. Transformation was performed by the method of Morrison (1979). After ligation, 100 ng of recombinant DNA were incubated with 100 ~1 of competent E. coli DG-75 cells (O’Farrell et ah, 1978). Cells were spread on plates containing either 10 pg/ml of tetracycline (for PstI-digested DNA) or 50 pg/ml of ampicillin (for HindIII-digested DNA). Drug-resistant colonies were replica plated, and the DNA was transferred to nitrocellulose filters (Grunstein and Hogness, 1975) and subjected to hybridization with single-stranded 32P-cDNA as follows:3 prehybridization was carried out in polyethylene bags at 65” for 3 hr in 4X SSC, 0.1% SDS, 100 pg/ml denatured salmon sperm DNA, 0.2% Ficoll, 0.2% polyvinylpyrrolidone, and 0.2% BSA (5 ml per 85-mm-diameter filter). The solution was discarded. Radioactive DNA probes (1 X lo6 cpm) were denatured at lOO”C, diluted with 2 ml of hybridization buffer and added to the bags; hybridization was carried out for 24 hr. Filters were blotted dry and washed 15 min in a solution containing 50% formamide, 2~ SSC, and 0.5% SDS at room temperature with gentle shaking. The wash was repeated first in the same solution but containing 0.5~ SSC and then three times in 2~ SSC. Detection of hybridized probe was by autoradiography. To determine the 3 F. Gannon,

unpublished

procedure.

ET

AL.

size of the recombinant plasmids, 5-ml cultures were grown and plasmid DNA isolated and analyzed by electrophoresis (Birnboim and Doly, 1979). Detection of TVMV-spetific DNA sequences. Recombinant plasmids or cDNA were digested with various restriction endonucleases and the fragments separated from pBR322 by electrophoresis in 1% agarose gels. The DNA was then transferred to nitrocellulose by the method of Southern (1975). Two types of probes were used. The first was TVMV RNA, labeled with 32P by partial alkaline hydrolysis and treatment with [T-~~P]ATP and polynucleotide kinase (Craig and McCarthy, 1980). The second was recombinant plasmids labeled with [(r-32P]dCTP by nick translation (Maniatis et al, 1975). Hybridizations were carried out as described above. RESULTS

Synthesis of cDNA. Hari et al. (1979) have demonstrated that the RNA of tobacco etch virus, another member of the potyvirus group, contains a poly(A) tract. We have previously shown that TVMV RNA is retained on columns of oligo(dT)-cellulose (Hellmann et aL, 1980) and therefore is likely to contain poly(A) as well. Consequently, we used oligo(dT) as a primer for AMV reverse transcriptase. A number of conditions were tested to optimize synthesis of full-length, single-stranded cDNA. Factors which proved to be important included a relatively high concentration of deoxynucleoside triphosphates (800 PM) and the use of RNasin (de Martynoff et al, 1980). The products of the first strand synthesis were analyzed on an alkaline agarose gel (Fig. 1A). A heterogeneous population of molecules was observed, but a significant amount of the radioactivity was present in the region of 10 kb, the presumed length of TVMV RNA (Hellmann et al, 1980). The overall synthesis corresponded to 35-45% copying of the template RNA. The product was tested for single-strandedness with nuclease S1 and found to be 70% sensitive to digestion. For synthesis of the second DNA strand, E. coli DNA polymerase I was used. The

CLONING

OF

TVMV

RNA

A. 23.7

kb-

9.8

kb-

6.6

kb-

4.5

2.5 2.2

I

‘36.6 ::I kb’ 4.5

ktr

kb-

2.5 2.2

kbkb’

kbkbf

0.5

kb-

C.

23.7 9.8 6.6 4.5

kbkbkP kb-

2.5 2.2

kbkb/

0.5

kb-

FIG. 1. Size and restriction analysis of “P-labeled TVMV cDNA. DNA was labeled with 32P, separated by electrophoresis in agarose gels, and detected by autoradiography. (A) Single-stranded TVMV cDNA separated on an alkaline 0.6% agarose gel. (B) Double-stranded TVMV cDNA digested with various restriction endonucleases and separated on neutral 1% agarose gels. (C) Same as B except that, in some cases, double digestions were performed. Numbers to the left of each figure indicate the mobility of fragments of phage X DNA digested with HindID, and run in parallel lanes as molecular weight markers.

213

214

HELLMANN

products of this reaction were analyzed on a neutral agarose gel (Figs. 1B and C, “uncut” lanes) and found also to be heterogeneous, with a maximum size of 10 kb. Overall synthesis by DNA polymerase I represented 40% copying of the singlestranded cDNA template. The product was substantially double-stranded, based on the finding that only 14% was sensitive to nuclease Si. Subsequent references to cDNA will designate the double-stranded product unless otherwise indicated. Construction of recombinant plasmids. Initial attempts to insert full length cDNA of TVMV RNA into bacterial plasmids proved unsuccessful. A different approach was therefore taken: digestion of the plasmid with a restriction endonuclease which cuts in a unique site, digestion of the cDNA with the same enzyme, and ligation of the two to produce a series of plasmids, each containing one restriction fragment of the cDNA. To determine which restriction enzymes would be suitable, cDNA was cut with several enzymes, the products separated on agarose gels, and the resulting DNA fragments detected by autoradiography (Fig. 1B). Discrete bands were produced by a number of restriction endonucleases. (The diffuse background of heterogeneous cDNA molecules was due to incomplete first and second strand synthesis reactions.) Two other enzymes, X&I and KpI, also cut the cDNA (data not shown). BaZI and AvaI appeared not to cleave the cDNA. From the molecular weight markers it was possible to determine the total length of fragments produced by each enzyme. The total length of DNA of each set was (in bp): EcoRI, 8360; HindIII, 9620; BamHI, 10,100; SalI, 1’750; PvuII, 9600; PstI, 1850; KpI, 9525; and X&I, 9000. In most cases, the sum approached 10,000 bp, the estimated length of TVMV RNA. Enzymes such as PstI and SalI may recognize only one site in the molecule, producing a strong band for one fragment near a terminus, and no discrete band for the remainder of the molecule because of size heterogeneity. Hind111 was chosen for cloning studies and used to cleave eDNA and linearize pBR322. To prevent religation of pBR322

ET

AL.

to itself, the linearized plasmid was treated with alkaline phosphatase. Transformed colonies were selected in ampicillin-eontaining medium. Transformants were analyzed for the presence of TVMV RNA sequences by colony hybridization, and positive transformants were screened by plasmid analysis. All transformants which were positive by colony hybridization contained plasmids into which DNA was inserted, as judged by the electrophoretic mobility of isolated plasmids in neutral agarose gels. The recombinant plasmids fell into four different size categories. Plasmids were then subjected to restriction endonuclease analysis. This demonstrated that cloned plasmids of apparent identical size contained identical DNA inserts, thus representing multiple examples of the same recombinant plasmid. It also demonstrated that the inserted sequences could be removed intact with HindIII, indicating that the restriction sites employed in the cloning procedure were preserved. The four different recombinant plasmid types were designated pTVHl (insert of 3000 bp), pTV-H2 (1850 bp), pTV-H3 (‘720 bp), and pTV-H4 (1100 bp). The same procedure was used with PstI instead of HindIII. In this case, selection was on tetracycline-containing plates. Of these transformants (more than 30), only a single type of recombinant plasmid was observed. This was designated pTV-Pl and was shown to contain an insert of 1850 bp. Identification of inserts as cloned TVMV RNA sequences. Inserts were excised with the homologous restriction enzyme (PstI or HindIII), separated on agarose gels, transferred to nitrocellulose, and hybridized with a [‘2P]TVMV RNA probe (Fig. 2A). The predominant hybridization in each case was to the insert. (Hybridization to high-molecular-weight species for plasmids pTV-Hl and -H2 was the result of incomplete digestion with HindIII). The RNA probe did not hybridize to pBR322 itself. In this same experiment [32P]TVMV cDNA digested with either Hind111 or PstI was run on parallel lanes. Each of the DNA inserts migrated to the same position as one of the cDNA restriction fragments (data not shown).

CLONING

OF

PLASMID

H3

H4

PI

TVMV

PLASMID

6.

HZ HI pEi?f? pBR

kb9.8 kb6.6 kb4.5 kb-

23.7

215

RNA

HI

H2

PI

H4

H3

-23.7kb -9.8 kb -6.6 kb -45kb

-0.5kb 05kbFrc;. 2. Recombinant plasmids hybridized to 32P-labeled TVMV RNA and cDNA. Nonradioactive recombinant plasmids pTV-Hl, pTV-HZ, pTV-H3, and pTV-H4 plus the vector pBR322 were digested with HkdIII, and the recombinant plasmid pTV-Pl, with PstI. The products were separated as in Fig. 1B and transferred to nitrocellulose. (A) Hybridization with “P-labeled TVMV RNA. (B) Hybridization with =P-labeled eDNA representing the 3’-terminal3000 nucleotide residues of TVMV RNA.

It was also of interest to determine which of the cloned fragments represented sequences closest to the 3’ terminus of the RNA. An experiment similar to that described above was performed, but a singlestranded cDNA probe corresponding to the 3’-terminal 3000 residues of TVMV RNA was used instead of total viral RNA (Fig. 2B). Hybridization was strongest to the insert of pTV-Pl, weaker to the insert of pTV-Hl and not detected with pTV-H2, -H3, and -H4. This would place the pTVPl insert closest to the 3’ terminus of the viral RNA, with that of pTV-Hl either near or overlapping pTV-Pl. Determination of uniqueness of plasmids. The question of whether the five

plasmids contained overlapping regions of the TVMV genome was examined by hybridization of each plasmid with each of the others. Radioactive probes were prepared by nick translation of each plasmid. Nonradioactive plasmids were digested with the homologous restriction enzyme,

separated by electrophoresis, transferred to nitrocellulose, and hybridized with the various probes (Fig. 3). A band representing pBR322 was evident in each lane because each probe contained labeled pBR322 sequences and each of the nonradioactive plasmids cut with the restriction enzyme generated a linearized pBR322 band. In addition to the pBR322, it can be seen that probes pTV-H2, -H3, and -H4 hybridized only to their homologous inserts. Thus, each insert represents unique sequences. The pTV-Pl probe hybridized strongly to the pTV-Pl insert and weakly to the pTVHl insert, while the pTV-Hl probe hybridized strongly to the pTV-Hl insert and weakly to the pTV-Pl insert. Thus, these two inserts share common sequences. This is also in keeping with the fact that only these two appear to represent sequences near the 3’-terminus of the RNA (Fig. 2B). These results suggest that each of the five plasmid types contains an insert corresponding to one of the Hind111 or P&I

216

HELLMANN

ET

AL.

PROBE :

HZ

PLASMID:

FIG. 3. Recombinant recombinant plasmids bridization was with

pBR HI

H2 PI

H4 H3

pBR HI

H2 PI

H4 H3

plasmids hybridized to “P-labeled recombinant plasmid DNA. The five plus the vector were digested, separated and transferred as in Fig. 2. Hyeach of the five recombinant plasmids labeled with ?P by nick translation.

restriction fragments of TVMV cDNA. The experiment also demonstrates that the inserts do not contain pBR322 sequences. If they did, the inserts would have hybridized to all of the probes (Fig. 3), since each probe contained labeled pBR322 sequences. Arrangement of TVMV cDNA restriction fragments. In order to help establish the position and orientation of the cloned DNA inserts, a restriction map of TVMV cDNA was constructed. This was accomplished by performing double digestions of the TVMV cDNA with seven restriction endonucleases. A portion of the results are shown in Fig. 1C. It can be seen, for example, that the largest EcoRI fragment (3200 bp) was cut by HindIII, BamHI, PstI,

and SalI, but not by PvuII or KpnI (digestion with XbaI was incomplete). Using data of this type, it was possible to arrange each set of restriction fragments end to end in an order which was consistent with all of the observed cutting sites. A map showing the deduced order and orientation of these restriction fragments is presented in Fig. 4A. H&ridization of recombinant plasmids to cDNA restriction fragments. From the size of restriction fragments derived from cDNA (Figs. 1 and 4A) and the size of inserts of recombinant plasmids (Fig. 2A), it appeared that each plasmid represented a cloned copy of one of the internal cDNA fragments of Hind111 or PstI (pTV-Hl contained the Hind111 fragment of 3000;

CLONING

8

A. 5’ F

1100

I,

kb

Hind 1850

,,

Pvu 1425

I

,( 720

,,

I,

TVMV

6

/

1000

OF

Pvu 2275

,,

Barn 2100

Born

217

RNA

4

2

Hind 3000

Pvu 1700

-4 3’

II

,,

Hhd 1950

Pvu 4200

5000

Born

Sol 8000 Kpn

1900

,,

Kpn 2275

,,

Eco

2400

,,

3000 ,

1000

,,

Eco 1350

,,

660

4 Sal 1750

Kpn 4350

,,

4 Pst 1650

I I

I

Eco 3200

PH 1’

IHH ,,

S

EP I

PI

6. HE V -+---i-r+

K BH

H4

H

H3

E

VK

H2

HE

KVE I -

B

P

H f

HI

FIG. 4. Map of cDNA fragments and recombinant plasmids. (A) TVMV cDNA was digested with seven restriction endonueleases, and the order of each set of fragments is shown. The size of each fragment (in bp) is indicated. A scale (in kb), equal to the approximate length of TVMV RNA, is shown at the top. (B) The insert portions of plasmids pTV-Pl, -Hl, -H2, -H3, and -H4 are shown in their proper alignment along the genome. The symbols refer to restriction sites: H, HindHI; S, SalI; V, PvuII; K, KpnI; B, BarnHI; P, PstI; E, EcoRI.

pTV-H2,1850; pTV-H3,720; pTV-H4,llOO; and pTV-Pl, the Pat1 fragment of 1850). Figure 4B shows a map locating the inserts of these five plasmids with respect to each other and to the restriction fragments of cDNA (Fig. 4A). If this correspondence is correct, the plasmid inserts should hybridize to the appropriate restriction fragments of cDNA. To test this, restriction nuclease digests of nonradioactive cDNA were separated by electrophoresis, transferred to nitrocellulose, and hybridized to each of the nick-translated plasmids (Fig. 5). pTV-Hl hybridized primarily to the region of the Hind111 fragment of 3000 bp (Fig. 5A, lane Hl), while pTV-HZ hybridized primarily to the region of the Hind111 fragment of 1850 bp (lane H2), pTV-H3, 720 (lane H3), and pTV-H4, 1100 (lane H4). (It must be kept in mind that the material on the gels was cDNA of heterogeneous length, and that not all

the sequences hybridizing to a given plasmid were the size of the full-length restriction fragment.) These results are in agreement with the assignments made previously on the basis of size alone. A similar result is shown with PstI-digested cDNA (Fig. 5B). Plasmid pTV-PI hybridized strongly to the PstI restriction fragment of 1850 bp (lane Pl) as did pTVHl (lane Hl), but none of the other plasmids hybridized to this restriction fragment. Yet pTV-H2, -H3, and -H4 did hybridize to cDNA of increasingly larger size, in agreement with their proposed order (Fig. 4B). Another type of information was obtained from this experiment. Digestions of cDNA were also made with EcoRI, PvuII, BamHI, and KpnI (Figs. 5C-F). If one assumes the order of the five cloned DNA fragments shown in Fig. 4B to be correct, then one can determine the order of each

T

T

--

PI

PI

HI

HI

H2

PROBE

H2

PROBE

-

-

H3

H3

H4

H4

E.

B. T

T

PI

PI

FIG. 5. Restriction fragments of cDNA hybridized with 32P-labeled recombinant 2. In one case, the cDNA was labeled with “P (A-F, lanes T). In all the other recombinant plasmid was performed as in Fig. 3. Digestion was with: A, HindHI;

Pvu 1700Pvu 1425-

Pvu 2275-

Pvu 4200-

D.

Hind 720-

Hind I IOO-1

Hind 1950Hind 185Oj

Hind 3000--/

A.

HI

-

-

H3

H3

-

H2

PROBE

H2

H4

H4

3200-

860-

1350-

E

EC0 550-

Eco

Eco

Eco 2400-

Eco

C. T

PI

HI

HI

-

-

-

-

-

H2

H3

H3

H4

H4

as in Fig. 32P-labeled

PROBE

H2

PROBE

plasmids. TVMV cDNA was digested, separated, and transferred cases, cDNA was nonradioactive, and hybridization with each B, P&I; C, EcoRI; D, PwuII; E, BarnHI; F, KpnI.

HI

PROBE

E

z

X

CLONING

OF

set of restriction fragments in a manner completely independent of the double digestion procedure used to derive the map in Fig. 4A. For example, in Fig. 5C (lane T), all five EcoRI restriction fragments of cDNA are shown. pTV-Pl (Lane Pl) hybridized to the EcoRI fragments of 3200 and 550 bp (plus fragments derived from less than full length cDNA molecules). pTV-Hl hybridized with EcoRI fragments of 3200 and 860 bp. pTV-HZ hybridized with the EcoRI fragment of 1350, and both pTVH3 and pTV-H4 hybridized with the EcoRI fragment of 2400. Thus, using the order of plasmid inserts shown in Fig. 4B, one would deduce the order of EcoRI cDNA frag‘ments, with respect to the viral RNA, to be 3’-550, 3200, 860, 1350, 2400-5’. This is exactly the order shown in Fig. 4A, which was derived from independent evidence. Similar information was obtained from digestions of cDNA with the other restriction endonucleases (Figs. 5D-F). As a final test, each of the recombinant plasmids was subjected to restriction analysis (Fig. 4B). In each case, restriction sites were located in positions predicted by the cDNA restriction map. According to this arrangement of the cloned DNA fragments, approximately 8200 of the estimated 10,000 nucleotide residues of the TVMV RNA genome are represented in recombinant plasmids. DISCUSSION

These results demonstrate that the major portion of the TVMV genome has been cloned to produce a set of five recombinant plasmids. Experiments were designed primarily to verify the authenticity of cloned DNA fragments (correspondence to viral RNA sequences) and to determine their order. Evidence for authenticity is as follows. The cloned inserts appear to represent the entire cDNA restriction fragments from which they were assumed to be derived since they contain the original restriction sites (Hind111 or P&I) at each end. Internal deletions are unlikely since cDNA fragments and cloned inserts, run side by side on agarose gels, were of identical size. Also, at least five isolates of each plasmid

TVMV

RNA

219

type were obtained after transformation of E. coli, and all contained inserts of identical size. Replacement of TVMV RNA sequences with pBR322 sequences within the insert is also unlikely since probes containing pBR322 sequences failed to hybridize to cloned inserts (Fig. 3). Finally, all of the plasmid inserts derived from HindIII digestion were unique, based on lack of cross-hybridization (Fig. 3). Evidence for the order and orientation of cloned DNA inserts shown in Fig. 4B was obtained from several types of experiments. pTV-Pl and pTV-Hl were the only plasmids to contain sequences near the Zterminus of TVMV RNA (Fig. 2B), and these cross-hybridized (Fig. 3). Digestion of cDNA with several restriction nucleases produced sets of fragments each of which added up to nearly 10 kb (Fig. 1B). Digestion with other restriction endonucleases indicated which additional sites were present in each of these fragments (Fig. 1C); this information could be used to limit the number of possible arrangements of these fragments, since all sites had to be accommodated. The map in Fig. 4A represents, to our knowledge, the only order which is internally consistent. Hybridization of recombinant plasmids to restriction fragments of cDNA independently confirmed this order (Fig. 5). A third independent piece of evidence was the restriction map of each recombinant plasmid (Fig. 4B), which matched the sites in cDNA. On the basis of the restriction map for the viral cDNA, it is clear why only four types of plasmids were obtained using HindHI; only the four internal Hind111 fragments of cDNA contain Hind111 sites at each end. The same reasoning explains the occurrence of only a single type of plasmid constructed with PstI. Indirect evidence suggests that there are additional PstI restriction sites near the 3’ terminus of the viral RNA, but recombinant plasmids resulting from these were apparently overlooked, presumably due to weak reactions during colony hybridization. It is also not impossible that there are very small Hind111 fragments between the major fragments which were overlooked for the same reason.

220

HELLMANN ACKNOWLEDGMENTS

The authors would like to acknowledge the generous assistance of Dr. Frank Gannon, in whose laboratory initial experiments were performed, and are indebted to Drs. G. Lomonossoff, S. Bhairi, K. Nelson, and R. Dickson for their helpful advice. This work was supported by Grant 59-2213-1-l-600-0 from the USDA/SEA and Grant 4E021 from the University of Kentucky Tobacco and Health Research Institute. REFERENCES BIRNBOIM, H. C., and DOLY, J. (1979). A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucl. Acio!s Res. 7, 1513-1523. BOTHWELL, A. L. M., PASKIND, M., RETH, M., IMANISHI-KARI, T., RAJEWSKY, K., and BALTIMORE, D. (1981). Heavy chain variable region contribution to the NPb family of antibodies: Somatic mutation evident in a y2a variable region. Cell 24. 625-637. CRAIG, E. A., and MCCARTHY, B. J. (1980). Four Dro sophila heat shock genes at 67B: characterization of recombinant plasmids. Nucl. Acids Res. 8,44414457. DEMARTYNOFF, G., PAYS, E., and VASSART, G. (1980). Synthesis of a full length DNA complementary to thyroglobulin 33 S messenger RNA. B&hem. Bi+ phys. Res. Commun. 93. 645-653. DOUGHERTY, W. G., and HIEBERT, E. (1980a). Translation of potyvirus RNA in a rabbit reticulocyte lysate: Reaction conditions and identification of capsid protein as one of the products of in vitro translation of tobacco etch and pepper mottle viral RNAs. Virology 101, 466-474. DOUGHERTY, W. G., and HIEBERT, E. (1980b). Translation of potyvirus RNA in a rabbit reticulocyte lysate: Identification of nuclear inclusion proteins as products of tobacco etch virus RNA translation and cylindrical inclusion proteins as a product of the potyvirus genome. Virology 104, 174-182. DOUGHERTY, W. G., and HIEBERT, E. (198Oc). Translation of potyvirus RNA in a rabbit reticulocyte lysate: Cell-free translation strategy and a genetic map of the potyviral genome. firoh 104, 183194. EDWARDSON, J. R. (1974). Some properties of the potato virus Y group. Florida Agric Stn. Monog. Ser. 4, 398. GOODMAN, H. M., and MACDONALD, R. J. (1979). Cloning of hormone genes from a mixture of cDNA molecules. In “Methods in Enzymology” (R. Wu, ed.), Vol. 68, pp. 75-80. Academic Press, New York. GRUNSTEIN, M., and HOGNESS, D. S. (1975). Colony hybridization: A method for the isolation of cloned DNAs that contain a specific gene. Proc. Nat. Acad Sci USA 72,3961-3965. HARI, V., SIEGEL, A., ROZEK, C., and TIMBERLAKE,

ET

AL.

W. E. (1979). The RNA of tobacco etch virus contains poly(A). Virology 92, 568-571. HELLMANN, G. M., SHAW, J. G., LESNAW, J. A., CHU, L.-Y., PIRONE, T. P., and RHOADS, R. E. (1980). Cellfree translation of tobacco vein mottling virus RNA. Virology 166,207-216. HELLMANN, G. M., THORNBURY, D. W., HIEBERT, E., SHAW, J. G., PIRONE, T. P., and RHOADS. R. E. (1983). Cell-free translation of tobacco vein mottling virus RNA. II. Immunoprecipitation of products by antisera to cylindrical inclusion, nuclear inclusion and helper component proteins. virology 124,434-444. MANIATIS, T., JEFFREY, A., and VAN DESANDE, H. (1975). Chain length determination of small double- and single-stranded DNA molecules by polyacrylamide gel electrophoresis. Biochemistry 14, 3787-3794. MANIATIS, T., JEFFREY, A., and KLEID, D. G. (1975). Nucleotide sequence of the rightward operator of phage X. Proc Nat. Acad Sci USA 72,118&1188. MCDONELL, M. W., SIMON, M. N., and STUDIER, F. W. (1977). Analysis of restriction fragments of T7 DNA and determination of molecular weight by electrophoresis in neutral and alkaline gels. J. Mol. BioL 110, 119-146. MCMASTER, G. K., and CARMICHAEL, G. G. (1977). Analysis of single- and double-stranded nucleic acids on polyacrylamide and agarose gels by using glyoxal and acridine orange. Proc. Nat. Acad. Sci USA 74, 4835-4838. MORRISON, D. A. (1979). Transformation and preservation of competent bacterial cells by freezing. In “Methods in Enzymology” (R. Wu, ed.), Vol. 68, pp. 326-331. Academic Press, New York. O’FARRELL, P. H., POLISKY, B., and GELFAND, D. H. (1978). Regulated expression by readthrough translation from a plasmid-encoded P-galactosidase. J. Bacterial. 134, 645-654. RHOADS, R. E., and HELLMANN, G. M. (1978). Chromatography of ovalbumin messenger ribonucleic acid on complementary deoxyribonucleic acid-cellulose. J. Biol. C&m. 253, 1687-1693. SIPPEL, A. E., LAND, H., LINDENMAIER, W., NGUYENHuu, M. C., WURTZ, T., TIMMIS, K. N., GIESECKE, K., and SCHUTZ, G. (1978). Cloning of chicken lysozyme structural gene sequences synthesized in vitro. Nucl. Acids Res. 5, 3275-3294. SOUTHERN, E. M. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mel; BioL 98, 503-517. WIGLER, M., SILVERSTEIN, S., LEE, L.-S., PELLICER, A., CHENG, Y.-C., and AXEL, R. (1977). Transfer of purified herpes virus thymidine kinase gene to cultured mouse cells. Cell 11, 223-232. ZAIN, B. S., and ROBERTS, R. J. (1979). Sequences from the beginning of the fiber messenger RNA of adenovirus-2. J. Mol. Bid 131, 341-352.