Cauliflowermosaicvirus VI symptomatic · 2005-04-23 · 734 Biochemistry: Baughmanet al. Asu 11 (7730) Asu11 Kpn I (Asp718) FIG. 1. Map of plasmids in Atl70 (pGB220) and Atll5...

Proc. Nati. Acad. Sci. USAVol.85, pp. 733-737, February 1988Biochemistry

Cauliflower mosaic virus gene VI produces a symptomaticphenotype in transgenic tobacco plants

(plant DNA virus/virus symptomology)

GAIL A. BAUGHMAN, JERRY D. JACOBS*, AND STEPHEN H. HOWELLtBiology Department C016, University of California at San Diego, La Jolla, CA 92093

Communicated by Myron K. Brakke, October 5, 1987 (received for review July 29, 1987)

ABSTRACT Gene VI of the cauliflower mosaic virus(CaMV) genome encodes a protein (P66) in virus-infectedplants that accumulates in cytoplasmic inclusion bodies. Whena segment of the CaMV genome bearing gene VI is transferredto tobacco plants by the Agrobacterium tumefaciens Ti plas-mid, the resulting transgenic plants display viral-like symp-toms. Symptoms produced by the DNA from two differentviral isolates (CaMV Cabb B-JI and CM1841) were distinct-symptoms from the first were mosaic-like, whereas the othercaused uniform bleaching of leaves. That gene VI was respon-sible for the symptomatic phenotype was demonstrated byshowing that symptom production was blocked by deletionsand by a frame-shifting linker mutation in gene VI. Further-more, in primary transformants, there was a strict correlationbetween the appearance of symptoms and the presence of geneVI product, P66, detected by immunoblots. Hence, a proteinencoded by the CaMV genome produces viral-like symptomsin transgenic tobacco plants.

The pathogenic effects of viruses on plant growth, develop-ment, and symptom production are described in rich detail inthe plant literature; however, the mechanisms by whichplant viruses cause disease are poorly understood at amolecular level (for review, see ref. 1). Symptoms producedby plant viruses are very diverse, but a common symptom insystemic infections is a mosaic leaf pattern that results fromthe yellowing (chlorosis, bleaching, or clearing) of vascularchannels that delimit "green islands" of tissue in the leaf.What causes the yellowing and the intensity of symptoms

is a matter of considerable interest. In plants infected withtobacco mosaic virus (TMV) (2) and cucumber mosaic virus(CMV) (3), virus is usually concentrated in the chloroticregions, whereas the dark green islands are relatively virus-free. Among different isolates of the same virus, thereappears to be direct correlation between the severity ofsymptoms (chlorosis) and the extent of TMV accumulation(4). Attenuated TMV strains, such as the L11A strain, whichdiffers from a normal strain by a single amino acid change inthe 183-kDa protein (5), accumulate to a lesser extent thanthe wild-type counterpart. However, among different vi-ruses, the severity of symptoms does not necessarily relateto the extent of virus accumulation (6). This may be due tothe fact that different viruses may have different means bywhich they produce symptoms.Most isolates of cauliflower mosaic virus (CaMV) produce

typical mosaic symptoms in the leaves of systemically in-fected host plants (7). At the cellular level, a prominentcytopathological effect of CaMV infection is the appearanceof viral inclusion bodies or viroplasms in the cytoplasm ofinfected cells. The most abundant viral-encoded protein, P66(8-11), in infected cells is a major protein of the matrix of the

inclusion body. The major inclusion body matrix protein isencoded by gene VI in the CaMV genome and is translatedfrom the 19S RNA-one of two major RNAs produced fromthe CaMV genome. Gene VI has been implicated in hostrange control and in symptom production (12). We demon-strate in this paper that transgenic tobacco plants thatexpress gene VI display a symptomatic phenotype charac-teristic of virus-infected plants.

MATERIALS AND METHODSPlasmid Constructions. The basic construct used in these

experiments was a 4840-base-pair (bp) Sal I-Xho I fragmentfrom the CaMV genome linked to a DNA fragment from the3' end of the nopaline synthetase (nos) gene, which providesa poly(A) signal (13). The CaMV DNA fragment (positions4833-1642) was obtained from pLW303, derived from aCaMV Cabb B-JI isolate (14), or pCaMV10, from a CaMVCM1841 isolate (15), and carries from left to right (Fig. 1) thedistal portion of open reading frame (ORF) V (V'), the 19Spromoter, ORF VI, the 35S promoter, ORF VII, ORF I, andthe proximal portion ofORF II (I'). To link the CaMV DNAto the nos 3' segment, the Sal I-Xho I fragment of the viralgenome was inserted into the Sal I site of pUC18Xnos 3',which carries the 1028-bp nos 3' fragment in a derivative ofpUC18 that has an Xho I linker inserted in the HindIII site ofthe polylinker.An Xho I-Kpn I fragment containing the linked CaMV

DNA-nos 3' segment was inserted into the Sal I-Kpn I sitesof Bin 19, an Agrobacterium tumefaciens binary Ti-plasmidvector (16), to create pCaMV201 (Cabb B-JI isolate) andpGB220 (CaMV CM1841 isolate). These plasmids weretransferred by conjugation to A. tumefaciens AtilO (LBA-4404 derivative) to yield At115 and At170, respectively (Fig.1).By making the appropriate deletions, we subdivided the

CaMV DNA insert into segments carrying either the 19S orthe truncated 35S transcription units. Deletions were madein pGB212, a progenitor plasmid (with pUC18 backbone) ofpGB220. A Sal I linker was inserted at the EcoRV site(CaMV map position 7342) or the Ava I site (position 6688) ofthe CaMV DNA insert, and the Sal I fragment containing aportion of the 19S transcription unit was deleted; then theremainder of the CaMV-nos 3' segment was inserted into Bin19 and transferred by conjugation into A. tumefaciens tocreate At171 and At174, respectively. These two plasmidsdiffer by length of the region upstream from the 35S pro-moter. In another construct, which gave rise to At172,nearly all of the truncated 35S transcription unit was deletedfrom the Asu II site (CaMV map position 7730), 295 bp

Abbreviations: CaMV, cauliflower mosaic virus; TMV, tobaccomosaic virus; CMV, cucumber mosaic virus; P66, CaMV gene VIproduct; ORF, open reading frame.*Present address: Zoology Department, University of Washington,Seattle, WA 98195.tTo whom reprint requests should be addressed.

733

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Mar

ch 1

5, 2

020

734 Biochemistry: Baughman et al.

Asu 11(7730) Asu 11

Kpn I(Asp 718)

FIG. 1. Map of plasmids in Atl70 (pGB220) and Atll5(pCaMV201) and plasmids derived from pGB220. Plasmids in Atl7Oand Atll5 were produced by inserting the 4.8-kilobase (kb) SalI-Xho I CaMV fragment (from CaMV isolates CM1841 and CabbB-JI, respectively) linked to the 1.0-kb nos 3' fragment into the A.tumefaciens binary Ti-plasmid vector, Bin 19 (16). Atl71, At172,and At173 have plasmids with deletions (A) as indicated. At173 hasan 8-bp Sal I linker insert at the Bal I site at position 5940 in theCaMV genome segment. RB, right border; LB, left border.

downstream from the start of transcription, to the Asu II sitein the nos 3' segment. A frame-shifting linker mutation inCaMV gene VI was created by inserting an 8-bp Sal I linkerat the Bal I site (CaMV map position 5940) in pGB212. Thisconstruct was inserted into Bin 19 and transferred to A.tumefaciens to create At173.DNA Extraction and Blots. DNA was extracted from

tobacco leaves by using a cetyltrimethylammonium bromideextraction procedure of Murray and Thompson (17). TheCsCI/ethidium bromide gradient-purified DNA was digestedat 0.25 ,g of DNA per ,ul with the indicated restrictionenzymes, precipitated with ethanol following digestion, sub-jected to electrophoresis on a 0.9%6 agarose gel, and trans-ferred to nitrocellulose paper according to standard proce-dures (18).

Protein Extraction and Immunoblots. Protein extracts wereprepared from young Nicotiana tabacum transgenic plantsby freeze-grinding leaves in liquid nitrogen, extracting in 50mM Tris-HCI, pH 9/3% NaDodSO4/5% mercaptoethanol(0.66 g of fresh leaf weight per 1 ml), and heating in a boilingwater bath for 2 min. Total leaf proteins were subjected toelectrophoresis on a 10% NaDodSO4/polyacrylamide geland transferred electrophoretically to nitrocellulose filters(19). P66 was detected by an indirect antibody-binding pro-cedure. The filters were incubated with P66 rabbit antiserum(antiserum 434, 1:200 dilution), and antibody-P66 complexeswere detected by secondary incubation with goat anti-rabbitIgG horseradish peroxidase conjugate (Bio-Rad) and visual-ized with a peroxidase color reagent (4-chloro-1-naphthol).Antiserum reactive to P66 was generated by injecting

rabbits with a synthetic peptide representing 10 amino acidsat the carboxyl terminus of the protein encoded by CaMVgene VI. The peptide, Tyr-Val-Pro-Thr-Thr-Ser-Ser-Lys-Gln-Val-Asp, was synthesized by the Merrifield method (20)and bore a tyrosine residue (not present in P66) at its amino

terminus so that it could be conjugated to carrier bovineserum albumin according to Walter et al. (21).

RESULTS

Transgenic tobacco plants in the Atl15, Atl70, or At172plant series (transgenic plants transformed by Atll5, Atl70,or Atl72, respectively) bearing a 4.8-kb segment of theCaMV genome (in AtllS and At170) or a 2.2-kb subfragment(in At172, Fig. 1) displayed an interesting phenotype. Theleaves of most plants showed typical symptoms of virusinfection of host plants. The leaves either had a blotchy(mosaic-like) appearance (Fig. 2A) or were uniformly lightgreen (Fig. 2B) when compared to asymptomatic plants (Fig.2C). We have tested the same DNA segment from twodifferent CaMV isolates and have found that the DNA fromthe CaMV Cabb B-JI isolate (Atll5 series) usually gave ablotchy phenotype, whereas the CaMV CM1841 isolate(At170 and At172 series) produced a more uniform lightgreen background.The 4.8-kb CaMV DNA segment introduced into these

plants is a complex region of the viral genome that waslinked to a 3' fragment from the nos gene (Fig. 1). The viralDNA segment contained two transcription units-the 19SRNA unit, which encodes ORF VI, and a truncated 35SRNA unit, which included ORFs VII and I. To find out whatviral determinants were responsible for the symptomaticphenotype, we produced various deletions in the viral DNAsegment and reintroduced the constructs into plants bymeans of the Ti plasmid. We found that all transformedplants were asymptomatic (Table 1) when we deleted part(At174 series) or nearly all (Atl71 series) of the 19S RNAtranscription unit, leaving the 35S RNA transcription unitintact with about 750 bases upstream from the start of 35SRNA transcription in the case of the At174 series and onlyabout 90 bases upstream in the Atl71 constructs. However,deletion of the 35S RNA transcription unit (At172 series,Fig. 1) had no effect on symptom production (Fig. 2B, Table1). These results pointed to gene VI in the 19S transcriptionunit as a possible determinant of symptom production. Toconfirm this, an 8-bp frame-shifting linker mutation (Sal Ilinker) was inserted near the start of the gene VI codingregion (166 bp downstream from the start of translation, BalI site, Fig. 1). No transgenic plants with a symptomaticphenotype were produced from the construct with a linkermutation in gene VI (At173 series, Fig. 2C, Table 1).Transgenic plantlets were routinely selected on antibiotic

medium, and explanted tissues from selected plants wererescreened on the same medium. However, we were con-cerned whether asymptomatic plants, especially those bear-ing the altered gene VI constructs, were truly transformed.Therefore, DNA was isolated from representative plants, cutwith Asp 718 and Sal I (to excise the CaMV-nos 3' insert),and analyzed by DNA blotting procedures. Plant Atl15j,which had a CaMV DNA insert (5.8 kb including the nos 3'segment) from the Cabb B-JI isolate and exhibited a"6severe" mosaic symptom phenotype, had about two tofour copies (from copy number reconstructions, not shown)of the CaMV DNA insert (Fig. 3). Plants At170-3 andAtl70-4, which had DNA from the CaMV CM1841 isolateand a less severe, uniform light green phenotype (as in Fig.2B), had about the same number of insert copies as plantAtll5j. Plants from the At172 series, At172-10, -11, and -25,were all symptomatic and each had different copy numbersof the CaMV inserts of the expected size (3.2 kb). (PlantAt172-11 had additional and presumably rearranged copiesof the insert at about 4.8 kb.) Plant At172-28 was asympto-matic and lacked a detectable CaMV insert. However, theplant was kanamycin resistant and, therefore, may have lostthe CaMV DNA insert by rearrangement. At173-13 (gene VI

Proc. Natl. Acad. Sci. USA 85 (1988)

Dow

nloa

ded

by g

uest

on

Mar

ch 1

5, 2

020

Proc. Natl. Acad. Sci. USA 85 (1988)

LI

FIG. 2. Transgenic tobacco plants (N. tabacum var. Wi-38) bearing the Sal I-Xho I fragment from the CaMV genome shown in Fig. 1. (A)Atll5j (primary transformant from Atll5 inoculation) shows blotchy (mosaic-like) symptoms after 2 months of growth. (B) At172-B63 (primarytransformant from At172 inoculation) shows uniform bleaching (chlorosis). (C) At173-B67 (primary transformant from At173 inoculation) isasymptomatic after 1.5 months of growth.

Biochemistry: Baughman et al. 735

Dow

nloa

ded

by g

uest

on

Mar

ch 1

5, 2

020

736 Biochemistry: Baughman et al.

Table 1. Plants with symptomatic phenotype express P66

Transgenic Plants with Plants expressing P66*plants symptoms With symptoms Without symptoms

Atll5 9/10 6/8t 0/1At170 7/12 5/5 0/3At172 11/17 7/7 0/5At171 0/8 NA 0/3At173 0/16 NA 0/8At174 0/14 NA 0/11NA, not applicable.

*The presence of P66 in all plants was determined with immunoblotsas shown for representative plants in Fig. 4.tTwo plants not expressing P66 were older plants not tested until 3months' growth.

linker mutation) and At174-22 and -24 (gene VI deletion)were asymptomatic but both bore CaMV inserts of theexpected size (4.7 kb for the At173 series and 4.0 kb for theAt174 series). At173 series plants also had a 1.1-kb CaMVDNA fragment that was not seen on these gels. Hence,representative plants of the At173 and At174 series con-tained CaMV inserts and were most likely asymptomaticbecause of the specific mutations introduced into gene VI.

Transformants in the Atl70 and At172 series (intact geneVI) differed in the severity of symptoms or in the penetranceof the symptomatic phenotype. Some were totally asympto-matic, which in the case of At172-28 was apparently due toa loss of the viral DNA insert. Other differences in theseverity of symptoms may be due to variation in expressionof the CaMV DNA segment associated with different num-bers of copies or sites of integration in the genome. In anycase, it was of interest to determine whether the expressionof symptoms correlated with the accumulation of P66, thegene VI protein, in representative plants. To detect P66, weused an immunoblot procedure and a polyclonal antiserumraised against a synthetic peptide corresponding to thecarboxyl-terminal 10 amino acids in the gene VI product.The antiserum recognized a single major band correspond-

ing in size to P66 (and a series ofminor bands seen on heavierloadings) in extracts from CaMV-infected turnip leaves and

m0M CM cm~ N._X t N N _- N N

CO 6 6 CjCt c j C4 CJ 4 4,- r~- _~- _~.. _~ _~. _~ _ 6

23.1 _

9.4 _6.5 _

4.3-

2.3 _-2.0 _

FIG. 3. CaMV DNA in transgenic tobacco plants. Total leafDNA (10 ,ug per lane) was cut with Asp 718 and Sal I to excise theCaMV-nos 3' insert, subjected to electrophoresis on a O.9o agarosegel, and hybridized to 32P-labeled pGB212 insert containing the SalI-Xho I CaMV-nos 3' DNA fragment shown in Fig. 1. The positionsand sizes (in kb) of HindIlI bacteriophage A DNA fragments areshown on the left.

(n

A-

U

._

CO) DCOLO

C'I) Cf) C')

- - -

49 <: 4

co cc- CI) 0N- N- N

0

CM _ ON N P- C0 N _ _

0 0 0 0 10 10 If) If) it) 10 0

N N. F. rN____ _ _ _ _ _ _C____ _ _ _ _ _ _~~~~<<<< 4t iz <1: zr <s <t V~~C

FIG. 4. Appearance of P66 in transgenic tobacco plants. Immu-noblots of total NaDodSO4-solubilized leaf proteins (20 ,ug ofprotein per lane) from transgenic plants in the At173, Atl70, andAtll5 series were subjected to electrophoresis on 10% NaDod-S04/polyacrylamide gels, transferred to nitrocellulose filters, andincubated with an antiserum to P66. P66 (arrowhead) was detected byan indirect antibody technique using goat anti-rabbit IgG conjugatedto horseradish peroxidase. P66 from CaMV-infected turnip leaves (11Lg of protein per lane) was used as a standard for comparison.

leaves of transgenic tobacco plants (Fig. 4). P66 was ob-served in the extracts from most, but not all, of the AT115and Atl70 series plants that had an intact 19S RNA tran-scription unit and gene VI. Nonetheless, there was nearly aperfect correlation between the appearance of symptomsand the presence of detectable levels of P66 protein (Table 1).No P66 (or any other cross-reacting protein) was found inextracts from the At173 series plants that have a frame-shifting linker mutation in gene VI (Fig. 4 and Table 1). Theamount of reacting protein was much higher in infectedturnip plants than in transgenic tobacco plants. The turnipextract was diluted by a factor of 20 more than extracts fromtobacco plants to obtain bands of equal intensity. Hence, weconclude that the product of gene VI, P66, is responsible forthe production of the symptomatic phenotype in transgenictobacco plants.

DISCUSSIONWe have shown that gene VI from the CaMV genomeproduces a symptomatic phenotype when introduced intotransgenic tobacco plants. Gene VI in the free, infectiousvirus has been implicated in symptom expression in hostplants. Daubert et al. (22) demonstrated that the introductionof a non-frame-shifting linker mutation in gene VI alteredsymptom expression. Most frame-shifting linker or deletionmutations in gene VI are lethal, which demonstrates theindispensability of gene VI to the propagation of the virus(22-24). Other experiments of Daubert et al. (12) involvingthe production of chimeric CaMV genomes between viralisolates that produce different symptoms on various hostplants also implicate gene VI in symptom production.

In our experiments, gene VI was shown to be responsiblefor a symptomatic phenotype in transgenic plants by thefollowing criteria: (i) a CaMV DNA fragment containingonly the 19S RNA transcription unit, which includes geneVI, produced symptoms; (ii) deletions or a frame-shiftinglinker mutation in gene VI blocked symptom production; (iii)production of symptoms was tightly correlated with the

Proc. Natl. Acad Sci. USA 85 (1988)

Dow

nloa

ded

by g

uest

on

Mar

ch 1

5, 2

020

Proc. Natl. Acad. Sci. USA 85 (1988) 737

appearance of P66, the gene VI product. Not all transgenicplants with an intact 19S RNA transcription unit weresymptomatic; in fact, 27 of 39 bona fide transformants(plants from Atll5, At170, and At172 series) showed asymptomatic phenotype. The asymptomatic plants mightarise for a variety of reasons, but none expressed detectablelevels of P66.The blotched or flecked (mosaic-like) phenotype observed

in leaves of many of the AtliS series transgenic plants wasunexpected and interesting. In virus-infected plants, theusual explanation given for the pattern of symptoms relatesto the tissue distribution of the virus (1); however, in thesetransgenic plants all cells in the leaf should have the samepotential to produce the virus-encoded protein. Transgenicplants in the At170 and At172 series were more uniformlybleached or chlorotic. The difference between the Atil5 andthe At170 series is the viral isolate from which the DNAsegment containing gene VI was derived. These two CaMVisolates have been shown to produce distinguishable symp-toms in host turnip plants, Brassica campestris var. rapa(12). The tobacco plants used in this study are not naturalhosts for CaMV; however, tobacco plants are easily trans-formed and ideal for transgenic plant studies.How gene VI produces a symptomatic phenotype is un-

known. The gene VI product, P66, is usually regarded as themajor inclusion body matrix protein and may play an impor-tant role in the multiplication of the virus. P66 is the mostabundant viral gene product accumulated in infected turnipplants (8-11), but even in infected turnip plants, P66 mayconstitute only a few percent of the total protein. In infectedturnips there may be one or a few inclusion bodies per cell(7), but the inclusion bodies do not appear, morphologically,to obstruct any cellular processes or structures. In thesymptomatic transgenic tobacco plants described here, weestimate that the amount of P66 produced is less by a factorof 20 than that in the turnip plant.

We thank R. F. Doolittle for synthesizing the peptide and advisingus in producing the P66 antiserum and David W. Ow for his help anduseful suggestions. This work was supported by the U.S. Depart-ment of Agriculture-Science and Education Administration Com-petitive Grants Program (85-CRCR-1-1831) and by a National Sci-ence Foundation postdoctoral fellowship in plant biology to G.A.B.

1. Fraser, R. S. S. (1987) Biochemistry of Virus-Infected Plants(Research Studies, Letchworth, U.K.).

2. Atkinson, P. H. & Matthews, R. E. F. (1970) Virology 40,344-356.

3. Loebenstein, G., Cohen, J., Shabtai, S., Coutts, R. H. A. &Wood, K. R. (1977) Virology 81, 117-125.

4. Fraser, R. S. S., Gerwitz, A. & Morris, G. E. L. (1986) Phys-iol. Mol. Plant Pathol. 29, 239-250.

5. Nishiguchi, M., Kikuchi, S., Kiho, Y., Ohno, T., Meshi, T. &Okada, Y. (1985) Nucleic Acids Res. 13, 5585-5590.

6. Zaitlin, M. (1979) in Plant Disease, an Advanced Treatise, eds.Horsfall, J. G. & Cowling, E. B. (Academic, New York), pp.257-271.

7. Shepherd, R. J. (1981) Descriptions of Plant Viruses (Holy-well, Oxford), No. 243.

8. Odell, J. T. & Howell, S. H. (1980) Virology 102, 349-359.9. Al Ani, R., Pfeiffer, P., Whitechurch, O., Lesot, A., Lebeu-

rier, G. & Hirth, L. (1980) Ann. Virol. (Inst. Pasteur) 131E,33-53.

10. Covey, S. N. & Hull, R. (1981) Virology 111, 463-474.11. Xiong, C., Muller, S., Lebeurier, G. & Hirth, L. (1982) EMBO

J. 1, 971-976.12. Daubert, S. D., Schoelz, J., Debao, L. & Shepherd, R. J.

(1984) J. Mol. Appl. Genet. 2, 537-547.13. Depicker, A., Stachel, S., Dhaese, P., Zambryski, P. &

Goodman, H. (1982) J. Mol. Appl. Genet. 1, 561-573.14. Walden, R. M. & Howell, S. H. (1982) Mol. Appl. Genet. 1,

447-456.15. Gardner, R. C., Howarth, A. J., Hahn, P., Brown-Luedi, M.,

Shepherd, R. J. & Messing, J. (1981) Nucleic Acids Res. 9,2871-2887.

16. Bevan, M. (1984) Nucleic Acids Res. 12, 8711-8721.17. Murray, M. G. & Thompson, W. F. (1980) Nucleic Acids Res.

8, 4321-4325.18. Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular

Cloning: A Laboratory Manual (Cold Spring Harbor Labora-tory, Cold Spring Harbor, NY).

19. Towbin, H., Staehlin, T. & Gordon, J. (1979) Proc. Natl.Acad. Sci. USA 76, 4350-4354.

20. Stewart, J. M. & Young, J. D. (1969) Solid Phase PeptideSynthesis (Freeman, San Francisco).

21. Walter, G., Scheidtmann, K.-H., Carbone, A., Laudano, A. P.& Doolittle, R. F. (1980) Proc. Natl. Acad. Sci. USA 77,5197-5200.

22. Daubert, S., Shepherd, R. J. & Gardner, R. C. (1983) Gene 25,201-208.

23. Howell, S. H., Walker, L. L. & Walden, R. M. (1981) Nature(London) 293, 483-486.

24. Dixon, L. K., Koenig, I. & Hohn, T. (1983) Gene 25, 189-199.

Biochemistry: Baughman et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 1

5, 2

020