VIROLOGY

190,

506-509

A Symptom

(1992)

Variant of Beet Curly Top Geminivirus

lnnes

Institute,

John

lnnes

Centre Received

for Plant March

by Mutation

of Open Reading

Frame C4

AND JONATHAN R. LATHAM

JOHN STANLEY’ John

Produced

Science 24,

Research, 1992;

accepted

Colney June

Lane,

Norwich

NR4

7lJH,

United

Kingdom

10, 1992

Two beet curly top virus (BCTV) mutants have been constructed in vitro that contain G-to-T transversions either at nucleotide 2682 or at nucleotide 2802 within the overlapping open reading frames (ORFs) Cl and C4. The mutations introduce termination codons in ORF C4 without affecting the amino acids encoded by ORF Cl. When agroinoculated into Nicotiana benthamiana the mutants caused stunting and yellowing of the plant and downward leaf curl but not the vein swelling and upward leaf curl symptoms that are characteristic of wild-type BCTV infection in this host. Levels of viral single- and double-stranded DNA forms were similar in mutant and wild-type infections. Symptoms induced by one such mutant in Nicotiana clevelandii and Datura stramonium were less severe than those in wild-type infections and were again qualitatively distinct. The mutants caused symptomless infections in Beta vulgaris, contrasting with stunting, severe leaf curl, and vein swelling symptoms associated with wild-type infection of this host. The levels of mutant DNA in newly expanding asymptomatic leaves frequently reached those of wild-type virus in leaves showing severe symptoms. The results suggest that ORF C4 encodes a protein that is a major determinant of pathogenesis that might affect the hyperplastic response of the host to BCTV infection. o 1992 Academic PWSS. IW.

The majority of geminiviruses that infect dicotyledonous plants have bipartite genomes (usually referred to as DNAs A and B). DNA A of African cassava mosaic virus (ACMV), typical of the bipartite geminiviruses, can autonomously replicate in protoplasts (I) and can be independently encapsidated (2) although it cannot spread efficiently within the plant or induce symptoms in the absence of DNA B. Therefore, DNA B plays a major role in pathogenesis by delivering the virus to the appropriate tissues in which it can replicate. In contrast, beet curly top virus (BCTV) has only a single component genome organized on a similar basis to DNA A (3; Fig. 1). A component equivalent to DNA B is not required for virus spread or symptom induction. This fundamental difference has been correlated with the variation in tissue specificity of these viruses and their unique symptom phenotypes (reviewed in 4). Although a great many BCTV phenotypic variants have been described (5 and references therein), little is known about the pathogenic determinants of the virus. A previous report from this laboratory (6) has demonstrated that intact ORF Cl is essential for viral DNA replication in /Vicotiana tabacum protoplasts, consistent with a role in viral DNA replication shown for analogous genes in ACMV and TGMV (7-9). In contrast to the bipartite geminiviruses, BCTV coat protein is required for systemic infection although a coat protein mutant retained the ability to replicate in Iv. tabacum

’ To whom 0042.6822/92

reprint

requests

should

$5.00

CopyrIght 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

protoplasts (6). In view of the differences that exist between BCTV and other geminiviruses that infect dicotyledonous plants, a more comprehensive study has been initiated with the view to identifying genes and/or gene functions that are unique to this virus. During this investigation mutants have been identified that have a profound effect on BCTV symptom development. The DNA insert of infectious clone pBCT028 (3) derived from a California BCTV isolate, was subcloned at its unique Sstl site (position 2056) in M 13mpl8 to give clone pBCT098. A G-to-T mutation was introduced into this clone at position 2682 by site-directed mutagenesis using the Bio-Rad Muta-Gene kit. This mutation introduced a termination codon in the complementarysense ORF C4 without affecting the amino acid sequence of the overlapping ORF Cl, removed the unique SalI site at position 2679 and created a second SnaBl site at position 2682 (Fig. 1). The Sstl inserts containing the parental (referred to here as wild-type) and mutant BCTV DNAs were cloned as partial repeats into the Agrobacterium tumefaciens binary vector pBinl9 (10) that contained the Sstl (2056)~EcoRI (2646) fragment of the BCTV genome (clone pBin0.2SE) to produce clones pBin1 .2SE and AC4/1, respectively. The clones were mobilized into A. tumefaciens containing the disarmed Ti plasmid PGV3850 (11) and agroinoculated as previously described (6). The mutant was as infectious as the wild-type clone when introduced into Mcotiana benthamiana (14 of 18 inoculated plants showing symptoms with both the mutant and wild-type BCTV over two experiments).

be addressed 506

SHORT

COMMUNICATIONS

1

SC

-

ss

-

2

3

507

4

5

6

7

8

9 1011

12131415

-2.19 - 1.96 1.03 - 0.82

-

FIG. 1. Left, Genome map of BCTV showing the position and orientation of ORFs and the conserved nonanucleotide (black dot) in relation to selected restriction sites. The SalI site in the wild-type clone is replaced with a SnaBl site in mutant AC4/1 and a Mae1 site has been introduced into mutant AC4/2. A mutation in ORF C4 that introduces an Xbal site will be described elsewhere. Right, Southern blot analysis of BCTV DNA forms in extracts of N. benthamiana (lanes 2, 3, and 1 O-l 3) and B. vu@aris (lanes 4-9) agroinoculated with wild-type virus (lanes 2, 4, 10, and 12) and mutant AC4/1 (lanes 3, 5-9, 1 1, and 13). Samples in lanes 5-9 were from five asymptomatic plants chosen at random. Equal amounts of total nucleic acids (2 pg) were loaded in each lane. Samples in lanes 1 O-l 3 were digested with SnaBl and those in lanes 12 and 13 were also treated with mung-bean nuclease to remove single-stranded DNA. Size markers (in kbp) were produced by digestion of clone pBCT028 with Sail (lane 1). EcaRl (lane 14), and SalI and SnaBl (lane 15). The positions of linear (lin), supercoiled (SC), and single-stranded (ss) DNAs and a population of subgenomic DNA forms (sg) are indicated. The blot was probed with labeled full-length BCTV genome insert of clone pBCT028.

The mutant induced stunting and a general yellowing of the plant as well as downward leaf curl more reminiscent of mild ACMV or TGMV than BCTV infection in this host (Fig. 2). In contrast to wild-type symptoms, there was no vein swelling and upward rolling of leaves infected with the mutant. Total nucleic acids were extracted from individual plants as described (12). In general, systemically infected leaves of the same age were sampled although this was not always possible due to the influence of symptom severity on plant growth. Samples were fractionated by agarose gel electrophoresis and depurinated (13), and viral DNA forms were analyzed by Southern blotting using oligolabeled (14) insert of pBCT028 as a BCTV-specific probe (Fig. 1). In all three samples examined, levels of both single- and

FIG. 2. A healthy N. benrhamiana plant with the mutant AC411 (left) and wild-type agroinoculation.

(center) BCTV

and plants infected (right) 18 days after

double-stranded DNAforms of the mutant were similar to those of the wild-type virus (compare lanes 2 and 3). Infection with both wild-type and mutant virus was invariably associated with the production of single- and double-stranded DNA forms of a variety of subgenomic DNAs that have recently been characterized (15, 16). Digestion with SnaBl linearized the wild-type doublestranded DNA (lanes 10 and 12) and produced two fragments of expected sizes 1965 and 1028 bp from the mutant DNA (lanes 11 and 13), verifying that the mutant progeny had retained the mutation. To ensure that the observed phenotype was due solely to a single point-mutation at position 2682, and was not the result of additional modifications introduced during the mutagenesis and cloning procedures, a T-to-G mutation was introduced at this position in the mutant to revert it to the wild-type sequence. The revertant was introduced into pBin0.2SE to produce the partial repeat pBin1 .2SErev that was mobilized into A. tumefaciens. Agroinoculation of this construct into N. benthamiana produced severe symptoms (in all 10 inoculated plants) that were indistinguishable from a wild-type infection, indicating that the point mutation was solely responsible for the mutant phenotype. In Beta vulgaris (cv. Giant Western), the wild-type clone induced typical BCTV symptoms of plant stunting, vein swelling, and severe upward leaf curl in 35 of 37 inoculated plants (over four experiments) 1 week after agroinoculation. In contrast, plants remained asymptomatic when agroinoculated with the mutant. Comparison of the viral DNA levels in B. vulgaris was particularly difficult due to the extremes of symptom phenotype exhibited by the wild-type and mutant virus.

508

SHORT

COMMUNICATIONS

However, dot blot analysis (17) of 53 plants showed that 40 supported a systemic infection, of which 14 contained mutant DNA levels in newly expanding leaves comparable to wild-type levels in similar, but severely affected leaves. Southern blot analysis of viral DNA forms from 5 randomly selected plants harvested 7 weeks after agroinoculation (Fig. 1, lanes 5-9) revealed typical supercoiled and single-stranded DNAs, although the relative level of the latter was reduced in comparison with the wild-type infection (lane 4) in this instance. However, the ratio of mutant DNA forms in younger plants (2 weeks afteragroinoculation) was similar to that for wild-type virus. As previously observed (15, 16) less subgenomic DNA accumulated in B. vulgar& than in N. benthamiana. Southern blot analysis of nucleic acids from the only B. vulgaris plant to develop mild leaf curl and vein swelling symptoms, using the discriminating enzymes Sal1 and SnaBI, revealed the presence of wild-type, presumably revertant, DNA in a background of mutant progeny. The mutant was also successfully introduced into Nhtiana clevelandii (8 of 10 inoculated plants) to produce stunting and yellowing that was less pronounced than in the wild-type infection. Upward curling of the expanding leaves was confined to wild-type infection although the effect was less dramatic than observed in N. benthamiana. When introduced into Datura stramonium, wild-type virus induced severe downward curling followed by vein swelling and severe upward curling of the edges of the newly expanding leaves in 4 of the 8 agroinoculated plants. In contrast, the mutant induced only extremely mild downward curling in this host although dot blot analysis showed that 3 of the 10 inoculated plants were infected and supported wildtype levels of mutant DNA. The results indicate that the point mutation has a profound effect on symptom development in several hosts. Early investigations on the cytopathology of BCTV infection in B. vulgaris by light microscopy (18) and subsequent electron microscopy studies of BCTVinfected Spinacia oleracea and 6. vulgaris ( 19,20) have suggested that the characteristic vein swelling associated with BCTV infection is due primarily to hyperplasia induced in the phloem in which the dividing cells differentiate predominantly into abnormal sieve elements. This stimulation of cell division is associated with upward curling of infected leaves. Since neither vein swelling nor upward leaf curl has been observed in mutant infections even though a host such as RI. benthamiana supports wild-type levels of viral DNA, the data suggest that the mutation affects the ability of the virus to induce the hyperplastic response. Outward signs of necrosis associated with wild-type BCTV infection, resulting from hyperplasia and subsequent

collapse of the tissue (18-20), were noticeably absent in mutant-infected plants, supporting this view. The phenotype might be attributable to one or more of several effects. First, although the mutation does not affect the amino acid sequence of Cl, the valine codon GUA in the mutant is somewhat less favorable in most eukaryotes including dicotyledonous plants than the wild-type codon GUC (21). The mutation might therefore modify the action of this essential gene, although this seems unlikely since this unfavorable codon is used on three occasions in ORF Cl which has 11 valine residues in total. Second, it is possible that the mutation disrupts a c&-acting element within the genome. For example, single nucleotide substitutions within the geminivirus maize streak virus that have no effect on encoded amino acids have been shown to alter the host range and symptom phenotype of the virus possibly by disrupting promoter activity (22). Finally, the mutation may serve to disrupt the product of ORF C4. These alternatives were tested by introducing a G-to-T mutation into clone pBCT098 at position 2802 which again introduced a termination codon into ORF C4 as well as an additional Mae1 site without affecting the amino acid sequence of Cl. The mutant was introduced into pBin0.2SE to produce the partial repeat AC4/2 that was mobilized into A. turnefaciens. The phenotype of this mutant was indistinguishable from that of AC4/1 in both N. benthamiana (23 of 28 inoculated plants showing symptoms over two experiments, with 7 of 8 plants tested showing wild-type DNA levels) and B. vulgaris (3 of 10 inoculated plants supporting an asymptomatic infection, of which one plant showed wild-type DNA levels). The results suggest that ORF C4 participates in protein coding. Currently, there is no supporting evidence for the existence of such a product although homologous ORFs are present in all geminiviruses that infect dicotyledonous plants examined to date. Minor transcripts initiate just upstream of this ORF in TGMV (23) but analogous transcripts have not been observed in either ACMV or abutilon mosaic virus (24, 25). Furthermore, disruption of this ORF in ACMV and TGMV had no effect on infectivity and symptom development in N. benthamiana (7, 9) although ORF C4 might be functional only in monopartite geminiviruses. The fact that the BCTV mutants induce unique symptoms in N. benthamiana suggests that caution is required when interpreting results using this convenient laboratory host. It might be necessary to reappraise the mutagenesis data accumulated on ACMV and TGMV concerning this and other ORFs (7-9, 26) following the introduction of mutants into the natural hosts of these viruses. The molecular and cytopathological basis of the

SHORT

mutant phenotype tion.

COMMUNICATIONS

is the subject of a current investiga10. 11.

ACKNOWLEDGMENTS We thank Prof. Jeff Davies cussion. J.R.L. was supported

and Dr. Peter by an AFRC

Markham for critical studentship.

dis-

12. 13. 14.

REFERENCES 15. 1. TOWNSEND, R., WATTS, J., and STANLEY, J., NucleicAcidsRes. 14, 1253-1265 (1986). 2. KLINKENBERG, F. A., and STANLEY, J.,/. Gen. Viral. 71, 1409-l 412 (1990). 3. STANLEY, J., MARKHAM, P. G., CALLIS, R. J., and PINNER, M. S., EAYBOJ. 5, 1761-1767 (1986). 4. STANLEY, J., In “Seminars in Virology” (R. Hull, Ed.), Vol. 2, pp. 139-149. Saunders Scientific, Philadelphia, 1991. 5. STENGER, D. C., CARBONARO, D., and DUFFUS, J. E., /. Gen. Viral. 71,2211-2215 (1990). 6. BRIDDON, R. W., WATTS, J., MARKHAM, P. G., and STANLEY, J., Virology 172, 628-633 (1989). 7. ETESSAMI, P., SAUNDERS, K., WATS, J., and STANLEY, J., /. Gen. !/ire/. 72, 1005-1012 (1991). 8. BROUGH, C. L., HAYES, R. J., MORGAN, A. J., Courts, R. H. A., and BUCK, K. W., J. Gen. Viral. 69, 503-514 (1988). 9. ELMER, J. S., BRAND, L., SUNTER, G., GARDINER, W. E., BISARO,

16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

509 D. M., and ROGERS, S. G., Nucleic Acids Res. 16, 7043-7060 (1988). BEVAN, M., Nucleic Acids Res. 12, 871 1-8721 (1984). ZAMBRYSKI, P., Joos, H., GENETELLO, C., LEEMANS, J., VAN MONTAGU, M., and SCHELL, J., EMBO/. 2, 2143-2150 (1983). COVEY, S. N., and HULL, R., virology 111, 463-474 (1981). WAHL, G. M., STERN, M., and STARK, G. R., Proc. Nat/. Acad. Sci. USA 76, 3683-3687 (1979). FEINBERG, A. P., and VOGELSTEIN, B., Anal. Biochem. 132, 6-l 3 (1983). STENGER, D. C., STEVENSON, M. C., HORMUZDI, S. G., and BISARO, D. M., J. Gen. Viral. 73, 237-242 (1992). FRISCHMUTH, T., and STANLEY, J., \/irology 189, 808-81 1 (1992). MAULE, A. J., HULL, R., and DONSON, J., J. Viral. Methods 6,2 l5224 (1983). ESAU, K., Am. J. Bof. 22, 149-163 (1935). ESAU, K., Ann. Bot. 40, 637-644 (1976). ESAU, K., and HOEFERT, L. L., Am. J. Bot. 65, 772-783 (1978). MURRAY, E. E., LOTZER, J., and EBERLE, M., Nucleic Acids Res. 17,477-498(1989). BOULTON, M. I., KING, D. I., DONSON, J., and DAVIES, J. W., viralogy 183, 114-121 (1991). SUNTER, G., and BISARO, D. M., virology 173, 647-655 (1989). WARD, A., Ph.D. Thesis, University of East Anglia. U.K. (1989). FRISCHMUTH, S., FRISCHMUTH, T., and JESKE, H., virology 185, 596-604 (1991). ETESSAMI, P., CALLIS, R., ELLWOOD, S., and STANLEY, J., Nucleic Acids Res. 16, 4811-4829 (1988).