Resistance to Squash Mosaic Comovirus in Transgenic Squash Plants Expressing Its Coat Protein Genes

Resistance to squash mosaic comovirus in transgenic squash plants expressing its coat protein genes

Sheng-Zhi Pang1,†, Fuh-Jyh Jan1,†, David M. Tricoli2, Paul F. Russell3,KimJ.Carney3, John S. Hu1, Marc Fuchs1, Hector D. Quemada3 & Dennis Gonsalves1,∗ 1Department of Plant Pathology, Cornell University, NYSAES, Geneva, NY 14456, USA (∗author for corres- pondence; Fax: 315-787-2389; e-mail: [email protected]); 2Seminis Vegetable Seeds, 37437 State Highway 16, Woodland, CA 95695, USA; 3Formerly of Asgrow Seed Company; †These authors contributed equally to this research.

Received 2 February 1999; accepted in revised form 19 July 1999

Key words: cosuppression, ﬁeld test, gene silencing, pathogen-derived resistance, SqMV, transgenic squash

Abstract The approach of pathogen-derived resistance was investigated as a means to develop squash mosaic comovirus (SqMV)-resistant cucurbits. Transgenic squash lines with both coat protein (CP) genes of the melon strain of SqMV were produced and crossed with nontransgenic squash. Further greenhouse, screenhouse and field tests were done with R1 plants from three independent lines that showed susceptible, recovery, or resistant phenotypes after inoculations with SqMV. Nearly all inoculated plants of the resistant line (SqMV-127) were resistant under greenhouse and field conditions and less so under screenhouse conditions. Plants of the recovery phenotype line (SqMV-3) were susceptible when inoculated at the cotyledon stage but leaves that developed later did not show symptoms. The susceptible line (SqMV-22) developed symptoms that persisted and spread throughout the plant. Plants were also analyzed for transcription rates of the CP transgenes and steady state transgene RNA levels. Results showed that the resistant line SqMV-127 displayed post-transcriptional silencing of the CP transgene as evidenced by high transcription rates but concomitant low accumulation of transgene transcripts. This is the first report on the development of transgenic squash that are resistant to SqMV.

Abbreviations: CP, coat protein; SqMV, squash mosaic comovirus; CPMV, cowpea mosaic comovirus

Introduction proteins are derived by proteolytic cleavages [9]. The virus capsid is composed of two distinct polypeptides Squash mosaic comovirus (SqMV) is a member of the with molecular weight of 42 kDa and 22 kDa, which genus Comovirus, with isometric virus particles about are encoded as a 64 kDa precursor by M-RNA [14]. 30 nm in diameter [4]. SqMV infects almost all spe- Nucleotide sequences of the coat protein (CP) genes cies in the Cucurbitaceae family and is transmitted by have been reported for a melon strain of SqMV [15]. beetles and through seeds [5]. The viral genome of Recently, cDNA clones of RNA-2 from two SqMV comoviruses consists of two single-stranded, positive- isolates were constructed and the full-length sequence sense RNA molecules identiﬁed as middle-component were determined [13]. RNA (M-RNA or RNA-2) and bottom-component Control of SqMV currently relies on the use of RNA (B-RNA or RNA-1) of about 4200 and 6000 virus-free seeds and insecticides for the control of nucleotides, respectively. Both M-RNA and B-RNA beetles [18]. Due to lack of SqMV resistant host genes are polyadenylated at the 30 end and have a genome- that can be incorporated into cucurbits by conventional linked protein (VPg) at the 50 end [9]. The RNAs are breeding [22], pathogen-derived resistance [24] could translated into polyproteins from which the functional provide an alternative to control SqMV and other co- 88 moviruses. Nida et al. [19] reported the production of transgenic tobacco expressing the 60 kDa precursor of the two capsid proteins of cowpea mosaic comovirus (CPMV) but resistance was not observed. Recently, reports from Sijen et al. [25, 26] have shown that Nico- tiana benthamiana plants expressing the full-length replicase, movement protein or CP genes were resistant to CPMV. However, transgenic cucurbits with resistance to comoviruses have not been reported. We report here that transgenic squash plants expressing the chimeric genes encoding the two SqMV CP genes are resistant to SqMV under greenhouse, screenhouse and ﬁeld conditions. Furthermore, the protection we observed was RNA-mediated via post- transcriptional transgene silencing [3, 21, 29].

Materials and methods

Cloning and squash transformation Figure 1. Agrobacterium binary vector pGA482G/SqMV. The loc- ations of important genetic elements within the binary vector are indicated: BR, right border; BL, left border; nos-npt, plant express- Two full-length CP genes from the melon strain of ible neomycin phosphotransferase gene; cos, cos site; tet, bacterial squash mosaic comovirus (SqMV m-88) were PCR tetracycline resistance gene; and gent, bacterial gentamicin resistance gene. Both 22 kDa and 42 kDa CP transgenes are driven by amplified and cloned in the sense orientation into a 0 plant expression vector pUC18cpexp [27], as pre- CaMV 35S promoter (35S-P), CMV-C 5 -untranslated leader and 35S terminator (35S-T). viously described by Hu et al. [15]. The resulting pUC18cpexp-22 kDa and -42 kDa CP plasmids were partially digested by HindIII (the CP open ELISA, Northern blot, nuclei isolation and nuclear reading frames contain internal HindIII sites), and run-on transcription assay both SqMV 22 kDa and 42 kDa CP expression cas- settes were cloned into the transformation vector Double-antibody sandwich enzyme-linked immun- pGA482G, a derivative of pGA482 described by An osorbent assay (DAS-ELISA) [10] was used to detect [2]. The original binary vector pGA482 contains the the accumulation of the two CP polypeptides and the right and left T-DNA borders of pTi37 which flank nptII enzyme in transgenic plants using antibodies the plant expressible neomycin phosphotransferase II against SqMV virions [15] and an nptII ELISA kit (5 gene (nptII), restriction enzyme polylinker, and bac- Prime to 3 Prime, Boulder, CO), respectively. Total teriophage λcos site. To improve the use of this vector, plant RNAs were isolated according to the procedure the bacterial gentamicin-(3)-N-acetyl-transferase gene described by Napoli et al. [17]. Total RNA (10 µg [1] was cloned into the SalI site located outside the per lane) was separated on a formaldehyde-containing T-DNA region, generating pGA482G. The plant ex- agarose gel [23] and the agarose gels were stained with pressible SqMV CP genes were inserted as tandem ethidium bromide to monitor the uniformity of total repeats and were oriented in the same direction as plant RNAs in each lane. Hybridization conditions the nptII gene (Figure 1). The resulting binary vec- were described previously [20]. Isolation of nuclei and tor was transferred into the disarmed T-DNA deletion nuclear run-on transcription assays were essentially derivative of the Agrobacterium tumefaciens strain performed as described by Dehio and Schell [7]. The C58. same amount of labeled RNA was used for hybridiz- Transgenic squash plants were obtained by in- ation to replicated dot blot membranes (BioRad, Her- oculating leaf tissues from an Asgrow inbred yellow cules, CA) that contained 0.2 µgof22kDaCP,42kDa crookneck squash line using the procedure essentially CP, actin and nptII genes. Images of some autoradio- described by Tricoli et al. [28]. grams were photographed with a COHU CCD camera, Model 4915-2000 (COHU, San Diego, CA). Signals 89 were quantitated using the US National Institutes of Eleven R0 transgenic squash plants were gener- Health Image program version 1.59. ated and crossed with an untransformed inbred line while only five of them were able to produce seeds. Inoculation of transgenic plants R1 plants from these five lines were initially analyzed at Kalamazoo, MI by inoculating seedlings and The strain SqMV m-88 was used in inoculation tests. monitoring the symptom development over a 35 day It was obtained from infected melon seeds [15] and period. Transgenic R1 seeds were germinated and as- typed as the melon strain by biological and serological sayed by nptII ELISA to identify the nontransgenic comparisons with the type strain [18]. Inocula were segregants from the R1 populations. All inoculated prepared by propagating the virus in zucchini squash transgenic plants of line SqMV-22 showed typical (Cucurbita pepo L.) and grinding infected leaves in symptoms about 10 days after inoculation and symp- 0.01 M potassium phosphate, pH 7.0. In most cases, toms were severe throughout the 35-day observation the cotyledons of small squash plants (ca. 10 days after period. These symptoms were similar to those of non- germination) were dusted with 400 mesh carborundum transformed controls. The R1 transgenic plants of line and rubbed with 1:15 dilution of SqMV m-88-infected SqMV-127, however, initially had moderate symp- squash leaf extract. Systemic infection were recorded toms but the symptoms either were not persistent or every day for at least 35 days. new leaves were symptomless by 21 days after inoculation. At 35 days after inoculation, none of the Field evaluation inoculated plants were symptomatic. The R1 plants of line SqMV-3 developed symptoms which persisted The R seeds of three transgenic squash lines as well 1 throughout the observation period but were, on the as those of nontransformed controls were germinated average, less severe than the nontransgenic plants. A on 9 June, seedlings were screened by nptII ELISA, percentage of the R plants of lines SqMV-136 (17%) and nptII positive seedlings were inoculated twice 1 and SqMV-69 (39%) were symptomless throughout with SqMV m-88 on 20 June and 22 June 1994. The the 35-days observations period. Three transgenic 35 inoculated R plants of the three transgenic squash 1 squash lines, SqMV-3, SqMV-22, and SqMV-127, lines and the untransformed control were planted in were selected for further investigations because their the field 3 feet within the row, and 6 feet between rows R plants exhibited differential symptom reactions. using a randomized block design on 30 June 1994. 1 All subsequent tests were done at Geneva, NY. Data were taken every other day for at least 60 days. Protection of transgenic squash plants against SqMV m-88 in greenhouse conditions Results Further test were done to evaluate the resistance of Production of transgenic squash and selection of test the selected transgenic squash in greenhouse condi- lines tions. Transgenic R1 seeds were germinated in soil Because the two SqMV CP polypeptides are encoded and the seedlings were assayed by nptII ELISA to as parts of a polyprotein, they do not contain trans- identify the nontransgenic segregants from the R1 pop- lation initiation codons in the original corresponding ulations. Cotyledons of 10-day-old nontransgenic and cDNA and the 42 kDa CP gene does not have a trans- transgenic plants were blindly inoculated with SqMV lation stop codon. Therefore, the two CP genes were m-88 to avoid human bias of inoculation. Inoculation first modified by polymerase chain reaction to restore results are summarized in Table 1. the open reading frames and were then cloned into All R1 nontransgenic segregants and untrans- the plant expression vector pUC18cpexp as previously formed control plants showed systemic symptoms 8– described by Hu et al. [15]. Each CP gene was thus 14 days after inoculation with no escapes. Similar to driven by the cauliflower mosaic virus (CaMV) 35S the controls, all of the transgenic R1 plants from line promoter, the cucumber mosaic virus 50-untranslated SqMV-22 displayed systemic symptoms that persisted leader and the 35S terminator (Figure 1). These plant throughout the life cycle of the test plants. All of the expressible SqMV coat protein genes were used to transgenic R1 plants from line SqMV-3 initially de- produce transgenic squash lines via Agrobacterium- veloped systemic symptoms; but 50% (10 of 20 plants) mediated transformation. of them did not show symptoms in newly developed 90

Table 1. Reactions of transgenic R1 squash expressing the coat protein genes of squash mosaic virus to inoculations by SqMV m-88 in different environments.1

Greenhouse Screenhouse Field Line n Sus Rec Res n Sus Rec Res n Sus Rec Res

Control39390 0 27270 0 35350 0 SqMV-2225250 0 23230 0 35350 0 SqMV-3 20 10 10 0 33 18 15 0 35 13 22 0 SqMV-127 21 2 1 18 54 9 11 34 35 0 0 35

1 The cotyledons of transgenic R1 plants were inoculated 10 days after germination with 15-fold diluted extracts of SqMV m-88 infected squash plants. The reactions could be grouped into three phenotypes: 1, Susceptible (Sus), typical systemic symptoms were observed at 8–14 days after inoculation (DAI); 2, Recovery (Rec), systemic symptoms were observed at 8–14 DAI, but not on the newly grown leaves at 20–48 DAI; 3, Resistant (Res): the plants remained symptom free throughout their life cycles. n, total number of plants inoculated.

Figure 2. Coordinated expression of 22 kDa and 42 kDa CP genes in R1 transgenic plants. Total RNAs isolated from transgenic plants (10 µg per lane) were analyzed by Northern hybridization using the 22 kDa or 42 kDa CP gene probe. Lane 1, a nontransgenic control; lanes 2–5, four SqMV-127 R1 plants; lanes 6–9, four SqMV-3 R1 Figure 3. Relation of CP RNA expressions with virus resistance in plants; lanes 10 and 11, two SqMV-22 R1 plants. Sizes of transcripts individual R1 plants. Total RNAs isolated from transgenic plants for 22 kDa and 42 kDa CP genes are ca. 900 bp and ca. 1400 bp, at 10 days after germination (10 µg per lane) were analyzed by respectively. Northern hybridization using the 22 kDa gene probe. The same plants were subsequently inoculated with SqMV m-88. Lanes 1 and 31, two nontransgenic controls; lanes 2–19, eighteen SqMV-127 R1 leaves 3–8 weeks after inoculation (Table 1). The plants; lanes 20–27, eight SqMV-3 R1 plants; lanes 28–30 and 32, asymptomatic young leaves did not contain the virus four SqMV-22 R1 plants. Reactions to SqMV are shown below the Northern bands: S represents susceptibility and R represents resist- as determined by DAS-ELISA or by back-inoculation ance. The signal intensities were quantified by NIH-Image program to susceptible plants and could not be infected by and normalized with the nontransgenic controls. The low (L) ex- subsequent mechanical inoculation (data not shown). pressors had density readings between 24 and 85 while the high These latter plants were referred to as showing a re- (H) expressors had density readings between 202 and 231. Size of 22 kDa CP gene transcripts is ca. 900 bp. covery phenotype. In contrast, 86% (18 of 21 plants) of the SqMV-127 transgenic progeny were highly resistant to the virus; no symptom was observed at any Virus resistance is correlated with low steady state developmental stages and no virus was recovered from levels of transgene SqMV mRNAs leaves of these plants. These plants were referred to as showing a resistance phenotype. The rest of the To investigate the mechanism of the resistance, an- SqMV-127 R1 plants showed the recovery phenotype other set of transgenic R1 plants were initially assayed (1 of 21 plants) or were susceptible (2 of 21 plants). by DAS-ELISA and western blot for the expression of the CP transgenes. To our surprise, no CPs were detected in any of these plants (data not shown) using the antibody that successfully detected SqMV CP 91 polypeptides in protoplast transient assay [15]. For this reason, individual transgenic R1 plants were analyzed before inoculation by Northern hybridizations using either the 22 kDa or the 42 kDa CP gene probe. Because the 22 kDa and 42 kDa CP genes are phys- ically linked within the same T-DNA, the levels of the 22 kDa CP transcript, as expected, correlated with those of the 42 kDa CP transcript in each R1 plants (Figure 2). As shown in Figure 3, 16 of 18 tested transgenic SqMV-127 R1 plants accumulated low levels of the CP transcripts and were resistant to the virus; the two R1 plants that accumulated high levels of the transcripts were susceptible. In contrast, Figure 4. Nuclear run-on transcription analysis on R1 progeny. Labeled nuclear RNAs were hybridized to 0.2 µg of restriction all of the SqMV-22 and nearly all of the SqMV-3 R1 plants accumulated high levels of the 22 kDa CP trans- enzyme-digested 22 kDa CP, 42 kDa CP, nptII and actin gene frag- ments which were dot blotted onto a membrane. The nuclei used in gene transcripts and were susceptible to the virus; the the assays were isolated from a SqMV-127 and a SqMV-3 R1 plants. one SqMV-3 R1 plant that accumulated a low level of The hybridization transcripts were quantified using NIH-Image pro- the 22 kDa CP transcripts was resistant. These data gram. The ratios were the relative CP and nptII transcription rates which were normalized to the actin transcription. The average of indicated that virus resistance correlated well with transcription rate of two actin genes was arbitrarily set as 1. low steady state levels of the 22 kDa CP transgene transcript. To determine whether the reduced steady state and in the field trial showed typical symptoms in the CP transgene mRNA levels in the resistant R1 plants middle of July. Generally, both SqMV-3 and SqMV- were due to post-transcriptional down-regulation of 127 R1 plants performed better under the field condi- the transgenes, nuclear run-on transcription analysis tions than in the screenhouse conditions. For example, was performed before inoculation using the endogen- 100% of the SqMV-127 transgenic progeny exhibited ous actin as a control. Both 22 kDa and 42 kDa CP the resistance phenotype under the field conditions as genes were found to be more efficiently transcribed in compared to only 63% under the screenhouse condi- the SqMV-127 R1 plants than in the SqMV-3 R1 plants tions. Similarly, more SqMV-3 progeny displayed the (Figure 4), even though the latter accumulated higher recovered resistance phenotype under the field con- levels of steady-state transgene transcripts (Figure 3). ditions (64%) than under the screenhouse conditions These results suggested that the reduced steady state (45%). levels of the transcripts were due to rapid turnover of the transgene transcripts rather than the reduction of the transgene transcription rates. Discussion

Protection of transgenic squash plants against SqMV SqMV is one of the major viruses that infect cucurbits in screenhouse and field conditions because it can be spread by beetles and is transmitted through seeds [5]. The latter can have severe economic To evaluate the resistance of the three lines (SqMV- consequences, in melons especially, because entire 3, -22, and -127) under different growing conditions, lots of seeds are rejected if some of the seeds are found transgenic R1 plants that had germinated 10 days to be contaminated with SqMV. Up to now, natural earlier in the greenhouse were inoculated with SqMV resistance to SqMV have not been found in cucurbits and subsequently moved to a screenhouse or to the [22]. We have shown that transgenic squash plants field 10 days after inoculation on 30 June 1994. Symp- expressing the chimeric genes encoding both CP poly- tom development was monitored until the fruits be- peptides of SqMV are protected against viral infection came mature (Table 1). Since other viruses moved into via the mechanism of post-transcriptional transgene si- the field test late in the season, symptomatic plants lencing [3, 21, 29]. This is the first report to show that were tested for SqMV by DAS-ELISA to distinguish pathogen-derived resistance is a potentially practical SqMV from other viruses. As shown in Table 1, all way to develop SqMV resistant cucurbits. controls and SqMV-22 R1 plants in the screenhouse 92

In this study, protected plants showed two dis- istance phenotype at Kalamazoo while similar plants tinct phenotypes: resistance or recovery. In the res- showed the complete resistance phenotype at Geneva istant phenotype (line SqMV-127), transgene silen- (Table 1). cing apparently occurred immediately after germin- Reports from Sijen et al. [25, 26] show homology- ation, resulting in a uniform transgene suppression dependent resistance CPMV in N. benthamiana plants and resistance to virus infection when inoculated at expressing the full-length replicase, movement pro- the cotyledon stage. In the recovery phenotype (line tein or CP genes. Previously, Nida et al. [19] pro- SqMV-3), plants were susceptible when inoculated at duced transgenic tobacco plants expressing the two CP the cotyledon stage but leaves that developed later genes of CPMV as a translational fusion but protec- did not show symptoms and could not be infected by tion against CPMV was not observed. Our transgenic subsequent mechanical inoculation. Our recent pre- squash plants expressed the CP transgenes of SqMV liminary experiments (data not shown) suggest that the separately and protection was found in the silenced recovery phenotype is due to the activation of post- transgenic squash plants. transcriptional gene silencing in upper leaves of R1 plants of SqMV-3. We are doing further genetic and molecular analysis of the recovery lines to determ- Acknowledgements ine which factors might contribute to the activation of post-transcriptional transgene silencing in these We thank D. Hummer and S. Day for technical as- plants. sistance. This work was partially supported by grants Similar recovery phenotype has been reported [11, from Asgrow Seed Company. F.J.J. was partially sup- 16]. Lindbo et al. [16] observed that when initially ported by a fellowship from the Ministry of Education, challenged with tobacco etch potyvirus (TEV), to- Taiwan. bacco plants transformed with a full-length CP gene of TEV or its truncated form at the N terminus displayed typical TEV-induced symptoms, but gradu- References ally outgrew infection by ca. 3 weeks after inoculation. Guo and Garcia [11] reported that a recovery 1. Allmansberger R, Brau B, Piepersberg W: Genes for phenotype was observed in N. benthamiana plants gentamincin-(3)-N-acetyl-transferases III and IV. II. Nucle- otide sequences of three AAC(3)-III genes and evolutionary transformed with mutated nuclear inclusion b (NIb) aspects. Mol Gen Genet 198: 514–520 (1985). gene sequences of plum pox potyvirus (PPV-R). The 2. An G: Binary Ti vectors for plant transformation and promoter recovered newly developing leaves were symptom- analysis. Meth Enzymol 153: 292–305 (1987). 3. Baulcombe DC: Mechanisms of pathogen-derived resistance less and virus free, and were resistant to inoculation to viruses in transgenic plants. Plant Cell 8: 1833–1844 with PPV-R but not PPV-PS or tobacco vein mottling (1996). potyvirus (TVMV) [11]. 4. Bruening G: Comovirus group. CMI/AAB Descriptions of Comparative results showed that the levels of pro- Plant Viruses, No 199 (1978). 5. Campbell RN: Squash mosaic virus. CMI/AAB Descriptions tection in transgenic squash are apparently affected of Plant Viruses, No 43 (1971). by the conditions under which the plants are grown 6. de Carvalho F, Gheysen G, Kushnir S, van Montagu M, Inze (Table 1). The environmental effect on gene silencing D, Castresana C: Suppression of β-1,3-glucanase transgene have been observed in a number of cases, presum- expression in homozygous plants. EMBO J 11: 2595–2602 (1992). ably by its profound influence on growth dynamics 7. Dehio C, Schell J: Identification of plant genetic loci involved and consequently the level of transgene expression. in a posttranscriptional mechanism for meiotically reversible Cosuppression of chalcone synthase (CHS) genes in transgene silencing. Proc Natl Acad Sci USA 91: 5538–5542 Petunia [30] and β-1,3-glucanase genes in tobacco [6] (1994). 8. Dorlhac de Borne FD, Vincentz M, Chupeau Y, Vaucheret H: are stimulated by high light intensities. Silencing of Co-suppression of nitrate reductase host genes and transgenes chitinase genes in N. sylvestris [12] and nitrate re- in transgenic plants. Mol Gen Genet 243: 613–621 (1994). ductase in tobacco [8] are dependent on germination 9. Goldbach, RW, Wellink J: Comoviruses: molecular biology and growth conditions. Thus, environmental condi- and replication. In: Harrison BD, Murant AF (eds) The Plant Viruses, vol 5: Polyhedral Virions and Bipartite RNA tions could be the major factors influencing transgene Genomes, pp. 35–76. Plenum Press, New York (1996). silencing derived protection in the field conditions. 10. Gonsalves D, Trujillo EE: Tomato spotted wilt virus in papaya This environmental effect may also explain why some and detection of the virus by ELISA. Plant Dis 70: 501–506 (1986). R1 plants of line SqMV-127 showed the recovered res- 93

11. Guo HS, Garcia JA: Delayed resistance to plum pox potyvirus 22. Provvidenti R: Resistance to viral diseases of cucurbits. In: mediated by a mutated RNA replicase gene: Involvement of Kyle MM (ed) Resistance to Viral Diseases of Vegetables, a gene-silencing mechanism. Mol Plant-Microbe Interact 10: pp. 8–43. Timber Press, Portland, OR (1993). 160–170 (1997). 23. Sambrook J, Fritsch EF, Maniatis T: Molecular Cloning: A 12. Hart CM, Fischer B, Neuhaus JM, Meins F: Regulated inactiv- Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory, ation of homologous gene expression in transgenic Nicotiana Cold Spring Harbor, NY (1989). sylvestris plants containing a defense-related tobacco chitinase 24. Sanford JC, Johnston SA: The concept of parasite-derived gene. Mol Gen Genet 235: 179–188 (1992). resistance-deriving resistance genes from the parasite’s own 13. Haudenshield JS, Palukaitis P: Diversity among isolates of genome. J Theor Biol 115: 395–405 (1985). squash mosaic virus. J Gen Virol 79: 2331–2341 (1998). 25. Sijen T, Wellink J, Hendriks J, Verver J, van Kammen A: Rep- 14. Hiebert E, Purcifull DE: Mapping of the two coat protein lication of cowpea mosaic virus RNA1 or RNA2 is speciﬁcally genes on the middle RNA component of squash mosaic virus blocked in transgenic Nicotiana benthamiana plants express- (comovirus group). Virology 113: 630–636 (1981). ing the full-length replicase or movement protein genes. Mol 15. Hu JS, Pang S-Z, Nagpala PG, Siemieniak DR, Slightom Plant-Microbe Interact 8: 340–347 (1995). JL, Gonsalves D: The coat protein genes of squash mosaic 26. Sijen T, Wellink J, Hiriart J-B, van Kammen A: RNA- virus: cloning, sequence analysis, and expression in tobacco mediated virus resistance: role of repeated transgenes and protoplasts. Arch Virol 130: 17–31 (1993). delineation of targeted regions. Plant Cell 8: 2277–2294 16. Lindbo JA, Sliva-Rosales L, Proebsting WM, Dougherty WG: (1996). Induction of a highly speciﬁc antiviral state in transgenic 27. Slightom JL: Custom polymerase-chain-reaction engineering plants: Implications for regulation of gene expression and of a plant expression vector. Gene 100: 251–255 (1991). virus resistance. Plant Cell 5: 1749–1759 (1993). 28. Tricoli DM, Carney KJ, Russell PF, McMaster JR, Groff 17. Napoli C, Lemieux C, Jorgensen R: Introduction of a chimeric DW, Hadden KC, Himmel PT, Hubbard JP, Boeshore ML, chalcone synthase gene into petunia results in reversible co- Quemada HD: Field evaluation of transgenic squash con- suppression of homologous genes in trans. Plant Cell 2: 279– taining single or multiple virus coat protein gene constructs 290 (1990). for resistance to cucumber mosaic virus, watermelon mosaic 18. Nelson MR, Knuhtsen HK: Squash mosaic virus variability: virus 2, and zucchini yellow mosaic virus. Bio/Technology 13: review and serological comparisons of six biotypes. Phyto- 1458–1465 (1995). pathology 63: 920–926 (1973). 29. van den Boogaart T, Lomonossoff GP, Davies JW: Can we ex- 19. Nida DL, Anjos JR, Lomonossoff GP, Ghabrial SA: Ex- plain RNA-mediated virus resistance by homology-dependent pression of cowpea mosaic virus coat protein precursor in gene silencing? Mol Plant-Microbe Interact 11: 717–723 transgenic tobacco plants. J Gen Virol 73: 157–163 (1992). (1998). 20. Pang S-Z, Slightom JL, Gonsalves D: Different mechan- 30. van der Krol AR, Mur LA, Beld M, Mol J, Stuitje AR: Flavon- isms protect transgenic tobacco against tomato spotted wilt oid genes in petunia: addition of a limiting number of copies and impatiens necrotic spot tospoviruses. Bio/Technology 11: may lead to a suppression of gene expression. Plant Cell 2: 819–824 (1993). 291–299 (1990). 21. Prins M, Goldbach R: RNA-mediated virus resistance in transgenic plants. Arch Virol 141: 2259–2276 (1996).