Cross-protection between Sunn-hemp Mosaic and Tobacco Mosaic Viruses

J. gen. Virol. (1986), 67, 1679-1687. Printed in Great Britain 1679 Key words: TMV/SHMV/coat protein/cross-protection/uneoating Cross-protection between Sunn-hemp Mosaic and Tobacco Mosaic Viruses By THOMAS M. ZINNEN*i" AND ROBERT W. FULTON Department of Plant Pathology, University of Wisconsin-Madison, Madison, Wisconsin 53706, U.S.A. (Accepted 25 April 1986) SUMMARY Cross-protection reactions of two tobamoviruses, sunn-hemp mosaic virus (SHMV) and common tobacco mosaic virus (TMV-C), were investigated and compared. A mutant of SHMV (SHMV-n), produced by nitrous acid treatment, induced necrotic lesions in bean. SHMV protected completely against this mutant and against SHMV-n RNA. SHMV in bean protected only weakly, however, against TMV-C. To determine whether the coat protein of these viruses affected the ability to superinfect, RNA of each virus was encapsidated in the coat protein of the other. TMV-C RNA encapsidated in SHMV coat protein was five- to 27-fold less infectious on SHMVinfected bean leaves than TMV-C RNA re-encapsidated in TMV-C coat protein. When homologous or heterologous coat protein was added to inocula, infectivity for healthy plants was diminished markedly more by homologous protein, suggesting that extraneous homologous protein diminished infectivity by inhibiting viral uncoating. SHMV-n RNA encapsidated in TMV-C coat protein did not superinfect SHMVinfected bean leaves. Thus, although coat protein was shown to be a factor in crossprotection in some situations, other factors must also be involved. INTRODUCTION Strains of most viruses cross-protect infected plants against superinfection by related strains. The strength of the protection varies, but it may be strong enough to provide a practical control of plant virus infection (Rast, 1972; Costa & Muller, 1980). The phenomenon is also of interest from a theoretical aspect. One theory of the mechanism of cross-protection (Kohler & Hauschild, 1947) postulates a sequestering of an essential metabolite by the protecting virus. The specificity of protection, however, necessitates postulating specific metabolites for each virus, for which there is no evidence. Ross (1974) suggested that the protecting virus sequestered the majority of the ribosomes in a cell due to the rapid increase in mrna. Thus, nucleic acid of a second virus would find few ribosomes left for binding and would be degraded prior to expression. This theory is compatible with observations (Wenzel, 1971) that the first virus often inhibits a second virus, even if the two viruses are not related. It does not, however, explain the specificity of cross-protection. Gibbs (1969) thought that the RNA of a challenge virus might become irreversibly but ineffectually bound to the replicase of the virus already in the cell. The RNA of unrelated viruses would not bind and thus there would be no protection. Palukaitis & Zaitlin (1984) proposed that RNA complementary to viral RNA, such as that produced during replication of the viral RNA, might bind to the challenge RNA and prevent its expression. DeZoeten & Fulton (1975) suggested that when RNA of a second virus enters a cell containing a related virus it may be encapsidated by free coat protein before it can become replicated. Zaitlin (1976), however, showed that a mutant of tobacco mosaic virus (TMV) whose coat protein did not encapsidate RNA nevertheless reduced susceptibility to superinfection by common TMV (TMV-C) by about 70~. Moreover, Sarkar & Smitamana (1981) showed that t Present address: Agrigenetics Advanced Research Division, 5649 East Buckeye Road, Madison, Wisconsin 53716, U.S.A. 0000-7030 1986 SGM

1680 T. M. ZINNEN AND R. W. FULTON half-leaves of Nicotiana tabacum L. cv. Samsun-EN previously inoculated with a coat proteinfree mutant of TMV were less susceptible to infection by a necrotizing strain of TMV than were the mock-inoculated opposite half-leaves. More recently, Sherwood & Fulton (1982) presented evidence suggesting that coat protein may be involved in cross-protection in another way. Necrotizing strains of TMV, if sufficiently concentrated, induce necrotic lesions in dark green areas in leaves of Nicotiana sylvestris Spegaz & Comes systemically infected with TMV-C, but not in light green leaf areas (Fulton, 1951). Sherwood & Fulton (1982) found that the light green tissue of 'mosaic leaves' of N. sylvestris was superinfected if inoculated with RNA of necrotizing strains or with this RNA encapsidated in coat protein of the unrelated brome mosaic virus. This RNA encapsidated in TMV-C coat protein, however, did not superinfect the light green areas. TMV-N RNA failed, however, to induce lesions on N. sylvestris leaves inoculated a few days previously with a high concentration of TMV-C. Sherwood & Fulton (1982) concluded that necrotizing TMV failed to infect the light green areas of mosaic leaves of N. sylvestris because it was unable to uncoat, presumably due to the coat protein already present at the infection site. On the other hand, in locally inoculated leaves where TMV-C was multiplying actively, other factors seemed to prevent superinfection. Wilson & Watkins (1986) showed that extraneous TMV-C coat protein inhibited the in vitro co-translational disassembly of TMV-C virions previously treated briefly at ph 8. They suggested that the added coat protein interfered with uncoating or induced re-encapsidation of partially uncoated viral RNA. They proposed, by analogy, that high levels of free coat protein on mosaic leaves of plants systemically infected with TMV may reverse or prevent the initial release of coat protein from the 5' end of challenge TMV. Dodds et al. (1985) found that RNA, but not virions, of a strain of cucumber mosaic virus (CMV) superinfected tomato leaves containing a mild strain of CMV. Superinfection occurred only in inoculated leaves, but symptoms of the challenge virus did not appear. Double-stranded RNA and virions of the challenge strain were identified and distinguished from those of the protecting strain by electrophoresis. There is other evidence that superinfection may occur and that the challenge virus may multiply without causing symptoms (Cassels & Herrick, 1977; Barker & Harrison, 1978). In spite of this we have taken as evidence of superinfection the development of symptoms (necrotic lesions in our system) following challenge inoculation. The presence of small amounts of challenge virus may be difficult to detect in the presence of large amounts of protecting virus in extracts of challenged leaves. If no symptoms occur following challenge, replication is presumably more restricted than in unprotected leaves, and such leaves may still be considered as protected. We have investigated cross-protection between two distinct tobamoviruses, sunn-hemp mosaic virus (SHMV) and TMV-C, and between SHMV and a necrotizing mutant (SHMV-n), to determine the effect of coat protein in such protection. SHMV and TMV-C are more distantly related than the strains of TMV investigated by Sherwood & Fulton (1982), and other factors in addition to coat protein might be expected to affect cross-protection. The host ranges of the viruses also differ and the host might be expected to affect cross-protection. METHODS Viruses. SHMV is a tobamovirus distinct from TMV-C (Zaitlin et al., 1977; Van De Walle & Siegel, 1982). The amino acid sequences of the coat proteins differ in 96 residues; moreover, SHMV coat protein has 162 amino acid residues (Meshi et al., 1982) while TMV-C has 158. SHMV is morphologically similar to TMV-C. It has been referred to as the bean strain or the cowpea strain of TMV. Kassanis & Verma (1975) have given it the status of a distinct virus and we follow this practice. Reactions of various plants to the viruses used are given in Table 1. Because we had no way to test SHMV for its ability to cross-protect against a closely related strain, a necrotizing mutant (SHMV-n) was produced by nitrous acid mutagenesis as described by Siegel (1965). The selected mutant induced necrotic primary lesions and systemic necrosis in bean (Phaseolus vulgaris L.) and induced systemic mosaic in cowpea [ Vigna unguiculata (L.) Walp.]. It reacted strongly with SHMV-specific antiserum. Inoculations were made on corundum-dusted leaves with virus suspended in 30 mm-phosphate buffer ph 8. SHMV was propagated in primary leaves of Pinto bean. TMV-C was propagated in N. sylvestris or in N. tabacum

Cross-protection between tobamoviruses 1681 Table 1. Viruses and the reaction of selected host plants* P. vulgaris CV r ~ ~ N. tabacum Virus Pinto Black Turtle Bountiful N. sylvestris cv. Xanthi-nc SHMV M M M LL LL SHMV-n~" LL + SL LL + SL LL + SL LL LL TMV-C LL LL - M LL SBMV LL M M * M, Mosaic; LL, local necrotic lesions; SL, systemic necrotic lesions and mosaic; -, no reaction. t Mutant generated from SHMV. L. cv. Havana 38. Both viruses were purified as described by Sherwood & Fulton (1982). Southern bean mosaic virus (SBMV) was purified from inoculated primary leaves of Bountiful bean as described by Hull (1977). Isolation of viral RNA and coat protein. RNA was prepared by the water-saturated phenol method of Ralph & Berquist (1967) as modified by Sherwood & Fulton (1982). Coat protein was prepared by the glacial acetic acid method of Fraenkel-Conrat (1957), except that pelleted coat protein was resuspended in 0.2 M-borate buffer ph 8. Encapsidation of viral RNA in coat protein. Viral RNA was encapsidated in either SHMV or TMV-C coat protein by the method of Richards & Williams (1972). Coat protein was dialysed against 50 m~t-acetate ph 4.7, until the preparation became cloudy. It was then dialysed at room temperature for 24 to 48 h against three changes of 41 ram-sodium phosphate ph 7.1 (equivalent to 0.1 ionic strength). Sufficient RNA to make a protein : RNA ratio of 40 : 1 was added to the protein. The mixture was incubated overnight at room temperature. Virus was collected by a cycle of low- and high-speed centrifugation. Pellets were resuspended in 3 mm-edta ph 7.0. Tests for superinfection, cross-protection and inoeulum interference. Cross-protection tests were done in two ways. In the first, one half of a leaf was inoculated with the protecting virus, and the opposite half was mock-inoculated. Four to 6 days later, the entire leaf was inoculated with the challenge virus. The extent of superinfection was expressed as an average of the ratios calculated by dividing the number of lesions in a protected half-leaf by the number in the opposite, mock-inoculated half-leaf. Because the protecting and challenge viruses were applied to the same leaf, this protection was called 'local protection'. In the second, seedlings were inoculated with the protecting virus; mock-inoculated seedlings served as controls. Challenge inoculations were made on systemically infected leaves and on comparable leaves of mock-inoculated plants, and therefore are referred to as tests for 'systemic protection'. Usually the first trifoliolate leaf of beans was used; second trifoliolate leaves were used when tests required distinct light and dark green areas, which were most distinct on these leaves. Both types of leaves are referred to as 'mosaic leaves'. When comparing two treatments, one was applied to the left leaflet and the other to the right leaflet. Healthy plants of the same age served as controls. In tests involving the effect of added coat protein on virus infectivity, which we termed inoculum interference, mixed inoculum was applied to one half of a leaf while inoculum containing only the same concentration of necrotizing virus was applied to the opposite half. The infectivity of the mixed inoculum was expressed as a percentage of lesions formed by the necrotizing virus alone. Values in tables are averages followed by standard errors. RESULTS Cross-protection of SHM V against TM V-C, SBM V and SHM V-n By selecting appropriate host plants, SHMV, TMV-C and SBMV could be evaluated as both protecting and challenge viruses in tests of local protection. SBMV is an icosahedral virus with single-stranded, messenger-sense RNA ; it is unrelated to the tobamoviruses and was included as an assay for any general decrease in susceptibility to superinfection due to infection by an unrelated virus (Wenzel, 1971). In tests for local protection, SHMV did not induce lesions in half-leaves of N. sylvestris inoculated 5 days before with 100 tig/ml TMV-C. In the reciprocal test, TMV-C induced 79/o as many lesions on half-leaves of Pinto bean inoculated 5 days before with 1 mg/ml SHMV, as compared to control half-leaves. SBMV, however, induced 76% as many lesions on SHMV-infected half-leaves of Pinto bean as on control half-leaves. Yet when SBMV (1 mg/ml) was tested as the protecting virus on P. vulgaris cv. Black Turtle bean, TMV-C induced only 6% as many lesions on half-leaves inoculated 6 days before with SBMV, as compared to control half-leaves.

1682 Table 2. T. M. ZINNEN AND R, W. FULTON Cross-protection by SHMV in Pinto bean against TMV-C particles and against equivalent amounts of TMV-C RNA Lesions per leaflet* Amount t J" Inoculum (!ag/ml) Healthy (H) Mosaic (Mi Ratio M/H RNA 0-5 15 _+ 6 3-5 + 1.0 0.23 Virions 10 283 _+ 38 109 _+ 24 0.39 RNA 1-0 27 _+ 9 2.3 + 0.9 0.09 Virions 20 703 _+ 98 86 + 12 0.12 * Values are averages, from six replicates per trial, of TMV-C lesions/leaflet on first trifoliolate leaves of healthy and of mosaic Pinto beans. To test the ability of SHMV to protect beans systemically against the necrotizing mutant, SHMV-n was applied at 12 ~tg/ml to first trifoliolate mosaic leaves of Pinto bean and to healthy plants. In two trials with five replicates per trial, previously healthy leaves developed averages of 6.4 and 6.9 lesions per cm 2, Mosaic leaves developed no lesions at all. An 80-fold increase of SHMV-n in the inoculum (to 1 mg/ml) also failed to induce lesions in protected leaves. Sherwood & Fulton (1982) found that the RNA of a necrotizing strain (TMV-N) was more efficient than intact TMV-N in superinfecting mosaic N. sylvestris leaves. To see whether SHMV behaved similarly, its RNA (50 ~tg/ml) and SHMV-n (10 gtg/ml) were applied to opposite leaflets of the first trifoliolate leaves of mosaic and healthy Pinto bean. In two trials with four replicates per trial, lesions averaging 67 and 178 per leaflet appeared in previously healthy leaflets inoculated with SHMV-n RNA and SHMV, respectively. No lesions were produced on mosaic leaves by either inoculum. Thus, unlike TMV-C, SHMV protected completely against an RNA inoculum of a necrotic mutant. It was of interest to determine whether this degree of protection would extend to the distantly related TMV-C. SBMV was also tested as a challenge virus to detect any general decrease in susceptibility to superinfection. In two trials in which Pinto beans systemically infected with SHMV were challenged with TMV-C and SBMV, SBMV induced as many lesions in previously healthy leaves as in SHMV-infected leaves (superinfection ratios of 1.02 and 1.13). TMV-C, however, induced fewer lesions in SHMV-infected leaves than in previously healthy leaves (superinfection ratios of 0.31 and 0.33). It should be noted that TMV-C induced lesions in both dark green and light green areas of mosaic leaves, in contrast to the inability of TMV-N to superinfect light green areas of mosaic leaves of N. sylvestris (Fulton, 1951). Moreover, the RNA of TMV-C was no more efficient than intact TMV-C in superinfecting bean leaves containing SHMV (Table 2). The protecting capacity of SHMV in bean was complete against a closely related mutant, but weak in both light green and dark green areas against the distantly related TMV-C. To test whether another protecting virus unrelated to TMV-C would provide incomplete protection, SBMV was tested as a systemic protectant against TMV-C. In two trials with SBMV as a protecting virus in Black Turtle bean and TMV-C (1, 5 and 25 ~tg/ml) as challenge virus, the superinfection ratios averaged 0.64, with a median of 0-59 and a range of 0.46 to 0.88. The suppression of superinfection provided by SBMV was about one-half of that provided by SHMV. Systemic cross-protection by TMV-C against SHMV in N. sylvestris Previous results on cross-protection between strains of TMV had been obtained mainly with N. sylvestris as the host. This host also could be used to test cross-protection by TMV-C against SHMV, which causes primary necrotic lesions in N. sylvestris. SHMV (30 I.tg/ml) was applied to leaves of N. sylvestris systemically infected with TMV-C and to healthy plants. Mosaic leaves developed on average 9.5 lesions per leaf, while previously healthy leaves averaged 83 lesions per leaf. SHMV superinfected mosaic leaves, but lesions developed only in dark green areas, behaving in this respect in the same way as TMV-N (Sherwood & Fulton, 1982).

Cross-protection between tobamoviruses 1683 Sherwood & Fulton (1982) showed that TMV-N RNA, or TMV-N treated with bentonite, induced lesions in both dark and light green areas of mosaic leaves of N. sylvestris. To determine if SHMV behaved similarly, half-leaves of TMV-C-infected N. sylvestris and of healthy plants were inoculated with 1.5 Ixg/ml SHMV RNA. Opposite half-leaves were inoculated with 30 Ixg/ml SHMV. SHMV induced an average of 522 lesions per half-leaf in previously healthy leaves and 20 lesions per half-leaf in mosaic leaves. Lesions induced by SHMV RNA induced an average of 131 lesions in control leaves and 5-5 lesions in mosaic leaves. Neither SHMV nor SHMV RNA induced lesions in light green areas of mosaic leaves, and neither did SHMV treated with 1 mg/ml bentonite. Thus, SHMV and TMV-N behaved differently in superinfection of N. sylvestris. These experiments provided no evidence that coat protein was involved in cross-protection reactions between TMV-C and SHMV. Heterologous eneapsidation and superinfection To determine whether SHMV-n RNA could superinfect SHMV-infected bean if encapsidated in coat protein of T MV-C (which readily superinfects), SHMV-n RN A was encapsidated in coat protein of either TMV-C or SHMV. When healthy leaflets of Bountiful bean were inoculated with SHMV-n RNA in TMV-C coat protein (50 ~tg/ml) an average of 449 lesions were formed per leaflet. When healthy leaflets were inoculated with SHMV-n RNA in SHMV coat protein (10 ~tg/ml) an average of 243 lesions were formed per leaflet. Mosaic leaves developed no lesions at all. Repeated attempts to detect SHMV-n in these leaves by subinoculations to healthy Pinto beans were unsuccessful. The complete protection of beans infected with SHMV against superinfection by SHMV-n raised a question regarding the specificity of superinfection. To test this, SHMV-n (10 ~tg/ml) and TMV-C (10 ~tg/ml) were applied to opposite leaflets of SBMV-infected Black Turtle beans and to healthy plants. Lesions induced by SHMV-n and TMV-C averaged 31 and 45 per leaflet, respectively, in infected bean, and 59 and 71 on previously healthy leaves. Superinfection ratios were 0.52 for SHMV-n and 0.56 for TMV-C. Prevention of superinfection by SHMV-n in bean infected by SHMV appeared to be a specific reaction. It seemed possible that TMV-C readily superinfected SHMV-infected bean leaves because the SHMV coat protein in the bean leaves was sufficiently different from the coat protein of the TMV-C for it not to inhibit uncoating. If this were so, then SHMV might inhibit the uncoating of and infection by a virus composed of TMV-C RNA encapsidated in SHMV coat protein. Such heterologously encapsidated virus (Het-TMV-C), and TMV-C RNA reconstituted in TMV-C coat protein (Recon-TMV-C) were prepared. To confirm that heterologously encapsidated particles containing TMV-C RNA were formed, preparations were examined in the electron microscope. Particles typical of TMV were seen. To test for full-length infectious particles, RNase at 10 ng/ml was added to the heterologously encapsidated preparations, to native TMV-C and to TMV-C RNA. After 30 min the infectivities of these preparations were compared on Pinto bean. Treatment with RNase eliminated all infectivity of TMV-C RNA, but reduced infectivity of the other preparations by only about 50~. The superinfecting capacity of the heterologously encapsidated TMV-C RNA and the reconstituted TMV-C were compared on opposite leaflets of SHMV-infected Pinto bean and on healthy beans. Protection against the heterologously encapsidated TMV-C RNA was five- to 27- fold stronger than against the reconstituted TMV-C (Table 3, Fig. 1). When heterologously encapsidated TMV-C RNA was inoculated to SHMV-infected bean with light and dark green areas, lesions developed in both areas, as with TMV-C. Interference by added coat protein on virus infectivity Previous results from this work and from others (Atabekov et al., 1970; Wilson & Watkins, 1986) suggested that an early stage in the infection process was affected by coat protein. It was thought that if coat protein present in protected leaves decreased superinfection, such a decrease might be mimicked by adding purified coat protein to the inoculum of challenge virus. To test for such interference, SHMV and TMV-C coat proteins (1 mg/ml prepared as for encapsidating RNA) were added separately to TMV-C and SHMV inocula (0.25 ~tg/ml, in 41 ram-phosphate,

1684 T. M. ZINNEN AND R. W. FULTON Fig. 1. Lesions induced by TMV-C RNA encapsidated in (a, c) SHMV coat protein or in (b, d) TMV-C coat protein on (a, b) healthy Pinto bean leaf or on (c, d) SHMV-infected Pinto bean leaf. T a b l e 3. Comparison of cross-protection by S H M V in Pinto bean against TMV-C RNA encapsidated in S H M V coat protein (Het-TMV-C) with cross-protection against TMV-C RNA encapsidated in TMV-C coat protein (Recon-TMV-C) Trial 1 2 3 Virus (100 lag/ml) Recon-TMV-C Het-TMV-C Recon-TMV-C Het-TMV-C Recon-TMV-C Het-TMV-C Lesions per leaflet* r '~ Healthy (H) Mosaic (M) 745 + 111 82 + 18 908 + 87 3.7 + 1 218 + 40 63 + 13 207 + 33 13 _ 6 510 + 29 318 + 34 586 + 87 20 + 5 Ratio M/H 0.11 0.004 0.29 0.062 0-62 0.035 * Values are averages, from five to seven replicates per trial, of lesions per leaflet on first trifoliolate leaves of healthy and of mosaic Pinto beans. p H 7.1). Infectivities o f m i x e d and control inocula were c o m p a r e d in N. tabacum cv. X a n t h i - n c. T M V - C i n o c u l u m s u p p l e m e n t e d w i t h T M V - C protein or w i t h S H M V protein i n d u c e d 11 ~ a n d 790/oo as m a n y lesions as the control respectively. T h e opposite effect was o b t a i n e d w i t h S H M V : w h e n m i x e d w i t h T M V - C protein or w i t h S H M V protein, S H M V particles i n d u c e d 6 9 ~ a n d 1 6 ~ as m a n y lesions as the control respectively. Thus, infectivity was d e c r e a s e d m o r e w h e n whole virus i n o c u l u m was m i x e d w i t h h o m o l o g o u s c o a t p r o t e i n t h a n w i t h heterologous c o a t protein.

160 "~ 140 8 120' 100 0.~. 80 6o I I I Cross-protection between tobamoviruses 1685 200 I ~ l l,6o t.. 8 120 0 ~ 40.o 20 O O 0 101 I I 10 l's 102 10 vs 103 Protein added (~tg/ml) 011 l l I II 101 101'5 102 102.5 10 3 Protein added (pg/ml) Fig. 2 Fig. 3 Fig. 2. Effect of TMV-C protein (O) and SHMV protein (ZS) on the infectivity of TMV-C RNA encapsidated in TMV-C coat protein on primary leaves of healthy Pinto bean. Fig. 3. Effect of TMV-C protein (O) and SHMV protein (Z~x) on the infectivity of TMV-C RNA encapsidated in SHMV coat protein on primary leaves of healthy Pinto bean. Similarly, TMV-C coat protein interfered more than SHMV coat protein with infection by reconstituted TMV-C in Pinto bean (Fig. 2). SHMV coat protein interfered more than TMV-C coat protein with the infectivity of TMV-C RNA encapsidated in SHMV protein (Fig. 3). Interference was consistently stronger when the added coat protein was the same as the protein in the virus particle. A time course trial showed that interference did not increase with time after mixing the inoculum with coat protein (data not shown). The decrease in TMV-C infectivity resulting from the addition of coat protein appeared to be specific because the addition of TMV-C coat protein to SBMV inoculum slightly increased its infectivity on Pinto bean. DISCUSSION SHMV and SBMV incompletely protected Black Turtle bean against TMV-C in tests of both local and systemic protection. SHMV incompletely protected Pinto bean against TMV-C but not against SBMV in tests of both local and systemic protection. Since SBMV conferred some protection against TMV-C, at least some of the protection by SHMV against TMV-C might have been due to a non-specific effect. The weak protection of SHMV against TMV-C in contrast to the lack of protection of SHMV against SBMV, however, suggests that the interaction between the two tobamoviruses is specific. Including SBMV as an assay for general effects leads apparently to contradictory conclusions, and shows the difficulty in describing protection as specific. Cross-protection by SHMV against its necrotizing mutant appeared to be complete, in contrast to the protection resulting from infection of N. sylvestris by TMV-C against the closely related TMV-N (Sherwood & Fulton, 1982) and against the distantly related SHMV. Furthermore, while SHMV induced lesions only in dark green areas of mosaic N. sylvestris leaves, TMV-C induced lesions in both light green and dark green areas of mosaic bean leaves. The host is evidently a factor in determining the completeness of the cross-protection reaction. The failure of SHMV-n RNA to cause lesions in SHMV-infected bean leaves, and of SHMV RNA to cause lesions in light green areas of mosaic leaves of N. sylvestris, indicates that

1686 T. M. ZINNEN AND R. W. FULTON unknown factors other than coat protein must prevent the development of the symptoms of and possibly the replication of the challenge virus. But because these unknown factors completely block superinfection, any inhibitory effect of coat protein would be masked. We could test, however, whether coat protein promotes superinfection. SHMV-n RNA encapsidated in TMV- C coat protein, however, failed to superinfect mosaic bean leaves; thus, there is no evidence that coat protein of a successful challenge virus promotes superinfection by a challenge RNA which alone did not superinfect. To detect inhibition of superinfection by coat protein, a challenge virus, or at least its RNA, must induce lesions in the protected leaves. The effect of coat protein on superinfectivity can then be measured by encapsidating the successful challenge RNA in various coat proteins, and assaying the infectivity of the virions in healthy leaves and in protected leaves. The decreased superinfectivity on SHMV-infected bean leaves of TMV-C RNA encapsidated in SHMV coat protein rather than in TMV-C coat protein indicated that SHMV coat protein somehow inhibited superinfection. In this case, the coat protein encapsidating the challenge RNA was the same as that of the protecting virus. This result led us to test directly the effect of added coat protein on the infectivity of virus in healthy plants. The addition of homologous coat protein to the inoculum decreased the infectivity of both native and re-encapsidated virus on healthy plants more than the addition of heterologous coat protein. Our results confirm those of Atabekov et al. (1970) with TMV-C and cucumber virus 4, although we disagree with their suggestion that the inhibition of infection by homologous coat protein indicates the presence of a specific binding site in plant cells for plant viruses. The inhibiting effect of added homologous protein on infectivity and superinfectivity is the result that would be expected if uncoating of a virus is affected by extraneous identical protein. Brakke & Van Pelt (1969) proposed that uncoating of TMV in vitro is a reversible reaction in which the virus particle is the reagent, and coat protein and partially stripped particles (and eventually naked RNA) are products. Pelcher et al. (1980) showed that added TMV-C coat protein slowed the rate of uncoating in vitro of TMV-C and a tomato isolate of TMV. They concluded that coat protein in solution may play a major role in the production and stabilization of partially uncoated virus particles. They noted, however, that extensive reversal of disassembly did not occur under their conditions. Wilson & Watkins (1986) showed that TMV coat protein preferentially inhibited the co-translational disassembly of TMV virions in vitro, and suggested the inhibition of uncoating proposed by Sherwood & Fulton (1982) results from the prevention of the initial release of a few coat protein subunits from the 5' end of the challenge virus RNA. Our results of tests with plants are consistent with the predictions of hypotheses derived from tests in vitro. We thank G. A. dezoeten for suggestions and S. A. Vicen for art work and photography. This research was supported by the College of Agricultural and Life Sciences, University of Wisconsin, Madison, Wisconsin. REFERENCES ATABEKOV, J. G., NOVIKOV, V. K., VISHNICHENKO, V. K. & JAVAKHIA, V. K. (1970). A study of the mechanisms controlling the host range of plant viruses. II. The host range of hybrid viruses reconstituted in vitro and of free viral RNA. Virology 41, 108-115. BARKER, H. & HARRISON, B. D. (1978). Double infection, interference and superinfection in protoplasts exposed to two strains of raspberry ringspot virus. Journal of General Virology 40, 647-657. BRAKKE, M. K. & VAN PELT, N. (1969). Influence of bentonite, magnesium, and polyamines on degradation and aggregation of tobacco mosaic virus. Virology 39, 516-533. CASSELS, A. C. & I-I]/RRICK, C. C. (1977). Cross protection between mild and severe strains of tobacco mosaic virus in doubly inoculated tomato plants. Virology 78, 253-260. COSTA, A. S. & MULLER, G. W. (1980). Tristeza control by cross protection: a U.S.-Braziiian cooperative success. Plant Disease 64, 538-541. DEZOETEN, G. A. & FULTON, R. w. (1975). Understanding generates possibilities. Phytopathology 65, 221-222. DODDS, J. A., LEE, S. Q. & TIFFANY, M. (1985). Cross protection between strains of cucumber mosaic virus: effect of host and type of inoculum on accumulation of virions and double-stranded RNA of the challenge strain. Virology 144, 301-309. FRAENKEL-CONRAT, H. (1957). Degradation of tobacco mosaic virus with acetic acid. Virology 4, 1-4. FULTON, R. W. (1951). Superinfection by strains of tobacco mosaic virus. Phytopathology 41, 579-592.

Cross-protection between tobamoviruses 1687 GIBBS, A. J. (1969). Plant virus classification. Advances in Virus Research 14, 263-328. HULL, a. (1977). The banding behaviour of the viruses of southern bean mosaic virus group in gradients of caesium sulphate. Virology 79, 50-57. KASSANIS, a. & VARMA, A. (1975). Sunn-hemp mosaic virus. Commonwealth Mycological Institute/Association of Applied Biologists Descriptions of Plant Viruses, No. 153. KOHLER, E. & nauscmld, I. (1947). Betrachtungen und versuche zum Problem der "erwarbenen Immunit~t" gegen Virusinfektionen bei Pflanzen. Zfichter 17/18, 97-105. MESHI, x., OnNO, T. ~ OKADA, V. (1982). Nucleotide sequence of the 30 K protein cistron of cowpea strain of tobacco mosaic virus. Nucleic Acids Research 10, 6111-6117. PALUKAITIS, P. & ZAITLIN, M. (1984). A model to explain the 'cross-protection' phenomenon shown by plant viruses and viroids. In Plant-Microbe Interactions: Molecular and Genetic Perspectives, pp. 420-429. Edited by T. Kosuge & E. W. Nester. London & New York: Macmillan. PELCHER, L. E., WALMSLEY, S. L. & MACKENZIE, S. L. (1980). The effects of heterologous and homologous coat protein on alkaline disassembly of tobacco and tomato isolates of tobacco mosaic virus. Virology 105, 287-290. RALPH, R. K. & BErtQUIST, B. L. (1967). Separation of viruses into components. Methods in Virology 2, 463-545. RAST, A. "r. B. (1972). M II-16, an artificial symptomless mutant of tobacco mosaic virus for seedling inoculation of tomato crops. Netherlands Journal of Plant Pathology 78, 110-112. RICttARDS, K. E. & WILLIAMS, R. C. (1972). Assembly of tobacco mosaic virus in vitro: effect of state of polymerization of the protein component. Proceedings of the National Academy of Sciences, U.S.A. 69, 1121-1124. Ross, A. F. (1974). Interaction of viruses in the host. Acta horticulturae 36, 247-260d. sarkar, S. & SMITAMANA, P. (1981). A proteinless mutant of tobacco mosaic virus: evidence against the role of a viral coat protein for interference. Molecular and General Genetics 184, 158-159. Snl~l~wooD, J. L. & fulton, R. W. (1982). The specific involvement of coat protein in tobacco mosaic virus cross protection. Virology 119, 150-158. SIEGEL, h. (1965). Artificial production of mutants of tobacco mosaic virus. Advances in Virus Research 11, 25-60. VAN DE WALLE, M. J. SIEGEL, A. (1982). Relationships between strains of tobacco mosaic virus and other selected plant viruses. Phytopathology 72, 390-395. WENZEL, t3. (1971). Vergleichende Untersuchungen des Interferenzverhaltens phytopathogener Viren auf Tabak. Phytopathologische Zeitschrift 71, 147-162. WILSON, T. M. A. & WATKINS, P. A. C. (1986). Influence of exogenous viral coat protein on the cotranslational disassembly of tobacco mosaic virus (TMV) particles in vitro. Virology 149, 132-135. ZAITLIN, M. (1976). Viral cross protection: more understanding is needed. Phytopathology 66, 382-383. ZAITLIN, M., BEACHY, R. N. & BRUENING, G. (1977). Lack of molecular hybridization between RNAs of two strains of TMV: reconsideration of the criteria for strain relationships. Virology 82, 237-241. (Received 9 January 1986)