Lack of Cross-Protection Between Citrus Exocortis Vir- Oid and Citrus Viroids Associated with Mild Symptoms in Etrog Citron*

Home , Citrus exocortis

Lack of Cross-Protection Between Citrus Exocortis Vir- oid and Citrus Viroids Associated with Mild Symptoms in Etrog Citron*

S. M. Garnsey, R. F. Lee, and J. S. Semancik

ABSTRACT. Etrog citron plants systemically infected with citrus viroid (CV) isolates, which cause mild symptoms in citron, developed the same typical severe symptoms of citrus exocortis viroid (CEV) infection as noninfected plants when challenge-inoculatedby graft or stem-slash inoculation with CEV. There was no difference inincubation period for expression of symptoms ofthe challenge isolate between plants with a primary infection of a mild viroid and noninfected plants. When extracts from citrons infected with the viroids inducing mild symptoms were subjected to sequential gel electrophoresis (sPAGE) under native and denaturing conditions and silver stained, no CEV band was found, but bands corresponding to citrus viroids CV-IIa and CV-IIIb were present. No hybridization was observed between cDNAprobes to CEV and either mild citronviroid. The lackof cross-protection, or interference by mild viroids with infection by the CEV challenge, supports the lack of relationship between CEV and the other mild citrus viroids studied. Index words. sequential gel electrophoresis, stem-slash inoculation

Etrog citron was used extensively tristeza virus (13), and cross-protec- for testing citrus for exocortis infection tion had been reported between mild following its introduction as a diagnos- and severe strains of potato spindle tic indicator for citrus exocortis (4). tuber viroid (PSTV) (2,14) and be- Some variation in symptom expression tween citrus exocortis viroid (CEV), in citron indicators was noted (17), and PSTV, and chrysanthemum stunt some inoculum sources caused only (CSV) viroids (14) in tomato or very mild symptoms (8,18). At the chrysanthemum. Infectivity assays time, it was assumed that the causal and purification attempts by PAGE in- agent of exocortis was a virus and that dicated that the mild isolates were the variation in symptom expression present in lower concentrations than indicated presence of isolates or strains CEV in comparable tissue (1,8). It was of different severities. The viroid na- also observed that mild symptoms ture of exocortis later became evident, could be obtained by infecting less reac- but the assumption remained that the tive clones of citron with CEV (8). mild symptoms caused by some isolates These observations suggested that reflected strain differences among a failure of cross-protection could be as- single viroid pathogen. sociated with either poor distribution Several cross-protection experi- or poor replication of the protectingiso- ments were initiated in Florida to test late, and this was investigated in sub- different isolates which produced mild sequent experiments. symptoms for ability to protect cit- As these protection studies were rons against several isolates that pro- progressing, evidence accumulated duced classical severe symptoms. No that the symptoms in Etrog citron, evidence was obtained for cross-pro- which were once all attributed to CEV, tection in these tests. These results were caused by several viroids with were puzzling since cross-protection distinct differences in molecular had been reported in citrus with citrus weight and other properties (3,6). Examination of the viroid isolates used in the various cross-protection studies *Mention of a trademark, warranty, prop- by sequential PAGE (sPAGE) proce- rietary product, or vendor does not constitute a dures revealed that there were several guarantee by the U.S. Department of Agricul- ture and does not imply its approval to the exclu- viroid species present, and that the sion of other products or vendors that may also mild isolates test for protection did not be suitable. contain typical CEV. Twelfth IOCV Conference

Recent reports have been pub- Agriculture and Consumer Services, lished on interference or protection be- Division of Plant Industry, code Te 20- tween two CEV isolates (7) and be- 3-5) to Etrog citron and subsequent tween members of the CV-I1 group passage by slash-cut transmission to (15,22). other Etrog citrons. It also causes typ- This paper reports that cross-pro- ical, severe CEV symptoms in citron tection did not occur between CV-I1 or and contains CEV, CV-IIa, and CV- CV-I11 and CEV. The lack of cross- IIIb (23). Source El0 was obtained protection observed is a further indica- originally by graft-inoculation to Etrog tion of a lack of relationship between citron from a Persian lime tree on Group I1 and Group I11 citrus viroids Rangpur lime rootstock and sub- which cause mild symptoms in Etrog sequent passage by slash-cut inocula- citron and CEV. A preliminary report tion to Etrog citron. It produces only ofthese results has been presented (9). mild symptoms in Etrog citron. Ex- tracts from ElO-infected plants do not MATERIALS AND METHODS carry CEV, but contain CV-111. Source Ell was obtained by graft-inoculation Host plants. All cross-protection to citron from a propagation of the Tem- studies were done with clonal propaga- ple orange Te 20-3-5 (see E22), which tions of three selections of Etrog citron: had undergone heat therapy treatment. Arizona 861,861-S1, and OES-4. Their It was also transferred by mechanical responses to CEV infection have been inoculation to citron. Ell causes mild described (8,18). All receptor plants symptoms in Etrog citron similar to were propagated as rooted cuttings those produced by E10. Extracts from and grown in individual containers in Ell-infected plants containboth CV-I1 a sterilized potting medium. Plants and CV-111. The CV-I1 is apparently were maintained in a partially shaded CV-IIa and not CV-IIb (21) since the glasshouse cooled with an evaporative original source tree had been propa- cooling system. Temperature condi- gated on Orlando tangelo for more than tions during the periods of experimen- 10 yr without cachexia symptoms (16). tation followed a normal diurnal fluctu- Isolates E16b and E25 used in one li- ation with night temperatures of 21-24 mited test produce typical CEV C and day temperatures of 27 to 34 C. symptoms in Etrog citron, but pres- Plants were fertilized as required to ence of other viroids was not deter- promote vigorous growth and sprayed mined. periodically to control pests. Precau- Inoculation procedures. Graft-in- tions were observed in handling and oculation was done by conventional T- pruning plants to avoid accidental vir- bud procedures with a chip of stem tis- oid infection by contamination (10,17). sue as the inoculum source. Mechanical Cutting tools were dipped in diluted inoculation was done by a stem-slash sodium hypochlorite or a mixture of procedure (10). The cutting blade was formalin and sodium hydroxide (10). contaminated either by dipping it into Viroid isolates. Four viroid an extract from infected plants, or by sources were used extensively in vari- cutting into the stem of a viroid-in- ous tests. Isolate E9 is a standard iso- fected plant, and then immediately late of CEV, which has been used in making a cut in the stem of the recep- many studies, and is also entered in the tor. Ten cuts were made per plant with collection of the American Type Cul- the blade freshly contaminated prior ture Collection as PV194. In our as- to each cut. Receptor plants were cut says, and in other independent tests back following inoculation to force new (12), it has been shown to contain only growth which was confined to a single CEV. It produces typical CEV symp- shoot. toms in Etrog citron. Source E22 was Challenge inoculations to previ- obtained by graft-inoculation from a ously inoculated plants were not made Temple orange tree (Florida Dept. of until definite symptoms of the primary Citms Viroids isolate had been observed, normally 4 within 20 days postinoculation if ex- to 8 months postinoculation. pressed in the first flush. Symptom evaluation. The leaf PAGE and cDNA hybridization epinasty symptoms typical of CEV in- procedures. Bark and leaf midrib tis- fection in citron (Fig. lb) were rated sue from young, symptomatic flushes visually on a 0 to 4 scale where 1 indi- of Etrog citron were powdered inliquid cated a deflection in the leaf blade from nitrogen and extracted immediately or zero to 90 degrees; 2, a deflection of 90 lyophilized for extraction at a later to 180 degrees; 3, a deflection of 180 to date. Extraction, concentration, and 270 degrees; and 4 indicated that the leaf electrophoresis were as described pre- had curled into at least a full circle (360 viously (5,20) and included phenol ex- degrees). A composite score per plant traction, partitioning with 2M lithium was compiled by totaling the reading chloride, partial purification on CF11 for all leaves on a flush and dividing by cellulose, and sPAGE. the number of leaves. Stunting was de- A cDNA probe to a California iso- termined by measuring shoot growth late of CEV was synthesized from the at different intervals after inoculation. viroid template by random priming in Incubation period was determined as the presence of 32P. Hybridizations the time between challenge inoculation were done as previously described (5). and appearance of symptoms of the RESULTS challenge isolate. Because of the periodicity in growth flushes in citrus, Cross-protection. In the first cross- incubation period data is less precise protection test attempted, Arizona 861 than in rapidly and continuously grow- citron plants systemically infected ing herbaceous plants. Under our con- with El0 and Ell isolates and non- ditions, symptoms of severe isolates of protected controls were challenged CEV normally appeared at the onset with E9. There were nine replications of the second flush of growth following per treatment. In addition, graft and inoculation that was approximately 40 mechanical challenge were tested for to 50 days. Symptoms can appear each combination. There was essen-

Fig. 1. Arizona 861 Etrog citron plants: A) uninoculated, B) systemically infected with mild citrus viroid isolate E10, C) systemically infected with citrus exocortis viroid isolate E9. 190 Twelfth ZOCV Conference tially no difference in results between mary isolate, but the challenge isolate the methods of inoculation. The results was E22 instead of E9, and 861-S1 and for the mechanical inoculation, which OES-4 citrons were used in place of was the less severe form of challenge, Arizona 861. The Ell and E22 protec- are summarized in Table 1 and also il- tion-challenge combination was chosen lustrated in Fig. 2. Based onincubation because at the time the test was in- period, leaf epinasty reaction, and itiated, it was felt that Ell had arisen stunting, there was no evidence of any as a heat therapy-induced mild variant modification in symptom severity or of E22 and could possibly be more delay in symptom expression resulting closely related to E22 than to E9, which from the primary infection by El0 and had a different origin. The OES-4 cit- Ell. There was, in fact, some slight ron showed a stronger reaction to mild indication of an additive effect of isolates than Arizona 861 (8) and was symptoms of the primary and the chal- considered a potentially more favored lenge isolates. host for those isolates. Challenge was A similar experiment was done by graft-inoculation 3 months after where Ell was used again as the pri-

TABLE 1 CITRUS EXOCORTIS VIROID SYMPTOMS, GROWTH RESPONSE, AND INCUBATION PERIOD IN HEALTHY ETROG CITRON AND ETROG CITRON INFECTED WITH TWO MILD CITRON VIROIDS

Inoculations Shoot Incubation growth Leaf period Primaryz ChallengeY (em)" epinastyw (days)"

None None 70 0.0 - El0 None 69 0.3 - Ell None 66 0.7 - None E9 24 2.5 51 El0 E9 22 2.6 50 Ell E9 19 2.8 49

"Plants graft-inoculated with mild isolates 8 months prior to challenge. YChallenge by stem-slash inoculation. "Growth measured 128 days after challenge. Average for 10 replicates. WO = no epinasty, 4 = 1eafcur1edfu11360".Average for allleaves on flushafter challengeinoculation. "Days after challenge inoculation.

TABLE 2 CEV SYMPTOMS AND GROWTH RESPONSE IN TWO ETROG CITRON SELECTIONS INOCU- LATED WITH A MILD VIROID ALONE, CEV ALONE, AND A MILD VIROID FOLLOWED BY CEV CHALLENGE

Inoculation

Host Primary Challenge" Growth (cm) SymptomsY

None None 176 0 Ell None 141 M Ell E22 25 S None E22 46 S None None 194 0 Ell None 164 M Ell E22 93 S None E22 84 S

"Plants challenged by graft-inoculation after symptoms of the primary mild isolate had appeared. Five replicates per treatment. YO = no symptoms, M = mild leaf epinasty, S = severe epinasty. Fig. 2. Arizona 861 citron plants 124 days after challenge inoculation with citrns exoeortis viroid isolate E9. A) uninoculated control with no primary orchallenge inoculation, B) challenge inoculation by E9 without wimarv inoculation by Ell. C) inoculated with mild citrus viroid isolate Ell without Ehallenge by ~9,and D) primary &ml&ion with Ell and subsequent challenge-inoculation by E9. Ell infected plants were showing systemic symptom at the time of dtdlange: by E9. All plants werecut hack at the time of challenge inoculation, and growth shown was formed after the challenge inoculation. symptoms of Ell appeared in the re- lenge plants clearly showed the severe ceptor plants receiving a primary in- symptoms of E22 infection along with oculation prior to challenge. There the severe stunting indicated. were five replications per treatment. An experiment was also conducted The results are summarized in to investigate the effect of challenge Table 2. Symptoms were not scored inoculum concentration. Healthy and numerically in this test, but all chal- Ell-infected citron plants were chal- 192 Twelfth ZOCV Conference lenge-inoculated with two dilutions of infected with all Florida isolates that three different CEV isolates as shown cause severe, typical CEV symptoms in Table 3. Again, there was no evi- in citron contained a prominent viroid dence of protection, even when chal- band that co-migrated with CEV stan- lenge inoculum titer was below the dards. In some cases, other faster mig- threshold to give 100% infection in rating viroidlike species were also healthy control plants. present. The CEV band was absent in Hybridization probes. Hybridiza- gels run from extracts of tissue infected tion tests with different Florida viroid with mild isolates including El0 and isolates to a labeled cDNA to a Califor- E 11. These mild sources contained one nia isolate of CEV showed strong posi- or more of the faster migrating species tive reactions to all sources, which which co-migrated with CV-I1 or CV- caused typical severe CEV symptoms I11 standards from California. Pat- in citron, but no reaction to mild terns for E9, E10, Ell, and E22 are sources, including E 10 and Ell. shown in Fig. 3. PAGE analysis. Initially, single- dimensional-PAGE (I)of extracts from DISCUSSION citrus infected with different Florida The results obtained in the differ- isolates revealed presence of an infec- ent cross-protection trials failed to tious viroid band from extracts of mild demonstrate that the viroids in the and severe sources. Concentrations mild isolates El0 and Ell conferred were consistently higher from extracts any protective effects to Etrog citron infected with severe isolates than from plants when these were subsequently those infected with mild sources, but challenge-inoculated by CEV. A delay no other physical discrimination was in expression of symptoms of the chal- made at that time. lenge isolate is commonly used as evi- Further evaluation of extracts of dence for protection or interference different Florida viroid isolates used (7,11,14). The lack of any delay in in the cross-protection tests described symptom expression in previously in- was attempted with the development oculated versus noninoculated plants of sPAGE procedures, coupled with suggests that infection by the chal- silver staining and refinements in dif- lenge isolate was unimpaired, and that ferential viroid species recovery from subsequent systemic movement of the CFll by elution with different concen- challenge isolate was also unrestricted. trations of ethanol (19,20). These There was no evidence that varying sPAGE tests were done several times. the host, the primary isolate, or the Different sources of tissue were used challenge isolate affected the results. for each assay. Extracts from tissue The level of challenge inoculum was

TABLE 3 INFECTION RATES BY THREE CITRUS EXOCORTIS VIROID ISOLATES CHALLENGE-IN- OCULATED AT TWO DILUTIONS INTO ETROG CITRON PLANTS WITH AND WITHOUT-. PRIMARY INOCULATION WITH A MILD CITRUS VIROID Challenge inoculcum concentrationY

10" 10-3

Primaryz None E16B E22 E25 E16B E22 E25

None 013" 313 313 213 213 213 013 Ell 013 313 313 313 313 113 113

"Ell inoculated plants were all showing symptoms at time of challenge. YInoculumconsisted of freshly prepared extract of young bark of Etrog citron systemically infected with the designated isolate. "Number of plants inoculatedlnumber showing typical severe CEV symptoms. Citrus Viroids

Because the sizes of CV-I1 and CV- I11 (297-302 and 280-292 nucleotides, respectively) are markedly different from CEV (371 nucleotides) (5,23) and cDNA probes made to CEV do not hy- bridize appreciably to CV-I1 or CV-I11 (19), the logical explanation for the lack of cross-protection observed is that there is no close relationship between CEV and either CV-I1 or CV-111. The results reported here are a biological confirmation of the molecular evidence already accumulated for this lack of relationship (6,23). The original assumption that the viroid isolates causing mild symptoms Fig. 3. Polyacrylamide gel electrophoresis in citron were mild strains of CEV of nucleic acid extracts from Etrog citron tis- seemed logical at that time, but again sues infected with: B) citrus exocortis viroid isolate E9, C) mild citrus viroid isolate E10, it illustrates the hazards of identifying D and F) mild citrus viroid isolate Ell, E) virus or viruslike pathogens solely on noninoculated, andG) CEV isolate E22. Lanes symptomatology. This is especially A and H contain standards for CEV (I), Citrus true with viroids where different vir- viroid CV-I1 (2), and citrus viroid CV-III(3). oids often induce very similar types of Note absence of CEV band in both El0 (C) and. Ell (D,F,) and that CV-111 present in E22 (G) symptoms. The previous references to is absent in E9 (B). the lack of cross-protection between CEV isolates are mostly suspect be- also not a factor. Plants inoculated by cause the assumed relationship be- stem-slash with diluted inocula were tween protecting and challenge iso- no less affected than those inoculated lates was based only on symptoms. by grafting, a much more severe chal- Lack of interference cannot be as- lenge. cribed to aunique property of the citron We cannot prove that prior to chal- host because evidence for interference lenge the primary infection of the vir- has been demonstrated recently be- oid isolates tested for protection was tween two biologically distinguishable systemic in all cells of the plants with isolates of CEV (7) and between two primary infections. However, in con- members of the Group I1viroids (22). trast to herbaceous hosts, use of citron Improved separation and staining hosts allowed an extensive incubation techniques have been critical for accu- period between primary and challenge rate identification of viroid species and inoculations for the primary isolate to played a key role in this study. For ex- become systemic, a factor that favors ample, early attempts to test relation- protection (7,ll). Challenge inocula- ships between our mild and severe tions were made only in sites where sources with end-labeled viroid from symptoms of mild isolate primary in- source E22, which had been obtained fection existed. via a single-phase-PAGE procedure The sPAGE and cDNA hybridiza- (I), gave misleading results because tion data clearly showed that CEV was the multiple viroid species in the probe absent in the protecting isolates were not separated. When techniques tested. The levels of CV-I1 and CV-I11 were developed to separate and purify observed in sPAGE gels indicated that the different species, the confusion was the samples were collected and pro- eliminated. cessed under conditions favorable for Our results illustrate the need to detection of viroids. Presence of CV- characterize by several methods all IIIb clearly did not suppress replica- components used in viroid cross-pro- tionordetectionofCEVinisolate E22. tection studies; and fortunately, good 194 Twelfth IOCV Conference collaboration was available to resolve teries which were observed in the the apparent discrepancies and mys- course of this investigation.

LITERATURE CITED

1. Baksh, N., R. F. Lee, and S. M. Garnsey 1984. Detection of citrus exocortis viroid by polyacrylamide gel electrophoresis, p. 343-452. In: Proc. 9th Conf. IOCV. IOCV, Riverside. 2. Branch, A. D., B. J. Benenfeld, E. R. Franck, J. F. Shaw, M. L. Varban, K. K. Willis, D. L. Rosen, and H. D. Robertson 1988. Interference between coinoculated viroids. Virology 163: 538-546. 3. Broadbent, P. and S. M. Garnsey

1987. Citrus exocortis., *D. 235-245. In: T. 0. Diener (ed.) The Viroids. Plenum Press. New York, 344 pp. 4. Calavan, E. C., E. F. Frolich, J. B. Carpenter, C. N. Roistacher, and D. W. Christiansen 1964. Ra~idindexing for exocortis of citrus. Phvto~atholom" -- 54: 998-999. 5. Duran-Vila, N., R. ~lor&,and J. S. Semancik 1986. Characterization of viroidlike RNAs associated with the citrus exocortis syndrome. Virolow 105: 75-84. ~ur~n-vil;"~.,J. A. Pina, J. F. Ballester, J. Juarez, C. N. Roistacher, R. Rivera-Bustamante, and J. S. Semancik 1988. The citrus exocortis disease: A complex of viroid-RNAs, p. 152-164. In: Proc. 10th Conf. IOCV. IOCV, Riverside. Duran-Vila, N. and J. S. Semancik 1990. Variations in the "cross-protection" effect between two strains of the citrus exocortis viroid (CEV). Ann. Appl. Biol. 117: 367-377. Feldman, A. W., S. M. Garnsey, and R. W. Hanks 1974. Coumarin accumulation in citrus tissue infected with exocortis virus, as influenced by virus titer, strain, and citron clone, p. 177-183. In: Proc. 6th Conf. IOCV. Univ. Calif. Press, Richmond. Garnsey, S. M., R. F. Lee, and J. S. Semancik 1988. Lack of cross-protection between citrus exocortis viroid and citrus viroids associated with mild symptoms in Etrog citron. Abstracts, 5th Int. Congr. Plant Pathol., Kyoto, Japan, Au~.20-27, 1988, p. 69. Garnsey, S. M. and L. G. Weathers 1972. Factors affecting mechanical spread of exocortis virus of citrus, p. 105-111. In: Proc. 5th Conf. IOCV. Univ. Florida Press, Gainesville. Khoury, J., R. P. Singh, A. Boucher, and D. H. Coombs 1988. Concentration and distribution of mild and severe strains of potato spindle tuber viroid in cross-protected tomato plants. Phytopathology 78: 1331-1336. La Rosa, R., G. Albanesse, A. Azzaro, F. Sesto, and F. Domina 1988. Suitability of nucleic acid analysis to diagnose viroid infections in citrus, p. 188-191. In: Proc. 10th Conf. IOCV. IOCV, Riverside. Miiller, G. W. and A. S. Costa 1977. Tristeza control in Brazil by preimmunization with mild strains. Proc. Int. Soc. Citricul- ture 3: 868-872. Niblett, C. L., E. Dickson, K. H. Fernow, R. K. Horst, and M. Zaitlin 1978. Cross-protection among four viroids. Virology 91: 198-203. Pina, J. A., N. Duran-Vila, and L. Navarro 1991. Interference of citrus viroids with cachexia symptoms on Parson's Special Mandarins, p. 206-208. In: Proc. 11th Conf. IOCV. IOCV, Riverside. Roistacher, C. N. 1988. The cachexia and xyloporosis diseases of citrus--a review, p. 116-124. In: Proc. 10th Conf. IOCV. IOCV, Riverside. Roistacher, C. N., E. C. Calavan, and R. L. Blue 1969. Citrus exocortis virus--chemical inactivation on tools, tolerance to heat and separation of isolates. Plant Dis. Rep. 53: 333-336. Roistacher, C. N., E. C. Calavan, R. Blue, L. Navarro, and R. Gonzales 1977. A new more sensitive citron indicator for detection of mild isolates of citrus exocortis viroid (CEV). Plant Dis. Rep. 61: 135. Semancik, J. S. 1988. Citrus exocortis disease - 1976 to 1986, p. 136-151. In: Proc. 10th Conf. IOCV. IOCV, Riverside. Semancik, J. S. 1991. Viroid purification and characterization, p. 233-248. In: C. N. Roistacher (ed.). Graft- transmissible diseases of citrus: Handbook for detection and diagnosis. FAO, Rome, 286 pp. Citrus Viroids 195

21. Semancik, J. S., C. N. Roistacher, and N. Duran-Vila 1988. A new viroid is the causal agent of the citrus cachexia disease, p. 125-135. In: 10th Conf. IOCV. IOCV, Riverside. 22. Semancik, J. S., D. J. Gumpf, and J. A. Bash 1992. Interference between viroids inducing exocortis and cachexia diseases of citrus. Ann. Appl. Biol. 121: 577-583. 23. Semancik, J. S., and N. Duran-Vila 1991. The grouping of citrus viroids: Additional physical and biological determinants and relationshipswithdiseaseofcitrus, p. 178-188. In:Proc. 1lthConf. IOCV. IOCV, Riverside.