Citrus Cachexia Viroid, a New Viroid of Citrus: Relationship to Viroids of the Exocortis Disease Complex

Transcription

1 J. gen. Virol. (1988), 69, Printed in Great Britain 3059 Key words: citrus cachexia viroid/exocortis disease/viroids Citrus Cachexia Viroid, a New Viroid of Citrus: Relationship to Viroids of the Exocortis Disease Complex By J. S. SEMANCIK, 1. C. N. ROISTACHER, 1 R. RIVERA-BUSTAMANTE 1 AND N. DURAN-VILA 2 1Department of Plant Pathology and Cell Interaction Research Group, University of California, Riverside, Calijbrnia 92521, U.S.A. and 2Instituto Valenciano de lnvestigaciones Agrarias, Moncada, Valencia, Spain (Accepted 4 July 1988) SUMMARY Recovery of highly purified citrus cachexia viroid (CCaV) was accomplished by serial elution following CF-11 cellulose chromatography of a 2 i-licl-soluble nucleic acid preparation. The alternative herbaceous host, cucumber (Cucumis sativus cv. Suyo), yielded greater quantities of the viroid than the highest yielding citrus host, citron (Citrus medica cv. Etrog). A randomly primed cdna probe to CCaV purified from cucumber reacted positively to extracts from citron and cucumber inoculated with the same isolate of CCaV. When tested against a broad range of other citrus viroids, the CCaV cdna hybridized to only one, CV-IIa, which has been identified as the causal agent of a mild form of the citrus exocortis disease. Because of the apparent homology between the nucleotide sequences of CV-IIa and CCaV, and a size difference of only five to ten nucleotides, these RNAs can be considered as members of a common subgroup of citrus viroids. These two viroids have been classified by bioassay reactions as the causal agents of two distinct types of citrus disease, an 'exocortis-like' syndrome and cachexia. The properties of and relationships between these two members of the citrus viroid II group and the definition of the exocortis and cachexia (xyloporosis) diseases are presented. INTRODUCTION The importance of viroid-induced diseases of citrus has been recognized since the citrus exocortis viroid (CEV) was identified as a type example for this group of small, pathogenic RNA molecules (Semancik & Weathers, 1972; Semancik, 1979). Until 1985, CEV remained the only well described viroid known to affect citrus, although a viroid aetiology had been proposed for the cachexia disease (Roistacher et al., 1983). A possible relationship between cachexia (Childs, 1950) and the xyloporosis disease found in the Mediterranean region has been speculated (Childs, 1952; Roistacher, 1988). With the report of a new viroid distinct from CEV but inducing moderate to severe exocortislike symptoms on citron (Citrus medica) and consequently named the citron variable viroid (CVaV) (Schlemmer et al., 1985), and the identification of a number of additional viroid RNAs (Duran-Vila et al., 1986, 1988) previously described as mild strains of CEV, it now appears that citrus harbours the largest collection of viroids of any single plant group. A proposal has been made for the ordering of this collection of over 12 citrus viroids into five groups using the parameters of molecular size, nucleotide sequence homology and host reaction (Duran-Vila et al., 1988; N. Duran-Vila et al., unpublished data; Semancik, 1988). Progress in the definition of this large number of citrus viroids has been greatly facilitated by recognition of the ability of citron cv. Etrog to serve as a common host for all the citrus viroids. A sensitive seedling selection of citron, Arizona 861-S 1, has functioned as the principal vehicle for indexing exocortis disease for the past 10 years (Roistacher et al., 1977). This selection of citron is also a symptomless host for the citrus cachexia viroid (CCaV) (Semancik et al., 1988) SGM

2 3060 J.s. SEMANCIK AND OTHERS Comparative properties of CCaV, the causal agent of the cachexia disease (Semancik, 1986; Semancik et al., 1988), and a citrus viroid, CV-IIa, which induces a mild form of the exocortis disease, are presented here. METHODS Plant materials. The cachexia disease isolate, designated Ca 902, was derived from an old line navel orange field source in The mild exocortis disease isolate, designated E 819, was derived from a Washington navel orange field source in These isolates were chosen for investigation since each was shown to contain a single viroid only. The causal agents of these diseases as well as all the other citrus viroids utilized in these studies were maintained in symptomless sweet orange seedlings as part of the University of California (Riverside, Ca., U.S.A.) citrus virus collection. These viroid reservoir hosts indexed negative for all other known citrus diseases. The citrus viroids were transmitted by bud inoculation to Arizona 86 I-S1 citron. Bud inoculation of citron with the mild exocortis source, E 819, results in a very mild leaf tip browning reaction and petiole wrinkling, while the cachexia source (Ca 902) produces no visible reaction. Unless otherwise indicated, nucleic acid extracts and viroid purifications were made from citron. Cucumber (Cucumis sat&us cv. Suyo) was inoculated with a 2 i-licl-soluble nucleic acid preparation by needle puncture of the hypocotyl through an inoculum drop as the first true leaves were emerging. After 3 weeks, the plants were trimmed to two or three nodes and all foliage was removed. After a further 3 weeks, the regrowth was collected for extraction. This process was repeated two or three times or until the new growth was severely retarded. Positive identification of the cachexia disease agent was by the characteristic browning reaction associated with gumming induced in the stem and bark of Parsons Special mandarin tissues when forced on rough lemon rootstock (Roistacher et al., 1973). Infection of the bioassay plants was by bud inoculation of the rough lemon rootstock or by slashing the citron tissue with a razor blade moistened in inoculum. Bioassay plants are normally monitored for 1 to 2 years for evidence of a positive reaction of cachexia. Extraction of nucleic acids and viroid purification. Tip tissue consisting of young leaves and stems was collected and either extracted as fresh tissue or frozen in liquid nitrogen, powdered and stored at - 20 C until extraction. Homogenization was performed in a medium containing water-saturated phenol, Tris-HC1 buffer ph 8.9, SDS, EDTA and mercaptoethanol as previously reported (Semancik et al., 1975; Duran-Vila et al., 1986). Total nucleic acid was concentrated by ethanol precipitation from the aqueous phase formed after centrifugation of the tissue homogenate. A 2 M-LiCl-soluble fraction was then collected and concentrated by ethanol precipitation and centrifugation. Serial elution of the LiCl-soluble fraction from CF-11 cellulose was employed for separation and concentration of some groups of viroids (Semancik, 1986). Viroids were isolated using the sequential 5 ~ PAGE system described in the following section. The purified viroid RNA was visualized in the final denaturing gel after staining with ethidium bromide rather than silver. Gel segments containing individual viroid bands were excised and the viroid was recovered by electroelution using an IBI Model UEA unidirectional electroelutor (Semancik et al., 1988). Analysis and characterization ofviroids. Detection of viroid RNA was by sequential 5~ PAGE under standard and denaturing conditions as described by Semancik & Harper (1984). Following the first PAGE, a section of the gel defining the viroid zone between CEV and 7S RNA (Rivera-Bustamante et al., 1986) was cut and placed on the top of the second 5 ~ gel containing 8 M-urea for denaturing PAGE. Viroid RNA was viewed on the second gel following silver staining (Igloi, 1983). This procedure has proved invaluable for the detection of citrus viroids (Duran-Vila et al., 1986, 1988) and is employed in Fig. 2, 3 and 10. For nucleotide homology studies, viroids were electrotransferred from denaturing gels onto nylon-based membranes (Nytran) using an LKB Transphor apparatus. A randomly primed CCaV cdna probe was made essentially by the procedure of Maniatis et al. (1982) using cloned Moloney murine leukaemia virus reverse transcriptase. Hybridization conditions were as reported by Garger et al. (1983). RESULTS Purification of the citrus cachexia viroid Although citron has been recognized as a common host in defining the citrus viroid groupings (Duran-Vila et al., 1988), significant differences can be observed in the rate of accumulation of the different viroids in infected citron. Whereas the bulk of citrus viroids can be readily detected in extracts from infected citron (Duran-Vila et al., 1986), CCaV is very difficult to resolve from the healthy plant background nucleic acid after PAGE (Fig. 1). The fact that CCaV is the only member of the citrus viroid family that is symptomless in citron may reflect an impaired rate of replication and/or accumulation.

3 Cachexia disease and the viroid complex (a) 1 2 (b) CEVc ~CEV 7/ CCaVc~ --7S RNA CCaV 1_ ~.~; : ~--7S RNA Fig. 1 Fig. 2 Fig. 1. Polyacrylamide (5 ~) gel containing 8 M-urea after denaturing PAGE of 2 M-LiCl-soluble nucleic acid fractions from healthy (lane 1) and CCaV-inoculated (lane 2) citrons after silver staining. The gel pictured is the second in the sequential PAGE procedure and results from removal of the section as defined by CEV and 7S RNA from the first PAGE separation and subsequent analysis of its contents by denaturing PAGE as depicted in Fig. 2. Fig. 2. Non-denaturing PAGE (a) of the single electrophoretic species of purified CCaV (lane 1) and a 2 M-LiCl-soluble extract from CEV-infected citron (lane 2) used to define the section between arrows that was removed and applied to a second gel and run under denaturing (8 M-urea) conditions (b). The bands from the non-denaturing gel would normally be stained with non-immobilizing ethidium bromide, but are stained with silver here for illustration purposes. The circular (c) and linear (1) forms of CCaY (lane 3) and CEV (lane 4) were observed after silver staining of the denaturing PAGE gel. Nevertheless, with repeated harvesting of new flush growth from the same CCaV-inoculated citron plants maintained under warm glasshouse conditions, and combining viroid recovered from several sets of denaturing gels, it has been possible to accumulate sufficient CCaV for partial characterization. The final purified viroid RNA species migrates as a single electrophoretic component under standard conditions of PAGE (Fig. 2, lane 1), and separates into the characteristic circular and linear forms of the viroid RNA (Fig. 2, lane 3) when subjected to denaturing PAGE in the presence of 8 M-urea. An estimated molecular size of approx. 300 nucleotides for CCaV was obtained by comparison of its migration relative to that of the 371 nucleotide CEV (Fig. 2, lanes 2 and 4). The highly purified CCaV RNA has been demonstrated to be transmissible in the symptomless carrier, citron, and to produce the characteristic cachexia disease reaction when bud-inoculated to its bioassay host, Parsons Special mandarin (Semancik et al., 1988). The viroid RNA accumulated to a much greater concentration in cucumber than in citron. Partially purified nucleic acid extracts from cucumber were subjected to elution from CF-11 cellulose in a graded ethanol series. The 25 ~ ethanol-ste buffer eluant was highly enriched in CCaV (Fig. 3). This manipulation greatly improved the recovery of viroid, resulting in quantities sufficient to function as a template in the production of a cdna probe.

4 3062 J.s. SEMANCIK AND OTHERS (a) (-- --CEV --7S (.-- --CEV Fig. 3. Polyacrylamide (5~) gels after nondenaturing PAGE (a), stained with ethidium bromide, and denaturing PAGE in the presence of 8 M-urea (b) stained with silver. Fractions of a 2 M-LiCl-soluble nucleic acid preparation from CCaV-inoculated citron were obtained by serial elution from CF-I1 cellulose by a series of buffered solutions with diminishing ethanol concentrations (lane 1, 30~; lane 2, 25~; lane 3, 20 ~; lane 4, 15 ~; lane 5, 10 ~; lane 6, 5 ~; lane 7, 0H). The section of gel (a) between the arrows as defined by CEV and 7S RNA contained in the standard lane (lane 8) was analysed by denaturing PAGE (b). Elution of CCaV is observed in the fractions containing 25 to 15 ~ ethanol (lanes 2 to 4). (a) 1 (b) Fig. 4. A section of an ethidium bromide-stained polyacrylamide (5~) gel (a) containing 8 M-urea after sequential PAGE and denaturing PAGE as in Fig. 2, with nucleic acid preparations from: citron inoculated with the Ca 902 cachexia disease isolate (lane 1), CCaV inoculum purified from infected cucumber (lane 2), healthy citron (lane 3), citron inoculated with purified CCaV inoculum as in lane 2 (lane 4), healthy cucumber (lane 5), cucumber infected with a purified CCaV inoculum as in lane 2 (lane 6) and a CEV standard (lane 7). Autoradiograph (b) of matching pattern (lanes 1 to 7) after electrotransfer onto Nytran membrane and hybridization with a 3Zp-labelled cdna probe to CCaV. A randomly primed CCaV cdna probe was constructed to verify that the viroid purified from cucumber was the same as the original source of cachexia disease agent obtained from the sweet orange reservoir. Comparable nucleic acid samples from citron, bud-inoculated with the sweet orange cachexia isolate, Ca 902, as well as citron and cucumber inoculated with purified CCaV were electrotransferred directly from gels containing 8 M-urea and hybridized with 32p labelled cdna to CCaV. The autoradiograph presented in Fig. 4 clearly confirms the identity in electrophoretic migration and sequence homology of the purified CCaV used as inoculum for citron and

5 Cachexia disease and the viroid complex CEV (a) (b) CV-IIa CV-IIb --CCaV CCaV) Fig. 5 Fig. 6 Fig. 5. Polyacrylamide (5%) gel containing 8 M-urea, stained with silver after processing by sequential PAGE as in Fig. 2. Nucleic acid extracts from citron infected with the mild exocortis isolate E 819 containing CV-IIa (lane 1), and cachexia isolate Ca 902 containing CCaV or CV-IIb (lane 3); identical samples as in lanes 1 and 3 were mixed before electrophoresis (lane 2); CEV was used as a standard (lane 4). Fig. 6. Polyacrylamide (5%) gel containing 8 M-urea, stained with silver after processing by sequential PAGE as in Fig. 2. Nucleic acid extracts from healthy citrons (lane 1) and citrons inoculated with Ca 902 (lane 2) containing CCaV (CV-IIb), a mild 'exocortis' isolate E 819 (lane 3) containing CV-IIa, and the 'citron variable' viroid source, E 821 (lane 4) containing CV-Ib, CV-IIa, CV-IIb and CV-IIIb, were eluted from CF-11 cellulose by step gradients of STE buffer containing 25% (a) and 0% (b) ethanol. cucumber with the original cachexia disease isolate derived from the sweet orange reservoir. Confirmation of the identity of the biological activity of the CCaV isolated from cucumber with the cachexia agent found in Ca 902 has been accomplished by transmission of the viroid purified from cucumber to citron and subsequent positive bioassay on Parsons Special mandarin. PAGE and chromatography analyses of CCa V and CV-IIa According to the description of citrus viroids made by Duran-Vila et al. (1988) and Semancik (1988), only two viroids were included in the Group II class. These two viroids, designated CV- IIa and CV-IIb, displayed very similar electrophoretic migration rates, corresponding to molecules of about 300 to 310 nucleotides, as well as a similar host range and similar properties on elution from CF-11 cellulose. When individually purified sources of CV-IIa (Fig. 5, lane 1) and CCaV (Fig. 5, lane 3) were subjected to coelectrophoresis as an artificial mixture under denaturing conditions in the same channel (Fig. 5, lane 2) the doublet visible indicated a slight but clear distinction in the migration rates and, accordingly, the molecular size of the two viroids. The position of CCaV relative to that of CV-IIa was identical to that expected for CV-IIb under comparable conditions. This analysis indicated that CCaV and CV-IIb constitute the same molecular entity. The strong homology reaction of CCaV cdna with CV-IIb presented later (Fig. 7) further supports this conclusion. The two very similar viroids, CV-IIa and CCaV, have been implicated as the causal agents of what are considered as two distinct diseases of citrus.

6 3064 J. S. SEMANCIK AND OTHERS Ca) 1 (b) CCaV) Fig. 7. A section of an ethidium bromide-stained polyacrylamide (5%) gel (a) containing 8 M-urea after sequential PAGE and denaturing PAGE as in Fig. 2, with 2 M-LiCl-soluble nucleic acid fractions from citron containing the CVs indicated; CEV and CV-Ia (lane 1), CV-Ib and CV-IIIb (lane 2), CV-/IIa and CV-IV (lane 3), CV-IIa (lane 4), and CV-IIb or CCaV (lane 5). Autoradiograph (b) of matching pattern after electrotransfer to Nytran membrane and hybridization with 3zp-labelled cdna CCaV probe. To investigate further the extent of this relationship, pure viroid isolates containing only CV- IIa or CCaV as well as an isolate, E 821, containing the citron variable viroid source (Schlemmer et al., 1985), and now known to comprise a mixture of four viroids, were analysed by sequential elution from CF-11 cellulose. This procedure has been utilized for the detection of viroid-like RNA present in low concentrations as well as in distinguishing among different viroids (Semancik, 1986). When samples eluted from cellulose as a 25 ~ ethanol-ste buffer solution were compared in denaturing PAGE with those eluted by STE buffer alone, both CV-IIa present in the E 819 source and CCaV present in the Ca 902 source were concentrated in the 25~ ethanol-ste fraction (Fig. 6). In addition, a similar fraction from the mixed viroid isolate, E 821, contained both Group II viroids. Particularly striking is the clear segregation of the Group II viroids from the other viroids present in the E 821 isolate in response to analytical CF-11 chromatography. Although both CV-IIa and CCaV can be eluted with 25 ~ ethanol-ste, CV-Ib as well as CV- IIIb require a lower ethanol concentration for elution and are evident only in the STE buffer fraction. After this analysis, it was determined that the E 821 isolate is also positive in the Parsons Special mandarin bioassay for the cachexia disease (unpublished data), thus confirming the presence of CCaV in this mixed viroid source. Nucleotide sequence homology between CCaV and other citrus viroids Nucleotide sequence homology as determined by molecular hybridization with specific cdna probes constructed from citrus viroids has been essential in establishing the general groupings of the citrus viroids (Duran-Vila et al., 1988; Semancik, 1988). When the CCaV cdna probe, as described above and employed in Fig. 4, was tested against a wide range of citrus viroids, only CV-IIa (Fig. 7b, lane 4) and CV-IIb (Fig. 7b, lane 5) reacted positively and with a similarly high intensity. Thus, what appears to be an exclusive and highly homologous relationship exists between CV-IIa and CCaV (CV-IIb). Possible biological relationships between CV-IIa and CCaV With the physical characterization data presented above, an implicit structural and conformational relationship between CV-IIa and CCaV (CV-IIb) can be proposed. Host range specificity and symptom expression in the Etrog citron utilized in defining the biological activity of citrus viroids (Duran-Vila et al., 1988) were tested to discriminate between members of the same group. Table I summarizes some of the physical properties of the Group II citrus viroids as well as the reactions on the two bioassay hosts for the exocortis and cachexia diseases.

7 Cachexia disease and the viroid complex 3065 Table 1. Comparative properties of the Group H citrus viroids Bioassay reactiont Isolate* Primary source PAGE band CCaV cdna r Exocortis ~ Cachexia E 819 Washington navel orange CV-IIa Ca 902 Old line navel orange CV-IIb (CCaV) * E indicates an exocortis disease source; Ca indicates a cachexia disease source. t Exocortis disease is bioassayed by leaf symptoms on citron (C. medica). Cachexia disease is bioassayed by browning reaction associated with gumming induced in the stem and bark in Parsons Special mandarin when forced on rough lemon root-stock. Although the citron isolate E 819 which contains CV-IIa has been designated as an exocortislike disease isolate, the distinct tip-browning of leaves, petiole wrinkling and browning of the citron bioassay host constitutes an unusual response for an exocortis source arid certainly the mildest reaction of all known exocortis isolates (Roistacher et al., 1977). In fact, this reaction is very difficult to reproduce unless ideal environmental conditions are available. Therefore, in most cases, the response of citron to CV-IIa is exceedingly difficult to distinguish from its response to CCaV, even though both viroids are present in similar detectable titres in this host. Evaluation of the root-stock and scion components of the bioassay system for the cachexia disease was accomplished by comparing a nucleic acid extract from infected citron with extracts from new flush tissues from Parsons Special mandarin scion displaying a strong cachexia reaction and from the rough lemon root-stock on which the mandarin was grafted. After denaturing PAGE, the characteristic CCaV band was visible only in the extracts from infected citron, suggesting an extremely low rate of replication and/or accumulation in mandarin and rough lemon. Therefore, the responses of the two indexing hosts to CV-IIa and CCaV are quite distinct. Citron remains a host that can maintain detectable titres of both viroids and a weak symptom expression to CV-IIa, while Parsons Special mandarin produces a strong response to CCaV, but in the absence of detectable levels of the viroid. Effect of temperature and mixed infection on CCa V replication It has been recognized that temperature can dramatically influence the incidence of viroid infection (S/inger, 1972) as well as the optimal conditions for the growth of viroid-containing cells in culture (Marton et al., 1982). The viroid structure is not only resistant to thermal inactivation, but it is also generally accepted that elevated temperatures favour viroid replication. These properties have almost become diagnostic indicators of infection of plants by viroids. Observations of the effect of temperature on viroid accumulation have been made for the most part with plant species which display a symptomatic response to viroid infection. Since Etrog citron remains symptomless during CCaV infection, a study was initiated to determine the influence of temperature on the accumulation of CCaV in this host as compared to the accumulation of CEV, a viroid which induces a severe stunting response in citron. Fig. 8 presents PAGE (a) and denaturing PAGE (b) patterns of extracts from CCaV- and CEV-infected citrons grown in the glasshouse under relatively hot (22 to 38 C) and cool (17 to 29 C) conditions of minimum and maximum temperatures, respectively. The growth.of uninoculated citron was noticeably retarded (approx. 50~) by the low temperature regime. Nevertheless, the nucleic acid preparations from plants grown under cool conditions displayed less DNA polydispersity in the gel (Fig. 8a, lanes 2 and 4) than those from plants grown under hot conditions (Fig. 8 a, lanes 1 and 3). Samples of CEV and CCaV were equalized as pairs on a extracted fresh weight basis. As anticipated, a greater concentration of CEV accumulated in plants grown under hot (Fig. 8, lanes 3) rather than cool (Fig. 8, lanes 4) conditions. This response is especially visible in the

8 3066 J. S. SEMANCIK AND OTHERS (a) 1 (b) i 3 CEV --CEV jccav --7S (---. RNA CCaV Fig. 8. Polyacrylamide (5 ~) gels after non-denaturing PAGE (a), stained with ethidium bromide, and subsequent denaturing PAGE (8 M-urea) (b), stained with silver. The section of the gel (a) between the arrows was analysed on the gel (b). Nucleic acid extracts (2 M-LiCl-soluble fraction) were made from cachexia (lanes 1, 2) and 'exocortis (lanes 3, 4) disease' citrons grown under high (22 to 38 C; lanes 1 and 3) or low (17 to 29 C; lanes 2 and 4) regimes of minimum-maximum temperatures. ethidium bromide-stained gel (Fig. 8a). The concentration of CCaV under both conditions appeared to be essentially equivalent or even slightly greater in plants grown under cool conditions (Fig. 8b, lanes 1 and 2), a response uncommon in viroid infection. A direct comparison cannot be made between the extracts from plants containing distinct viroids grown at either high or low temperature, because the CEV-containing gel samples were made approximately half as concentrated as the CCaV samples to avoid overstaining of the CEV band in the denaturing gel. DISCUSSION With the recent advances in the detection of viroids and the increase in the reported numbers of viroids described, the general grouping of citrus plant species has been found to harbour the largest collection of naturally occurring viroids. However, only two diseases, exocortis and cachexia (xyloporosis), have thus far been either defined or suggested to result from viroid infection in citrus. From this, significant questions of major importance to citriculture present themselves such as, what the implications of this large reservoir of more than 12 viroid species are to plant development and productivity, and what physical and biological relationships exist among the citrus viroids. CEV, as the well characterized type citrus viroid, has long been recognized as the causal agent of the exocortis disease. With this report, a second citrus viroid, the citrus cachexia viroid (CCaV), has now been identified and demonstrated to be the causal agent of the cachexia disease. Comparative properties including molecular size, conformation and nucleotide sequence homology suggest that CCaV is closely related to CV-IIa and is probably identical to CV-IIb in the scheme of Duran-Vila et al. (1988). Rapid indexing techniques employing PAGE and molecular hybridization offer the potential to shorten the existing 1 to 2 year cachexia disease bioassay procedure. It should be cautioned, however, that most field samples analysed contained a mixture of viroids, and if CV-IIa is

9 Cachexia d&ease and the viroid complex 3067 present the similarity in size and sequence homology with CCaV could pose potential sources of false positive reactions in rapid detection procedures such as dot blot hybridizations or PAGE with crude samples. Therefore, the bioassay indexing procedure for the ultimate identification of the cachexia disease should be retained to confirm the results of any rapid detection method. Although exocortis and cachexia have been demonstrated as quite distinct diseases of citrus based on distinct field symptomatology, the high degree of nucleotide sequence homology between the causal agents, CV-IIa and CCaV, suggests the possibility of a much closer relationship between these diseases. The basic question that remains is whether the responses of bioassay plants are sufficient to establish the existence of distinct diseases and whether fundamental relationships between diseases are minimized by the acceptance of apparently different host reactions. In the case of CCaV and CV-IIa two very similar molecules have been indicated as distinct pathogenic agents basically on the symptom reaction of citron for the bioassay of the 'exocortis' disease and mandarin for the 'cachexia' disease. Since the ethanol-dependent elution of viroids from CF-11 cellulose appears to be based on molecular conformation (Semancik, 1986), it is possible to suggest also a similarity in this parameter between CV-IIa and CCaV. Because viroid RNA sequences are not translated (Hall et al., 1974; Semancik et al., 1977) and, thus, biological effects on the host are not exerted via viroid-specified proteins, the property of molecular conformation is of special significance for viroids. It would not seem reasonable that CCaV (CV-IIb) is related to CV-IIa by being a deletion variant which somehow acquired an additional capacity, i.e. the ability to cause cachexia in the mandarin bioassay host, in spite of a net reduction in total genetic information. The precise effect on conformation caused by the postulated loss of approx, five to ten nucleotides from CV- IIa can only be conjectured at this time. Nevertheless, these data support the view that molecular conformation and not the total genetic capacity as judged by nucleotide content may be the more significant factor in the expression of biological activity of viroids. With a total nucleotide range of about 300 to 305, both of these Group II citrus viroids are in the size range of the hop stunt viroid (HSV) and cucumber pale fruit viroid (CPFV) and may be similar to the HSV-related viroid of 302 nucleotides recently isolated from Etrog citron (Sano et al., 1986, 1988). A further relationship can be suggested by the transmission and the symptom expression of CV-IIa and CCaV in cucumber, a common host of HSV and CPFV. The response of CCaV replication to temperature is unlike that normally expected for viroids. No significant increase in the accumulation of CCaV in citron could be detected with increase in growing temperatures, whereas CEV replication in the same host is enhanced by high temperatures. Since CEV can be incorporated into the nucleic acid population of tomato cells in culture (Lin & Semancik, 1985) and even enhance the viability of CEV-containing cells (Marton et al., 1982), the absence of any demonstrable symptom expression in citron containing CCaV may also reflect a significant level of coordinated synthesis of viroid with host nucleic acid. The bioassay reaction for CCaV in the indicator, Parsons Special mandarin, is enhanced by high temperatures, but only when CCaV is indexed as a pure viroid species (J. A. Pina, personal communication). When a mixture of citrus viroids, including both the cachexia and those citrus viroids producing a symptom response on citron and collectively known as exocortis disease agents, is bioassayed on Parsons Special mandarin, a more complex picture emerges. Index readings are more consistent in response to viroid mixtures containing CCaV when bioassay plants are incubated at lower temperatures, whereas the CCaV detection rate is lower when bioassay plants are maintained at higher temperatures. This effect may result from an interference or competition by the additional viroid(s) in the mixture with the cachexia bioassay. Temperature mediation of this effect might be explained if CCaV replication is not seriously affected by temperature while the replication of the other citrus viroids, as with CEV, is preferentially enhanced at higher temperatures. It can also be predicted that the closely related Group II viroid, CV-IIa, would be probably the most effective of the citrus viroids to accomplish the interference with the cachexia bioassay. It should be cautioned that this interference phenomenon should not be equated with a classical crossprotection reaction since no report of this type of reaction exists for viroid infection of citrus.

10 3068 J. S. SEMANCIK AND OTHERS The excellent assistance of Mr J. Bash with the bioassay procedures and Kathy Harper with illustration and manuscript preparation are gratefully acknowledged. These studies were supported in part by a grant from the U.S.-Spain Joint Committee for Scientific and Technological Cooperation. REFERENCES CHILDS, J. F. L. (1950). The cachexia disease of Orlando tangelo. Plant Disease Reporter 34, CHILDS, J. F. L. (1952). Cachexia disease: its bud transmission and relation to xyloporosis and tristeza. Phytopathology 42, DURAN-VILA, N., FLORES, R. & SEMANCIK, I. S. (1986). Characterization of viroid-like RNAs associated with the citrus exocortis syndrome. Virology 150, DURAN-VILA, N., PINA, J. A., BALLESTER, J. F., JUAREZ, J., ROISTACHER, C. N., RIVERA-BUSTAMANTE, R. & SEMANCIK, J. S. (1988). The citrus exocortis disease: a complex of viroid RNAs. Proceedings of the International Organization of Citrus Virologists 10, GARGER, S. l., TURPEN, T., CARRINGTON, J. C., MORRIS, T. J., DODDS, I. A. & GRILL, L. K. (1983). Rapid detection of plant viruses by dot blot hybridization. Plant Molecular Biology Reporter 1, HALL, T. C., WEPPRICH, R. K., DAVIES, J. W., WEATHERS, L. G. & SEMANCIK, J. S. (1974). Functional distinctions between the ribonucleic acids from citrus exocortis viroid and plant viruses: cell-free translation and aminoacylation reactions. Virology 61, IGLOI, G. L. (1983). A silver stain for the detection of nanogram amounts of trna following two-dimensional electrophoresis. Analytical Biochemistry 134, LIN, J. J. & SEMANCIK, J. S. (1985). Coordination between host nucleic acid metabolism and citrus exocortis viroid turnover. Virus Research 3, MANIATIS, T., FRITSCH, E. V. & SAMBROOK, T. (1982). Molecular Cloning: A Laboratory Manual, pp , New York: Cold Spring Harbor Laboratory. MARTON, L., DURAN-VILA, N., LIN, L-L & SEMANCIK, J. S. (1982). Properties of cell cultures containing the citrus exocortis viroid. Virology 122, RIVERA-BUSTAMANTE, R., GIN, R. & SEMANCIK, J. S. (1986). Enhanced resolution of circular and linear molecular forms of viroid and viroid-like RNA by electrophoresis in a discontinuous-ph system. AnalyticalBiochemistry 156, ROISTACHER, C. N. (1988). Cachexia and xyloporosis diseases. A review. Proceedings of the International Organization of Citrus Virologists 10, ROISTACHER, C. N., BLUE, R. L. & CALAVAN, E. C. (1973). A new test for citrus cachexia. Citrograph 58, ROISTACHER, C. N., CALAVAN, E. C., BLUE, R. L., NAVARRO, L. & GONTALES, R. (1977). A new more sensitive citron indicator for detection of mild isolates of citrus exocortis viroid (CEV). Plant Disease Reporter 53, ROISTACHER,. N., GUMPF, D. J., NAUER, E. M. & GONZALES, R. (1983). Cachexia disease: virus or viroid. Citrograph 68, S)~NGER, H. L. (1972). An infectious and replicating RNA of low molecular weight: the agent of the exocortis disease of citrus. Advances in Bioscience 8, SANO, T., HATAYA, T., SASAKI, A. & SHIKATA, T. (1986). Etrog citron is latently infected with hop stunt viroid-like RNA. Proceedings of the Japan Academy 62, SANO, T., HATAYA, T. & SHIKATA, T. (1988). Complete nucleotide sequence of a viroid isolated from Etrog citron, a new member of hop stunt viroid group. Nucleic Acids Research 16, 347. SCHLEMMER, A., ROISTACHER, C. N. & SEMANCIK, J. S. (1985). A unique, infectious RNA which causes symptoms in citron typical of citrus exocortis disease. Phytopathology 75, SEMANCIK, J. S. (1979). Small pathogenic RN A in plants - the viroids. Annual Review of Phytopathology 17, SEMANClg, J. S. (1986). Separation of viroid RNAs by cellulose chromatography indicating conformational distinctions. Virology 155, SEMANCIK, J. S. (1988). The citrus exocortis disease. A review Proceedings of the International Organization of Citrus Virologists 10, SE~t~NCIK, J. S. & HARPER, K. L. (1984). Optimal conditions for cell-free synthesis of citrus exocortis viroid and the question of specificity of RNA polymerase activity. Proceedings of the National Academy of Sciences, U.S.A. 81, SEMANCIK, J. S. & WEATHERS, L. G. (1972). Exocortis disease : evidence for a new species of'infectious' low molecular weight RNA in plants. Nature New Biology 237, SEMANCIK, J. S., MORRIS, r. L, WEATHERS, L. G., ROROORF, n. F. & KEARNS, D. R. (1975). Physical properties of a minimal infectious RNA (viroid) associated with the exocortis disease. Virology 63, SEMANCIK, J. S., CONEJERO, V. & GERHART, J. (1977). Citrus exocortis viroid: survey of protein synthesis in Xenopus laevis oocytes following addition of viroid RNA. Virology 80, SEMANCIK, J. S., ROISTACHER, C. N. & DURAN-VILA, N. (1988). A new viroid is the causal agent of the citrus cachexia disease. Proceedings of the International Organization of Citrus Virologists 10, (Received 3 February 1988)