Cultures of callus and sprouts

Citron (C. medica) budwood, 8 cm long and 3 mm in diameter, were dissected from plants, both healthy and diseased with CEVd isolate E811 planted in greenhouses. The plant material infected with CEVd was obtained from the collection of the DGSV-SENASICA in Mexico, from the USDA-ARS. The budwood were washed with liquid soap and faucet water, disinfested with 0.5% sodium hypochlorite solution for one minute, 70% ethanol for 50 seconds, and finally rinsed three times with sterile distilled water in aseptic conditions inside a laminar flow cabinet. The budwood were divided into 3 cm long segments, dissected longitudinally, and cultured in vitro with the cut surface in contact with the medium. The explants were cultured in each jar, 8 cm in diameter and 12 cm tall. The callus induction medium contained inorganic ^{Murashige and Skoog salts (1962)} supplemented with i-inositol (100 mg L^-1), thiamine HCl (0.1 mg L^-1), pyridoxine HCl (0.5 mg L^-1), nicotinic acid (0.5 mg L^-1), sucrose (30 g L^-1), naphthaleneacetic acid (10 mg L^-1) and 6-benzylaminopurine (0.25 mg L^-1) according to the procedure by ^{Duran et al. (1989)} and ^{Navas et al. (1995)}. The medium was adjusted to a pH of 5.7, solidified with 2.8 g L^-1 of phytagel® Sigma and sterilized in an autoclave at 116 °C for 20 minutes. The cultivars were kept in the dark at 26 °C ± 2 °C for 3 weeks until the callus was formed, those without contamination by saprophytic fungi or bacteria, were chosen. Small 3 mm portions of callus were used for the indexing in vitro once the presence of CEVd was found using RT-PCR.

Citron plant stalks were obtained with sprouts 4 to 5 cm in length, planted in a greenhouse, with no record of infections and negative for CEVd by RT-PCR. The stalks with sprouts were washed and superficially disinfested using the procedure described previously, and submerged in water for 1 minute in a streptomycin solution (200 mg L^-1) to eliminate bacterial from the tissue. The sprouts were dissected to cut stalk sections of 1.5 ± 0.5 cm long from the plants and those were cultured in a MS medium (^{Murashige and Skoog, 1962}), adding 30 g L^-1 of sucrose. The medium pH was adjusted to 5.7 and dispensed in 25 ml in test tubes and sterilized in an autoclave at 116 °C for 20 minutes. The medium was solidified with 3.0 g L^-1 of phytagel^® (Sigma). The sprouts were incubated at 25 ± 2 °C with a photoperiod of 12 hours and after 48 hours of incubation, the sprouts without contamination by saprophytic fungi or bacteria were selected to indexing.

Indexing in vitro

The sprouts were inoculated by a lateral cutting, 2-3 mm long, tongue-shaped graft with callus. The inoculated sprouts were transferred to new culture tubes solid sterile MS medium, and were incubated for 50 days at 25 ±2 °C with a photoperiod of 16 hours. The blades used during inoculation were disinfested between treatments by immersion in a 1 % sodium hypochlorite solution for 15 seconds to prevent cross-infections (^{Kovalskaya and Hammond, 2014}). During the test, the number of sprouts with symptoms were recorded, and the height above the sprout was measured using a digital caliper in the laminar flow cabinet. The presence of CEVd in the sprouts was detected by the chain reaction of polymerase with reverse transcriptase (RT-PCR).

Indexing in vivo

Citron indicator plants with negative CEVd diagnoses by RT-PCR were inoculated with bud grafts and non-grafted plants were used as a control. The grafted buds came from plants previously diagnosed as healthy of infected with CEVd. During the setup of the test, the plants were trimmed to a height of 30 cm and all their leaves were eliminated for grafting to help new sprouts develop. The bud graft was carried out following the method by ^{Roistacher (1991)}. The blades used during inoculation were disinfested between treatments and repetitions by submerging them in a 1 % sodium hypochlorite solution for one minute to prevent cross-infections (^{Kovalskaya and Hammond, 2014}). The plants were planted individually in 4L plastic pots and kept in a greenhouse at 28-32 °C. For one year and four months, a record was kept of symptoms described by ^{Roistacher (1991)} and ^{Bernard et al. (2009)}, the presence of CEVd was detected using RT-PCR and the height was measured from ground level to the tallest sprout.

Detection of CEVd by RT-PCR

The citron budwood used as a source of plant material for the culture of callus and sprouts were diagnosed for the presence or absence of CEVd one week before placing the material. Indexed sprouts were diagnosed 60 days after inoculation and citron plants of the in vivo indexing were diagnosed one week before and 180 days after grafting. The presence of CEVd in callus, buds, sprouts indexed in vitro and in a greenhouse was confirmed by RT-PCR. Total RNA was extracted with the reagents and the protocol by RNeasy Plant Mini^® by Qiagen and the cDNA synthesis was carried out with the enzyme SuperScrip^TM II Reverse Transcriptase (Invitrogen Corp.) following manufacturer instructions. The amplification of viroidal cDNA by PCR was carried out in primers CEVd1 5’-CCC TGA AGG ACT TCT TCC CC- 3’ and CEVd2 5’-ATC CCC GGG GAA ACC TGG AGG AA -3’ that amplify a fragment of 371 pb (^{Yang et al., 1992}) according to conditions described by Bernad et al. (2009) using Platinum® Taq polymerase high fidelity (Invitrogen Corp.). The products of the reaction were analyzed by electrophoresis in a 2 % agarose gel stained with ethidium bromide (0.5 µg ml^-1), visualized under UV light and photodocumented in a Gel Doc EZ system^® (Biorad). The fragments obtained in the RT-PCR were sequenced in an Applied Biosystem Corp Genetic Analyzer 3100. The sequences were analyzed using the program Chromas V. 2.5.0, aligned with Clustal W in the program Mega v.7.0.14 (^{Kumar et al., 2016}), and a consensual sequence was obtained using the program Seaview v.4.5.4 (^{Gouy et al., 2010}). The consensual sequence was compared with the original isolation sequence and with sequences deposited in the National center for biotechnology information (^{NCBI, 2010}) through the option BLASTn ver 2.3.0 (^{Altschul et al., 1990}).

Sensitivity, specificity, and accuracy of the RT-PCR: The parameters related to the diagnosis by RT-PCR were calculated for the treatments grafted with tissue infected with CEVd in each indexing procedure according to ^{Papayiannis, 2014}. The calculation formulas were as follows: sensitivity= (positive samples / positive samples+ false negatives)* 100; specificity= (negative samples / negative samples + false positives)* 100; accuracy= (negative samples + positives/ negatives+ positives+ false negatives+ false positives)* 100.

Indexing efficiency: The indexing efficiency (Ɵ) was determined as the probability of the expression of symptoms (P) in samples inoculated with infected tissue (n) with a level of confidence of 95 %, according to the methodology by ^{Vidalakis et al. (2004)}. Efficiency was calculated with the formula Ɵ= 1- (1-P_L) ⁿ using the lower probability limit by Clopper and Pearson (P_L= 0.7935 calculated for 16 units) and n= 4 as the minimum number of samples recommended by ^{Roistacher (1991)} and ^{Vidalakis et al. (2004)} for the diagnosis of CEVd by indexing.

Experimental design and statistical analysis: We used an experimental design of fields divided in time by each indexing condition (in vitro and in vivo). The largest field was for the treatments, and the small field, the days after inoculation (dai). The biological indexing in vitro of CEVd was composed of four experiments, separated in time, which included three treatments with four repetitions: the inoculation of sprouts with callus infected by CEVd (T1), inoculation with a healthy callus (T2) and y a non-inoculated sprout as a control (Ta) with a total of 48 samples analyzed. The variables evaluated were the incidence of sprouts with symptoms at 10, 15 y 20 dai and accumulated height from the initial height at 10, 20, 30 and 40 dai.

The in vivo biological indexing consisted in evaluating four experiments with three treatments: plant inoculated with buds infected by CEVd (T3), plant inoculated with buds without evidence of infection by CEVd by RT-PCR and indexing (T4) and non-inoculated plant as controls (Tb) with a total of 48 samples analyzed and four repetitions per treatment. The variables evaluated were the incidence at 40, 50, and 60 days after inoculation (dai) and accumulated height from the initial height at 10, 20, 30, 40, and 180 dai. The data of the variables underwent an analysis of variance and Tukey comparison of averages test, with a trust level of 95 % using the ^{Statistical Analysis System ver 9.0 (2002)}. The area under the disease progress curve was calculated for infected sprouts and plants using the trapeze method in the AUDPC program ver 1.1 (^{Mora and Acevedo, 2016}).

Height, incidence, and ABCP

Highly significant differences (p ≤ 0.01) were observed between treatments, time (dai) and the interaction treatment* time for the variables evaluated in both indexing procedures (Table 2). The average accumulated height of the treatments in vitro was 35.8 mm for the control; 32.1 mm for sprouts grafted with healthy callus and 19.3 mm for sprouts grafted with infected callus; whereas for indexing in the winter, these were 156.5 mm for the control, 162.9 mm for plants grafted with uninfected buds, and 106.7 mm for plants grafted with infected buds. Significant differences were observed in the heights of the sprouts in vitro in all the evaluation dates. In plants indexed in greenhouses, significant differences (p≤ 0.05) were found 10 days after inoculation, and highly significant after 180 days. A roughly consistent increase in height was observed in time in the treatments inoculated with healthy callus or buds with respect to the control, while the height of the infected treatments was 46 % lower (Figure 2A-B). The differences in height of the inoculated sprouts and the controls of the in vitro indexing observed in this study are similar to results obtained by ^{Černi et al. (2012)}, who compared the heights of Gynura aurantiaca sprouts infected by CEVd planted in vitro with greater statistical values for height in healthy sprouts. In the in vivo indexing, the height of plants grafted with infected buds was 42% lower than the control, symptom related to the infection with CEVd in citrus plants (^{Kovalskaya and Hammond, 2014}). The height differences between treatments seem to be related to the differential expression of auxins and cytokinins, including the function of apical and lateral meristems, the plant immunity processes and the response to biotic and abiotic stresses (^{Argueso et al., 2010}; ^{Naseem and Dandekar, 2012}) such as the one cause by an infection with CEVd. The auxin content may have promoted the apical dominance in healthy sprouts and plants in this work, resulting in a greater height, while the high contents of cytokinins in plants and sprouts infected by CEVd could have inhibited it (^{Laplaze et al., 2007}; ^{Shimizu et al., 2009}; ^{Naseem and Dandekar, 2012}) due to the impact of the cytokinins on the flow, distribution, and signaling of the auxin (^{Shimizu et al., 2009}).

Table 2. Average squares of the height and accumulated incidence in two indexing conditions. 

FV: factor of variation, GL: degrees of freedom, Trat: treatment, CV: coefficient of variation, **: highly significant (p ≤ 0.01).

Figure 2. Comparison of averages of height and incidence curve in sprouts planted in vitro (A, C) and plants grafted in vivo (B, D). Different letters in each column are statistically different with Tukey (p=0.01). T1: sprouts grafted with infected callus, T2: sprouts grafted with healthy callus, Ta: non-grafted sprouts, T3: plants grafted with infected bud, T4: plants grafted with healthy bud, Tb: non-grafted plants. The line and model (C, D) represent the adjustment with a quadratic regression. 

The incidence on the sprouts planted in vitro increased with time, reaching the highest rate 20 dai by grafting, in comparison with the progress curve of the disease in vivo (Figure 2C-D), where the highest rate was reached on day 60 after grafting. The in vitro indexing decreased 75 % in the time required for the appearance of typical symptoms of the disease after inoculation compared with the tests of the indexing in vivo, which required another 30 days. A detailed evaluation of the time needed to carry out an indexing could include the time of formation of callus in vitro (3 weeks). This study only considered the variable dai as the moment as of which pathogen and host are in contact for the expression of symptoms. Although it is true that indexing in vitro requires time to develop callus that will act as inoculants, it is possible to use another type of tissue such as the cortex of infected plants (^{Kapari et al., 2008}). Also, the presence of CEVd in sprouts and callus planted in vitro could contribute to develop new lines of research of aspects related to the pathogenicity of the viroids, since infected callus cultures are easy to preserve and multiply, and it is a useful tissue when the sample needs to be homogenized. Future investigations that contribute to determine if callus infected with CEVd can infect plants grown in greenhouses could be useful in traditional in vivo indexing.

The disease progress was adjusted to a quadratic regression model, where y= -0.0009x²+ 0.0693x- 0.0057, R²= 0.994 with p≤ 0.05 and AUDPC= 10.781 for the indexing in vitro, while for the indexing in vivo they were y= 0.0004x2+ 0.0048x- 0.0061, R²= 0.947 not significant, and the AUDPC = 9.844. The Area under the disease progress curve method helped obtain a better comparison of the environments in which the viroid developed in regard to the quadratic regression method que did not result significant in the in vivo indexing, due to the irregular shape of the data in regard to time. The highest initial incidence in vitro during the appearance of the disease represented a lower tolerance to the development of the CEVd symptoms in comparison with the initial incidence in vivo. The use of the same indicator variety in both indexing indicates that the concentration and the size of the host between indexing (5-30 cm) could be the main factors that influence the development of the disease. The speed with which the symptoms appear in vitro could be due to the time required by the viroids to replicate and move in the inoculated plant (^{Kapari et al., 2008}). Possibly, a higher concentration derived from the size of the sprouts (5 cm) in regard to the size of the plants (30 cm) favored the quick development of symptoms in vitro in this study. It has been proven that the temperatures (24-40 °C) the biotests undergo contributes to increasing the concentration of the viroid for the expression of symptoms of CEVd (^{Škorić et al., 2001}; ^{Bernard et al., 2009}; ^{Bani et al., 2015}) potentiating in smaller hosts. Another factor implied in the infection processes by viroids is the long distance movement inside the phloem that helps reach leaves, roots, and organs far away from those initially infected (^{Bani et al., 2010}). In this investigation, the distance travelled by CEVd in plants with roots of the in vivo indexing was greater than in sprouts planted in vitro, delaying the infection process, and consequently, the expression of symptoms.

Indexing efficiency

The results of this study indicate that both in vitro and in vivo indexing were 99 % efficient (Table 1). An indexing protocol with a sufficient amount of biological indicators can be considered efficient if at least one biological indicator expresses symptoms. The amount of biological indicators is determined by the type of isolation, the type of infection or coinfection with other pathogens and the ability to have an efficiency of at least 99% with a trust of 95 % (^{Vidalakis et al., 2004}). The sprouts planted in vitro grafted with infected callus showed symptoms of epinasty, growth reduction, leaves with reduced sizes, rugged, and with dry tips (Figure 3), coinciding with reports by ^{Kapari et al. (2008)} for sprouts grafted with cortex infected with CEVd. No symptoms were observed in the sprouts grafted with healthy callus or in control sprouts. The plants grafted with buds infected with CEVd showed symptoms of slight to severe epinasty, leaves wrinkled and twisted to the reverse with light to dark brown cracks in petiole and branches, blisters in the petiole and reduced growth, which coincide with CEVd transmission reports of biotests for C. medica Arizona 861-S plants grown in greenhouses (^{Bani et al., 2015}; ^{Lin et al., 2015a}). The use of the crop in vitro in this study proved to be useful to carry out the transmission and development of characteristic CEVd symptoms with an efficiency similar to the in vivo indexing procedures, as reported in other studies for diseases related to viroids in economically important plants (^{El-Dougdoug et al., 2010}; ^{Mahfouze et al., 2010}; ^{Černi et al., 2012}). The presence of symptoms in the samples can simplify the molecular diagnosis of the viroids (^{Camps et al., 2014}) since they distribute themselves homogenously in the plant tissues (^{Bani et al., 2015}; ^{Lin et al., 2015b}) and it is difficult to determine is the sampled tissue corresponds to the infected tissue, unless the expression of symptoms serves as an indicator.

Figure 3. Symptoms related to Citrus exocortis viroid (CEVd) in Citrus medica sprouts inoculated in vitro (A epinastia and B: dry tips) and in plants planted in vivo (D: epinastia and E: cracks on the reverse). Sprouts (C: without symptoms) and control plants (F: without symptoms). 

The evaluation of the efficiency of indexing methods can provide tools to personalize the processes to specific needs. The conditions, as well as the number of indicators used in the diagnosis, can be adjusted based on the probability of the expression of symptoms to minimize the overall cost of the diagnosis (^{Vidalakis et al., 2004}). The in vitro conditions open up a possibility of improvement from the current dynamics of indexing applicable in the propagative material certification programs for citrus plants to quickly discriminate the material infected by CEVd. Viroidal diseases in germplasm banks are prevented with the joint diagnosis of indexing and RT-PCR tests. The compliance of certification programs established by the European and Mediterranean Plant Protection Organization (EPPO) and the North American Plant Protection Organization (NAPPO) has resulted in an effective control of diseases caused by viroids (^{Barba et al., 2003}; ^{Gergerich et al., 2015}). The in vitro indexing procedure developed in this study is another addition to the efforts by the germplasm quarantine and certification programs with alternative methods to the traditional diagnosis that minimize the risks of epidemics caused by viroids.